In this notebook, we'll build a character-wise RNN trained on Anna Karenina, one of my all-time favorite books. It'll be able to generate new text based on the text from the book.
This network is based off of Andrej Karpathy's post on RNNs and implementation in Torch. Also, some information here at r2rt and from Sherjil Ozair on GitHub. Below is the general architecture of the character-wise RNN.
In [1]:
import time
from collections import namedtuple
import numpy as np
import tensorflow as tf
First we'll load the text file and convert it into integers for our network to use. Here I'm creating a couple dictionaries to convert the characters to and from integers. Encoding the characters as integers makes it easier to use as input in the network.
In [2]:
with open('anna.txt', 'r') as f:
text=f.read()
vocab = sorted(set(text))
vocab_to_int = {c: i for i, c in enumerate(vocab)}
int_to_vocab = dict(enumerate(vocab))
encoded = np.array([vocab_to_int[c] for c in text], dtype=np.int32)
Let's check out the first 100 characters, make sure everything is peachy. According to the American Book Review, this is the 6th best first line of a book ever.
In [3]:
text[:100]
Out[3]:
'Chapter 1\n\n\nHappy families are all alike; every unhappy family is unhappy in its own\nway.\n\nEverythin'
And we can see the characters encoded as integers.
In [4]:
encoded[:100]
Out[4]:
array([31, 64, 57, 72, 76, 61, 74, 1, 16, 0, 0, 0, 36, 57, 72, 72, 81,
1, 62, 57, 69, 65, 68, 65, 61, 75, 1, 57, 74, 61, 1, 57, 68, 68,
1, 57, 68, 65, 67, 61, 26, 1, 61, 78, 61, 74, 81, 1, 77, 70, 64,
57, 72, 72, 81, 1, 62, 57, 69, 65, 68, 81, 1, 65, 75, 1, 77, 70,
64, 57, 72, 72, 81, 1, 65, 70, 1, 65, 76, 75, 1, 71, 79, 70, 0,
79, 57, 81, 13, 0, 0, 33, 78, 61, 74, 81, 76, 64, 65, 70], dtype=int32)
Since the network is working with individual characters, it's similar to a classification problem in which we are trying to predict the next character from the previous text. Here's how many 'classes' our network has to pick from.
In [5]:
len(vocab)
Out[5]:
83
Here is where we'll make our mini-batches for training. Remember that we want our batches to be multiple sequences of some desired number of sequence steps. Considering a simple example, our batches would look like this:
We have our text encoded as integers as one long array in encoded. Let's create a function that will give us an iterator for our batches. I like using generator functions to do this. Then we can pass encoded into this function and get our batch generator.
The first thing we need to do is discard some of the text so we only have completely full batches. Each batch contains $N \times M$ characters, where $N$ is the batch size (the number of sequences) and $M$ is the number of steps. Then, to get the number of batches we can make from some array arr, you divide the length of arr by the batch size. Once you know the number of batches and the batch size, you can get the total number of characters to keep.
After that, we need to split arr into $N$ sequences. You can do this using arr.reshape(size) where size is a tuple containing the dimensions sizes of the reshaped array. We know we want $N$ sequences (n_seqs below), let's make that the size of the first dimension. For the second dimension, you can use -1 as a placeholder in the size, it'll fill up the array with the appropriate data for you. After this, you should have an array that is $N \times (M * K)$ where $K$ is the number of batches.
Now that we have this array, we can iterate through it to get our batches. The idea is each batch is a $N \times M$ window on the array. For each subsequent batch, the window moves over by n_steps. We also want to create both the input and target arrays. Remember that the targets are the inputs shifted over one character. You'll usually see the first input character used as the last target character, so something like this:
y[:, :-1], y[:, -1] = x[:, 1:], x[:, 0]
where x is the input batch and y is the target batch.
The way I like to do this window is use range to take steps of size n_steps from $0$ to arr.shape[1], the total number of steps in each sequence. That way, the integers you get from range always point to the start of a batch, and each window is n_steps wide.
Exercise: Write the code for creating batches in the function below. The exercises in this notebook will not be easy. I've provided a notebook with solutions alongside this notebook. If you get stuck, checkout the solutions. The most important thing is that you don't copy and paste the code into here, type out the solution code yourself.
In [15]:
def get_batches(arr, n_seqs, n_steps):
'''Create a generator that returns batches of size
n_seqs x n_steps from arr.
Arguments
---------
arr: Array you want to make batches from
n_seqs: Batch size, the number of sequences per batch
n_steps: Number of sequence steps per batch
'''
# Get the number of characters per batch and number of batches we can make
characters_per_batch = n_seqs * n_steps
n_batches = len(arr) // characters_per_batch
# Keep only enough characters to make full batches
arr = arr[:n_batches * characters_per_batch]
# Reshape into n_seqs rows
arr = np.reshape(arr, (n_seqs, -1))
for n in range(0, arr.shape[1], n_steps):
# The features
x = arr[:, n:n+n_steps]
# The targets, shifted by one
y = np.zeros_like(x)
y[:, :-1] = x[:, 1:]
y[:, -1] = x[:, 0]
yield x, y
Now I'll make my data sets and we can check out what's going on here. Here I'm going to use a batch size of 10 and 50 sequence steps.
In [16]:
batches = get_batches(encoded, 10, 50)
x, y = next(batches)
In [17]:
print('x\n', x[:10, :10])
print('\ny\n', y[:10, :10])
x
[[31 64 57 72 76 61 74 1 16 0]
[ 1 57 69 1 70 71 76 1 63 71]
[78 65 70 13 0 0 3 53 61 75]
[70 1 60 77 74 65 70 63 1 64]
[ 1 65 76 1 65 75 11 1 75 65]
[ 1 37 76 1 79 57 75 0 71 70]
[64 61 70 1 59 71 69 61 1 62]
[26 1 58 77 76 1 70 71 79 1]
[76 1 65 75 70 7 76 13 1 48]
[ 1 75 57 65 60 1 76 71 1 64]]
y
[[64 57 72 76 61 74 1 16 0 0]
[57 69 1 70 71 76 1 63 71 65]
[65 70 13 0 0 3 53 61 75 11]
[ 1 60 77 74 65 70 63 1 64 65]
[65 76 1 65 75 11 1 75 65 74]
[37 76 1 79 57 75 0 71 70 68]
[61 70 1 59 71 69 61 1 62 71]
[ 1 58 77 76 1 70 71 79 1 75]
[ 1 65 75 70 7 76 13 1 48 64]
[75 57 65 60 1 76 71 1 64 61]]
If you implemented get_batches correctly, the above output should look something like
x
[[55 63 69 22 6 76 45 5 16 35]
[ 5 69 1 5 12 52 6 5 56 52]
[48 29 12 61 35 35 8 64 76 78]
[12 5 24 39 45 29 12 56 5 63]
[ 5 29 6 5 29 78 28 5 78 29]
[ 5 13 6 5 36 69 78 35 52 12]
[63 76 12 5 18 52 1 76 5 58]
[34 5 73 39 6 5 12 52 36 5]
[ 6 5 29 78 12 79 6 61 5 59]
[ 5 78 69 29 24 5 6 52 5 63]]
y
[[63 69 22 6 76 45 5 16 35 35]
[69 1 5 12 52 6 5 56 52 29]
[29 12 61 35 35 8 64 76 78 28]
[ 5 24 39 45 29 12 56 5 63 29]
[29 6 5 29 78 28 5 78 29 45]
[13 6 5 36 69 78 35 52 12 43]
[76 12 5 18 52 1 76 5 58 52]
[ 5 73 39 6 5 12 52 36 5 78]
[ 5 29 78 12 79 6 61 5 59 63]
[78 69 29 24 5 6 52 5 63 76]]although the exact numbers will be different. Check to make sure the data is shifted over one step for y.
Below is where you'll build the network. We'll break it up into parts so it's easier to reason about each bit. Then we can connect them up into the whole network.
First off we'll create our input placeholders. As usual we need placeholders for the training data and the targets. We'll also create a placeholder for dropout layers called keep_prob. This will be a scalar, that is a 0-D tensor. To make a scalar, you create a placeholder without giving it a size.
Exercise: Create the input placeholders in the function below.
In [31]:
def build_inputs(batch_size, num_steps):
''' Define placeholders for inputs, targets, and dropout
Arguments
---------
batch_size: Batch size, number of sequences per batch
num_steps: Number of sequence steps in a batch
'''
# Declare placeholders we'll feed into the graph
inputs = tf.placeholder(tf.int32, [batch_size, num_steps], name='inputs')
targets = tf.placeholder(tf.int32, [batch_size, num_steps], name='targets')
# Keep probability placeholder for drop out layers
keep_prob = tf.placeholder(tf.float32, name='keep_prob')
return inputs, targets, keep_prob
Here we will create the LSTM cell we'll use in the hidden layer. We'll use this cell as a building block for the RNN. So we aren't actually defining the RNN here, just the type of cell we'll use in the hidden layer.
We first create a basic LSTM cell with
lstm = tf.contrib.rnn.BasicLSTMCell(num_units)
where num_units is the number of units in the hidden layers in the cell. Then we can add dropout by wrapping it with
tf.contrib.rnn.DropoutWrapper(lstm, output_keep_prob=keep_prob)
You pass in a cell and it will automatically add dropout to the inputs or outputs. Finally, we can stack up the LSTM cells into layers with tf.contrib.rnn.MultiRNNCell. With this, you pass in a list of cells and it will send the output of one cell into the next cell. Previously with TensorFlow 1.0, you could do this
tf.contrib.rnn.MultiRNNCell([cell]*num_layers)
This might look a little weird if you know Python well because this will create a list of the same cell object. However, TensorFlow 1.0 will create different weight matrices for all cell objects. But, starting with TensorFlow 1.1 you actually need to create new cell objects in the list. To get it to work in TensorFlow 1.1, it should look like
def build_cell(num_units, keep_prob):
lstm = tf.contrib.rnn.BasicLSTMCell(num_units)
drop = tf.contrib.rnn.DropoutWrapper(lstm, output_keep_prob=keep_prob)
return drop
tf.contrib.rnn.MultiRNNCell([build_cell(num_units, keep_prob) for _ in range(num_layers)])
Even though this is actually multiple LSTM cells stacked on each other, you can treat the multiple layers as one cell.
We also need to create an initial cell state of all zeros. This can be done like so
initial_state = cell.zero_state(batch_size, tf.float32)
Below, we implement the build_lstm function to create these LSTM cells and the initial state.
In [39]:
def build_lstm(lstm_size, num_layers, batch_size, keep_prob):
''' Build LSTM cell.
Arguments
---------
keep_prob: Scalar tensor (tf.placeholder) for the dropout keep probability
lstm_size: Size of the hidden layers in the LSTM cells
num_layers: Number of LSTM layers
batch_size: Batch size
'''
### Build the LSTM Cell
# Use a basic LSTM cell
lstm = tf.contrib.rnn.BasicLSTMCell(lstm_size)
# Add dropout to the cell outputs
drop = tf.contrib.rnn.DropoutWrapper(lstm, output_keep_prob=keep_prob)
# Stack up multiple LSTM layers, for deep learning
cell = tf.contrib.rnn.MultiRNNCell([drop] * num_layers)
initial_state = cell.zero_state(batch_size, tf.float32)
return cell, initial_state
Here we'll create the output layer. We need to connect the output of the RNN cells to a full connected layer with a softmax output. The softmax output gives us a probability distribution we can use to predict the next character, so we want this layer to have size $C$, the number of classes/characters we have in our text.
If our input has batch size $N$, number of steps $M$, and the hidden layer has $L$ hidden units, then the output is a 3D tensor with size $N \times M \times L$. The output of each LSTM cell has size $L$, we have $M$ of them, one for each sequence step, and we have $N$ sequences. So the total size is $N \times M \times L$.
We are using the same fully connected layer, the same weights, for each of the outputs. Then, to make things easier, we should reshape the outputs into a 2D tensor with shape $(M * N) \times L$. That is, one row for each sequence and step, where the values of each row are the output from the LSTM cells. We get the LSTM output as a list, lstm_output. First we need to concatenate this whole list into one array with tf.concat. Then, reshape it (with tf.reshape) to size $(M * N) \times L$.
One we have the outputs reshaped, we can do the matrix multiplication with the weights. We need to wrap the weight and bias variables in a variable scope with tf.variable_scope(scope_name) because there are weights being created in the LSTM cells. TensorFlow will throw an error if the weights created here have the same names as the weights created in the LSTM cells, which they will be default. To avoid this, we wrap the variables in a variable scope so we can give them unique names.
Exercise: Implement the output layer in the function below.
In [41]:
def build_output(lstm_output, in_size, out_size):
''' Build a softmax layer, return the softmax output and logits.
Arguments
---------
lstm_output: List of output tensors from the LSTM layer
in_size: Size of the input tensor, for example, size of the LSTM cells
out_size: Size of this softmax layer
'''
# Reshape output so it's a bunch of rows, one row for each step for each sequence.
# Concatenate lstm_output over axis 1 (the columns)
seq_output = tf.concat(lstm_output, axis=1)
# Reshape seq_output to a 2D tensor with lstm_size columns
x = tf.reshape(seq_output, (-1, in_size))
# Connect the RNN outputs to a softmax layer
with tf.variable_scope('softmax'):
# Create the weight and bias variables here
softmax_w = tf.Variable(tf.truncated_normal((in_size, out_size), stddev=0.1), name='w')
softmax_b = tf.Variable(tf.zeros(out_size), name='b')
# Since output is a bunch of rows of RNN cell outputs, logits will be a bunch
# of rows of logit outputs, one for each step and sequence
logits = tf.matmul(x, softmax_w) + softmax_b
# Use softmax to get the probabilities for predicted characters
out = tf.nn.softmax(logits, name='predictions')
return out, logits
Next up is the training loss. We get the logits and targets and calculate the softmax cross-entropy loss. First we need to one-hot encode the targets, we're getting them as encoded characters. Then, reshape the one-hot targets so it's a 2D tensor with size $(M*N) \times C$ where $C$ is the number of classes/characters we have. Remember that we reshaped the LSTM outputs and ran them through a fully connected layer with $C$ units. So our logits will also have size $(M*N) \times C$.
Then we run the logits and targets through tf.nn.softmax_cross_entropy_with_logits and find the mean to get the loss.
Exercise: Implement the loss calculation in the function below.
In [49]:
def build_loss(logits, targets, lstm_size, num_classes):
''' Calculate the loss from the logits and the targets.
Arguments
---------
logits: Logits from final fully connected layer
targets: Targets for supervised learning
lstm_size: Number of LSTM hidden units
num_classes: Number of classes in targets
'''
# One-hot encode targets and reshape to match logits, one row per sequence per step
y_one_hot = tf.one_hot(targets, num_classes)
y_reshaped = tf.reshape(y_one_hot, (-1, num_classes))
# Softmax cross entropy loss
loss = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits=logits, labels=y_reshaped))
return loss
Here we build the optimizer. Normal RNNs have have issues gradients exploding and disappearing. LSTMs fix the disappearance problem, but the gradients can still grow without bound. To fix this, we can clip the gradients above some threshold. That is, if a gradient is larger than that threshold, we set it to the threshold. This will ensure the gradients never grow overly large. Then we use an AdamOptimizer for the learning step.
In [27]:
def build_optimizer(loss, learning_rate, grad_clip):
''' Build optmizer for training, using gradient clipping.
Arguments:
loss: Network loss
learning_rate: Learning rate for optimizer
'''
# Optimizer for training, using gradient clipping to control exploding gradients
tvars = tf.trainable_variables()
grads, _ = tf.clip_by_global_norm(tf.gradients(loss, tvars), grad_clip)
train_op = tf.train.AdamOptimizer(learning_rate)
optimizer = train_op.apply_gradients(zip(grads, tvars))
return optimizer
Now we can put all the pieces together and build a class for the network. To actually run data through the LSTM cells, we will use tf.nn.dynamic_rnn. This function will pass the hidden and cell states across LSTM cells appropriately for us. It returns the outputs for each LSTM cell at each step for each sequence in the mini-batch. It also gives us the final LSTM state. We want to save this state as final_state so we can pass it to the first LSTM cell in the the next mini-batch run. For tf.nn.dynamic_rnn, we pass in the cell and initial state we get from build_lstm, as well as our input sequences. Also, we need to one-hot encode the inputs before going into the RNN.
Exercise: Use the functions you've implemented previously and
tf.nn.dynamic_rnnto build the network.
In [43]:
class CharRNN:
def __init__(self, num_classes, batch_size=64, num_steps=50,
lstm_size=128, num_layers=2, learning_rate=0.001,
grad_clip=5, sampling=False):
# When we're using this network for sampling later, we'll be passing in
# one character at a time, so providing an option for that
if sampling == True:
batch_size, num_steps = 1, 1
else:
batch_size, num_steps = batch_size, num_steps
tf.reset_default_graph()
# Build the input placeholder tensors
self.inputs, self.targets, self.keep_prob = build_inputs(batch_size, num_steps)
# Build the LSTM cell
cell, self.initial_state = build_lstm(lstm_size, num_layers, batch_size, self.keep_prob)
### Run the data through the RNN layers
# First, one-hot encode the input tokens
x_one_hot = tf.one_hot(self.inputs, num_classes)
# Run each sequence step through the RNN with tf.nn.dynamic_rnn
outputs, state = tf.nn.dynamic_rnn(cell, x_one_hot, initial_state=self.initial_state)
self.final_state = state
# Get softmax predictions and logits
self.prediction, self.logits = build_output(outputs, lstm_size, num_classes)
# Loss and optimizer (with gradient clipping)
self.loss = build_loss(self.logits, self.targets, lstm_size, num_classes)
self.optimizer = build_optimizer(self.loss, learning_rate, grad_clip)
Here are the hyperparameters for the network.
batch_size - Number of sequences running through the network in one pass.num_steps - Number of characters in the sequence the network is trained on. Larger is better typically, the network will learn more long range dependencies. But it takes longer to train. 100 is typically a good number here.lstm_size - The number of units in the hidden layers.num_layers - Number of hidden LSTM layers to uselearning_rate - Learning rate for trainingkeep_prob - The dropout keep probability when training. If you're network is overfitting, try decreasing this.Here's some good advice from Andrej Karpathy on training the network. I'm going to copy it in here for your benefit, but also link to where it originally came from.
Tips and Tricks
Monitoring Validation Loss vs. Training Loss
If you're somewhat new to Machine Learning or Neural Networks it can take a bit of expertise to get good models. The most important quantity to keep track of is the difference between your training loss (printed during training) and the validation loss (printed once in a while when the RNN is run on the validation data (by default every 1000 iterations)). In particular:
- If your training loss is much lower than validation loss then this means the network might be overfitting. Solutions to this are to decrease your network size, or to increase dropout. For example you could try dropout of 0.5 and so on.
- If your training/validation loss are about equal then your model is underfitting. Increase the size of your model (either number of layers or the raw number of neurons per layer)
Approximate number of parameters
The two most important parameters that control the model are
lstm_sizeandnum_layers. I would advise that you always usenum_layersof either 2/3. Thelstm_sizecan be adjusted based on how much data you have. The two important quantities to keep track of here are:
- The number of parameters in your model. This is printed when you start training.
- The size of your dataset. 1MB file is approximately 1 million characters.
These two should be about the same order of magnitude. It's a little tricky to tell. Here are some examples:
- I have a 100MB dataset and I'm using the default parameter settings (which currently print 150K parameters). My data size is significantly larger (100 mil >> 0.15 mil), so I expect to heavily underfit. I am thinking I can comfortably afford to make
lstm_sizelarger.- I have a 10MB dataset and running a 10 million parameter model. I'm slightly nervous and I'm carefully monitoring my validation loss. If it's larger than my training loss then I may want to try to increase dropout a bit and see if that helps the validation loss.
Best models strategy
The winning strategy to obtaining very good models (if you have the compute time) is to always err on making the network larger (as large as you're willing to wait for it to compute) and then try different dropout values (between 0,1). Whatever model has the best validation performance (the loss, written in the checkpoint filename, low is good) is the one you should use in the end.
It is very common in deep learning to run many different models with many different hyperparameter settings, and in the end take whatever checkpoint gave the best validation performance.
By the way, the size of your training and validation splits are also parameters. Make sure you have a decent amount of data in your validation set or otherwise the validation performance will be noisy and not very informative.
In [29]:
batch_size = 10 # Sequences per batch
num_steps = 50 # Number of sequence steps per batch
lstm_size = 128 # Size of hidden layers in LSTMs
num_layers = 2 # Number of LSTM layers
learning_rate = 0.01 # Learning rate
keep_prob = 0.5 # Dropout keep probability
This is typical training code, passing inputs and targets into the network, then running the optimizer. Here we also get back the final LSTM state for the mini-batch. Then, we pass that state back into the network so the next batch can continue the state from the previous batch. And every so often (set by save_every_n) I save a checkpoint.
Here I'm saving checkpoints with the format
i{iteration number}_l{# hidden layer units}.ckpt
Exercise: Set the hyperparameters above to train the network. Watch the training loss, it should be consistently dropping. Also, I highly advise running this on a GPU.
In [50]:
epochs = 20
# Save every N iterations
save_every_n = 200
model = CharRNN(len(vocab), batch_size=batch_size, num_steps=num_steps,
lstm_size=lstm_size, num_layers=num_layers,
learning_rate=learning_rate)
saver = tf.train.Saver(max_to_keep=100)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
# Use the line below to load a checkpoint and resume training
#saver.restore(sess, 'checkpoints/______.ckpt')
counter = 0
for e in range(epochs):
# Train network
new_state = sess.run(model.initial_state)
loss = 0
for x, y in get_batches(encoded, batch_size, num_steps):
counter += 1
start = time.time()
feed = {model.inputs: x,
model.targets: y,
model.keep_prob: keep_prob,
model.initial_state: new_state}
batch_loss, new_state, _ = sess.run([model.loss,
model.final_state,
model.optimizer],
feed_dict=feed)
end = time.time()
print('Epoch: {}/{}... '.format(e+1, epochs),
'Training Step: {}... '.format(counter),
'Training loss: {:.4f}... '.format(batch_loss),
'{:.4f} sec/batch'.format((end-start)))
if (counter % save_every_n == 0):
saver.save(sess, "checkpoints/i{}_l{}.ckpt".format(counter, lstm_size))
saver.save(sess, "checkpoints/i{}_l{}.ckpt".format(counter, lstm_size))
Epoch: 1/20... Training Step: 1... Training loss: 4.4204... 0.3138 sec/batch
Epoch: 1/20... Training Step: 2... Training loss: 4.3124... 0.0398 sec/batch
Epoch: 1/20... Training Step: 3... Training loss: 3.5444... 0.0397 sec/batch
Epoch: 1/20... Training Step: 4... Training loss: 3.4641... 0.0445 sec/batch
Epoch: 1/20... Training Step: 5... Training loss: 3.2677... 0.0423 sec/batch
Epoch: 1/20... Training Step: 6... Training loss: 3.3969... 0.0431 sec/batch
Epoch: 1/20... Training Step: 7... Training loss: 3.2000... 0.0508 sec/batch
Epoch: 1/20... Training Step: 8... Training loss: 3.2086... 0.0421 sec/batch
Epoch: 1/20... Training Step: 9... Training loss: 3.2669... 0.0446 sec/batch
Epoch: 1/20... Training Step: 10... Training loss: 3.2446... 0.0398 sec/batch
Epoch: 1/20... Training Step: 11... Training loss: 3.3109... 0.0455 sec/batch
Epoch: 1/20... Training Step: 12... Training loss: 3.1732... 0.0472 sec/batch
Epoch: 1/20... Training Step: 13... Training loss: 3.0846... 0.0425 sec/batch
Epoch: 1/20... Training Step: 14... Training loss: 3.2104... 0.0430 sec/batch
Epoch: 1/20... Training Step: 15... Training loss: 3.0916... 0.0402 sec/batch
Epoch: 1/20... Training Step: 16... Training loss: 3.1691... 0.0441 sec/batch
Epoch: 1/20... Training Step: 17... Training loss: 3.1519... 0.0477 sec/batch
Epoch: 1/20... Training Step: 18... Training loss: 3.2325... 0.0410 sec/batch
Epoch: 1/20... Training Step: 19... Training loss: 3.2152... 0.0425 sec/batch
Epoch: 1/20... Training Step: 20... Training loss: 3.1383... 0.0404 sec/batch
Epoch: 1/20... Training Step: 21... Training loss: 3.2184... 0.0394 sec/batch
Epoch: 1/20... Training Step: 22... Training loss: 3.1650... 0.0480 sec/batch
Epoch: 1/20... Training Step: 23... Training loss: 3.1928... 0.0433 sec/batch
Epoch: 1/20... Training Step: 24... Training loss: 3.1707... 0.0416 sec/batch
Epoch: 1/20... Training Step: 25... Training loss: 3.1421... 0.0412 sec/batch
Epoch: 1/20... Training Step: 26... Training loss: 3.1201... 0.0432 sec/batch
Epoch: 1/20... Training Step: 27... Training loss: 3.0508... 0.0508 sec/batch
Epoch: 1/20... Training Step: 28... Training loss: 3.1901... 0.0388 sec/batch
Epoch: 1/20... Training Step: 29... Training loss: 3.1385... 0.0425 sec/batch
Epoch: 1/20... Training Step: 30... Training loss: 3.1557... 0.0412 sec/batch
Epoch: 1/20... Training Step: 31... Training loss: 3.1511... 0.0470 sec/batch
Epoch: 1/20... Training Step: 32... Training loss: 3.1910... 0.0412 sec/batch
Epoch: 1/20... Training Step: 33... Training loss: 3.1819... 0.0431 sec/batch
Epoch: 1/20... Training Step: 34... Training loss: 3.2099... 0.0414 sec/batch
Epoch: 1/20... Training Step: 35... Training loss: 3.0961... 0.0403 sec/batch
Epoch: 1/20... Training Step: 36... Training loss: 3.2143... 0.0431 sec/batch
Epoch: 1/20... Training Step: 37... Training loss: 3.2283... 0.0420 sec/batch
Epoch: 1/20... Training Step: 38... Training loss: 3.1766... 0.0442 sec/batch
Epoch: 1/20... Training Step: 39... Training loss: 3.0873... 0.0441 sec/batch
Epoch: 1/20... Training Step: 40... Training loss: 3.1165... 0.0410 sec/batch
Epoch: 1/20... Training Step: 41... Training loss: 3.1081... 0.0445 sec/batch
Epoch: 1/20... Training Step: 42... Training loss: 3.2164... 0.0422 sec/batch
Epoch: 1/20... Training Step: 43... Training loss: 3.0885... 0.0428 sec/batch
Epoch: 1/20... Training Step: 44... Training loss: 3.1866... 0.0408 sec/batch
Epoch: 1/20... Training Step: 45... Training loss: 3.0868... 0.0417 sec/batch
Epoch: 1/20... Training Step: 46... Training loss: 3.1800... 0.0409 sec/batch
Epoch: 1/20... Training Step: 47... Training loss: 3.1451... 0.0438 sec/batch
Epoch: 1/20... Training Step: 48... Training loss: 3.0776... 0.0397 sec/batch
Epoch: 1/20... Training Step: 49... Training loss: 3.1228... 0.0410 sec/batch
Epoch: 1/20... Training Step: 50... Training loss: 3.1759... 0.0406 sec/batch
Epoch: 1/20... Training Step: 51... Training loss: 3.0936... 0.0399 sec/batch
Epoch: 1/20... Training Step: 52... Training loss: 3.1041... 0.0736 sec/batch
Epoch: 1/20... Training Step: 53... Training loss: 3.1167... 0.0558 sec/batch
Epoch: 1/20... Training Step: 54... Training loss: 3.1020... 0.0526 sec/batch
Epoch: 1/20... Training Step: 55... Training loss: 3.0486... 0.0646 sec/batch
Epoch: 1/20... Training Step: 56... Training loss: 3.0679... 0.0606 sec/batch
Epoch: 1/20... Training Step: 57... Training loss: 3.0852... 0.0494 sec/batch
Epoch: 1/20... Training Step: 58... Training loss: 3.0284... 0.0568 sec/batch
Epoch: 1/20... Training Step: 59... Training loss: 3.0548... 0.0412 sec/batch
Epoch: 1/20... Training Step: 60... Training loss: 2.9848... 0.0564 sec/batch
Epoch: 1/20... Training Step: 61... Training loss: 3.0835... 0.0933 sec/batch
Epoch: 1/20... Training Step: 62... Training loss: 3.0462... 0.0667 sec/batch
Epoch: 1/20... Training Step: 63... Training loss: 3.0026... 0.0536 sec/batch
Epoch: 1/20... Training Step: 64... Training loss: 3.0344... 0.0430 sec/batch
Epoch: 1/20... Training Step: 65... Training loss: 3.0015... 0.0423 sec/batch
Epoch: 1/20... Training Step: 66... Training loss: 3.0020... 0.0420 sec/batch
Epoch: 1/20... Training Step: 67... Training loss: 2.9822... 0.0425 sec/batch
Epoch: 1/20... Training Step: 68... Training loss: 3.0505... 0.0446 sec/batch
Epoch: 1/20... Training Step: 69... Training loss: 3.0957... 0.0387 sec/batch
Epoch: 1/20... Training Step: 70... Training loss: 2.8703... 0.0392 sec/batch
Epoch: 1/20... Training Step: 71... Training loss: 2.9697... 0.0423 sec/batch
Epoch: 1/20... Training Step: 72... Training loss: 2.9910... 0.0394 sec/batch
Epoch: 1/20... Training Step: 73... Training loss: 2.8785... 0.0473 sec/batch
Epoch: 1/20... Training Step: 74... Training loss: 2.9604... 0.0393 sec/batch
Epoch: 1/20... Training Step: 75... Training loss: 2.9384... 0.0397 sec/batch
Epoch: 1/20... Training Step: 76... Training loss: 2.9751... 0.0396 sec/batch
Epoch: 1/20... Training Step: 77... Training loss: 2.9026... 0.0441 sec/batch
Epoch: 1/20... Training Step: 78... Training loss: 2.9640... 0.0438 sec/batch
Epoch: 1/20... Training Step: 79... Training loss: 2.8856... 0.0448 sec/batch
Epoch: 1/20... Training Step: 80... Training loss: 2.7832... 0.0453 sec/batch
Epoch: 1/20... Training Step: 81... Training loss: 3.0270... 0.0534 sec/batch
Epoch: 1/20... Training Step: 82... Training loss: 2.8523... 0.0567 sec/batch
Epoch: 1/20... Training Step: 83... Training loss: 2.8655... 0.0616 sec/batch
Epoch: 1/20... Training Step: 84... Training loss: 2.7837... 0.0422 sec/batch
Epoch: 1/20... Training Step: 85... Training loss: 2.8120... 0.0418 sec/batch
Epoch: 1/20... Training Step: 86... Training loss: 2.7909... 0.0486 sec/batch
Epoch: 1/20... Training Step: 87... Training loss: 2.7633... 0.0521 sec/batch
Epoch: 1/20... Training Step: 88... Training loss: 2.6897... 0.0489 sec/batch
Epoch: 1/20... Training Step: 89... Training loss: 2.6589... 0.0474 sec/batch
Epoch: 1/20... Training Step: 90... Training loss: 2.7838... 0.0504 sec/batch
Epoch: 1/20... Training Step: 91... Training loss: 2.7257... 0.0398 sec/batch
Epoch: 1/20... Training Step: 92... Training loss: 2.8268... 0.0439 sec/batch
Epoch: 1/20... Training Step: 93... Training loss: 2.7865... 0.0489 sec/batch
Epoch: 1/20... Training Step: 94... Training loss: 2.8845... 0.0478 sec/batch
Epoch: 1/20... Training Step: 95... Training loss: 2.8134... 0.0500 sec/batch
Epoch: 1/20... Training Step: 96... Training loss: 2.6651... 0.0515 sec/batch
Epoch: 1/20... Training Step: 97... Training loss: 2.8010... 0.0487 sec/batch
Epoch: 1/20... Training Step: 98... Training loss: 2.6969... 0.0397 sec/batch
Epoch: 1/20... Training Step: 99... Training loss: 2.6533... 0.0389 sec/batch
Epoch: 1/20... Training Step: 100... Training loss: 2.6581... 0.0487 sec/batch
Epoch: 1/20... Training Step: 101... Training loss: 2.7888... 0.0533 sec/batch
Epoch: 1/20... Training Step: 102... Training loss: 2.6086... 0.0500 sec/batch
Epoch: 1/20... Training Step: 103... Training loss: 2.7221... 0.0487 sec/batch
Epoch: 1/20... Training Step: 104... Training loss: 2.5053... 0.0485 sec/batch
Epoch: 1/20... Training Step: 105... Training loss: 2.6037... 0.0498 sec/batch
Epoch: 1/20... Training Step: 106... Training loss: 2.7989... 0.0491 sec/batch
Epoch: 1/20... Training Step: 107... Training loss: 2.6868... 0.0414 sec/batch
Epoch: 1/20... Training Step: 108... Training loss: 2.6604... 0.0400 sec/batch
Epoch: 1/20... Training Step: 109... Training loss: 2.6473... 0.0410 sec/batch
Epoch: 1/20... Training Step: 110... Training loss: 2.7488... 0.0393 sec/batch
Epoch: 1/20... Training Step: 111... Training loss: 2.5691... 0.0436 sec/batch
Epoch: 1/20... Training Step: 112... Training loss: 2.6554... 0.0449 sec/batch
Epoch: 1/20... Training Step: 113... Training loss: 2.6475... 0.0433 sec/batch
Epoch: 1/20... Training Step: 114... Training loss: 2.5471... 0.0397 sec/batch
Epoch: 1/20... Training Step: 115... Training loss: 2.6617... 0.0397 sec/batch
Epoch: 1/20... Training Step: 116... Training loss: 2.6785... 0.0400 sec/batch
Epoch: 1/20... Training Step: 117... Training loss: 2.5657... 0.0441 sec/batch
Epoch: 1/20... Training Step: 118... Training loss: 2.4622... 0.0431 sec/batch
Epoch: 1/20... Training Step: 119... Training loss: 2.6098... 0.0388 sec/batch
Epoch: 1/20... Training Step: 120... Training loss: 2.5236... 0.0412 sec/batch
Epoch: 1/20... Training Step: 121... Training loss: 2.4844... 0.0399 sec/batch
Epoch: 1/20... Training Step: 122... Training loss: 2.5700... 0.0466 sec/batch
Epoch: 1/20... Training Step: 123... Training loss: 2.6462... 0.0413 sec/batch
Epoch: 1/20... Training Step: 124... Training loss: 2.5720... 0.0496 sec/batch
Epoch: 1/20... Training Step: 125... Training loss: 2.5556... 0.0543 sec/batch
Epoch: 1/20... Training Step: 126... Training loss: 2.4815... 0.0516 sec/batch
Epoch: 1/20... Training Step: 127... Training loss: 2.4441... 0.0587 sec/batch
Epoch: 1/20... Training Step: 128... Training loss: 2.5062... 0.0519 sec/batch
Epoch: 1/20... Training Step: 129... Training loss: 2.4985... 0.0405 sec/batch
Epoch: 1/20... Training Step: 130... Training loss: 2.4766... 0.0398 sec/batch
Epoch: 1/20... Training Step: 131... Training loss: 2.4856... 0.0418 sec/batch
Epoch: 1/20... Training Step: 132... Training loss: 2.5440... 0.0422 sec/batch
Epoch: 1/20... Training Step: 133... Training loss: 2.5115... 0.0458 sec/batch
Epoch: 1/20... Training Step: 134... Training loss: 2.4978... 0.0491 sec/batch
Epoch: 1/20... Training Step: 135... Training loss: 2.5948... 0.0568 sec/batch
Epoch: 1/20... Training Step: 136... Training loss: 2.3607... 0.0558 sec/batch
Epoch: 1/20... Training Step: 137... Training loss: 2.5406... 0.0548 sec/batch
Epoch: 1/20... Training Step: 138... Training loss: 2.4827... 0.0398 sec/batch
Epoch: 1/20... Training Step: 139... Training loss: 2.5343... 0.0390 sec/batch
Epoch: 1/20... Training Step: 140... Training loss: 2.5233... 0.0398 sec/batch
Epoch: 1/20... Training Step: 141... Training loss: 2.4143... 0.0453 sec/batch
Epoch: 1/20... Training Step: 142... Training loss: 2.5191... 0.0515 sec/batch
Epoch: 1/20... Training Step: 143... Training loss: 2.5645... 0.0493 sec/batch
Epoch: 1/20... Training Step: 144... Training loss: 2.4917... 0.0514 sec/batch
Epoch: 1/20... Training Step: 145... Training loss: 2.4141... 0.0504 sec/batch
Epoch: 1/20... Training Step: 146... Training loss: 2.4734... 0.0453 sec/batch
Epoch: 1/20... Training Step: 147... Training loss: 2.5478... 0.0417 sec/batch
Epoch: 1/20... Training Step: 148... Training loss: 2.3886... 0.0390 sec/batch
Epoch: 1/20... Training Step: 149... Training loss: 2.4303... 0.0403 sec/batch
Epoch: 1/20... Training Step: 150... Training loss: 2.4390... 0.0397 sec/batch
Epoch: 1/20... Training Step: 151... Training loss: 2.4864... 0.0482 sec/batch
Epoch: 1/20... Training Step: 152... Training loss: 2.4208... 0.0398 sec/batch
Epoch: 1/20... Training Step: 153... Training loss: 2.5448... 0.0387 sec/batch
Epoch: 1/20... Training Step: 154... Training loss: 2.4759... 0.0435 sec/batch
Epoch: 1/20... Training Step: 155... Training loss: 2.5724... 0.0394 sec/batch
Epoch: 1/20... Training Step: 156... Training loss: 2.5408... 0.0433 sec/batch
Epoch: 1/20... Training Step: 157... Training loss: 2.3371... 0.0403 sec/batch
Epoch: 1/20... Training Step: 158... Training loss: 2.5363... 0.0429 sec/batch
Epoch: 1/20... Training Step: 159... Training loss: 2.6215... 0.0445 sec/batch
Epoch: 1/20... Training Step: 160... Training loss: 2.4853... 0.0448 sec/batch
Epoch: 1/20... Training Step: 161... Training loss: 2.4100... 0.0458 sec/batch
Epoch: 1/20... Training Step: 162... Training loss: 2.4222... 0.0419 sec/batch
Epoch: 1/20... Training Step: 163... Training loss: 2.3604... 0.0399 sec/batch
Epoch: 1/20... Training Step: 164... Training loss: 2.3886... 0.0416 sec/batch
Epoch: 1/20... Training Step: 165... Training loss: 2.4792... 0.0410 sec/batch
Epoch: 1/20... Training Step: 166... Training loss: 2.4197... 0.0486 sec/batch
Epoch: 1/20... Training Step: 167... Training loss: 2.4095... 0.0458 sec/batch
Epoch: 1/20... Training Step: 168... Training loss: 2.5265... 0.0508 sec/batch
Epoch: 1/20... Training Step: 169... Training loss: 2.6209... 0.0499 sec/batch
Epoch: 1/20... Training Step: 170... Training loss: 2.3811... 0.0536 sec/batch
Epoch: 1/20... Training Step: 171... Training loss: 2.4600... 0.0510 sec/batch
Epoch: 1/20... Training Step: 172... Training loss: 2.4396... 0.0519 sec/batch
Epoch: 1/20... Training Step: 173... Training loss: 2.5087... 0.0402 sec/batch
Epoch: 1/20... Training Step: 174... Training loss: 2.5351... 0.0450 sec/batch
Epoch: 1/20... Training Step: 175... Training loss: 2.4909... 0.0529 sec/batch
Epoch: 1/20... Training Step: 176... Training loss: 2.3359... 0.0514 sec/batch
Epoch: 1/20... Training Step: 177... Training loss: 2.3444... 0.0503 sec/batch
Epoch: 1/20... Training Step: 178... Training loss: 2.3965... 0.0480 sec/batch
Epoch: 1/20... Training Step: 179... Training loss: 2.4834... 0.0491 sec/batch
Epoch: 1/20... Training Step: 180... Training loss: 2.3229... 0.0469 sec/batch
Epoch: 1/20... Training Step: 181... Training loss: 2.2539... 0.0487 sec/batch
Epoch: 1/20... Training Step: 182... Training loss: 2.3941... 0.0516 sec/batch
Epoch: 1/20... Training Step: 183... Training loss: 2.4052... 0.0492 sec/batch
Epoch: 1/20... Training Step: 184... Training loss: 2.3865... 0.0543 sec/batch
Epoch: 1/20... Training Step: 185... Training loss: 2.3584... 0.0527 sec/batch
Epoch: 1/20... Training Step: 186... Training loss: 2.4075... 0.0465 sec/batch
Epoch: 1/20... Training Step: 187... Training loss: 2.3269... 0.0426 sec/batch
Epoch: 1/20... Training Step: 188... Training loss: 2.3429... 0.0407 sec/batch
Epoch: 1/20... Training Step: 189... Training loss: 2.4349... 0.0441 sec/batch
Epoch: 1/20... Training Step: 190... Training loss: 2.3724... 0.0448 sec/batch
Epoch: 1/20... Training Step: 191... Training loss: 2.3737... 0.0395 sec/batch
Epoch: 1/20... Training Step: 192... Training loss: 2.2575... 0.0409 sec/batch
Epoch: 1/20... Training Step: 193... Training loss: 2.3937... 0.0517 sec/batch
Epoch: 1/20... Training Step: 194... Training loss: 2.5574... 0.0633 sec/batch
Epoch: 1/20... Training Step: 195... Training loss: 2.4531... 0.0516 sec/batch
Epoch: 1/20... Training Step: 196... Training loss: 2.4170... 0.0502 sec/batch
Epoch: 1/20... Training Step: 197... Training loss: 2.2850... 0.0480 sec/batch
Epoch: 1/20... Training Step: 198... Training loss: 2.4261... 0.0464 sec/batch
Epoch: 1/20... Training Step: 199... Training loss: 2.5210... 0.0469 sec/batch
Epoch: 1/20... Training Step: 200... Training loss: 2.4640... 0.0421 sec/batch
Epoch: 1/20... Training Step: 201... Training loss: 2.3980... 0.0533 sec/batch
Epoch: 1/20... Training Step: 202... Training loss: 2.3492... 0.0535 sec/batch
Epoch: 1/20... Training Step: 203... Training loss: 2.4700... 0.0509 sec/batch
Epoch: 1/20... Training Step: 204... Training loss: 2.4600... 0.0496 sec/batch
Epoch: 1/20... Training Step: 205... Training loss: 2.3601... 0.0443 sec/batch
Epoch: 1/20... Training Step: 206... Training loss: 2.3810... 0.0387 sec/batch
Epoch: 1/20... Training Step: 207... Training loss: 2.4243... 0.0478 sec/batch
Epoch: 1/20... Training Step: 208... Training loss: 2.4146... 0.0415 sec/batch
Epoch: 1/20... Training Step: 209... Training loss: 2.3805... 0.0391 sec/batch
Epoch: 1/20... Training Step: 210... Training loss: 2.3341... 0.0422 sec/batch
Epoch: 1/20... Training Step: 211... Training loss: 2.3579... 0.0399 sec/batch
Epoch: 1/20... Training Step: 212... Training loss: 2.3681... 0.0383 sec/batch
Epoch: 1/20... Training Step: 213... Training loss: 2.4190... 0.0585 sec/batch
Epoch: 1/20... Training Step: 214... Training loss: 2.3994... 0.0400 sec/batch
Epoch: 1/20... Training Step: 215... Training loss: 2.3526... 0.0414 sec/batch
Epoch: 1/20... Training Step: 216... Training loss: 2.3367... 0.0408 sec/batch
Epoch: 1/20... Training Step: 217... Training loss: 2.3999... 0.0390 sec/batch
Epoch: 1/20... Training Step: 218... Training loss: 2.3199... 0.0447 sec/batch
Epoch: 1/20... Training Step: 219... Training loss: 2.2657... 0.0395 sec/batch
Epoch: 1/20... Training Step: 220... Training loss: 2.3288... 0.0412 sec/batch
Epoch: 1/20... Training Step: 221... Training loss: 2.4128... 0.0386 sec/batch
Epoch: 1/20... Training Step: 222... Training loss: 2.3910... 0.0413 sec/batch
Epoch: 1/20... Training Step: 223... Training loss: 2.3892... 0.0442 sec/batch
Epoch: 1/20... Training Step: 224... Training loss: 2.3998... 0.0410 sec/batch
Epoch: 1/20... Training Step: 225... Training loss: 2.4130... 0.0438 sec/batch
Epoch: 1/20... Training Step: 226... Training loss: 2.3708... 0.0424 sec/batch
Epoch: 1/20... Training Step: 227... Training loss: 2.4184... 0.0479 sec/batch
Epoch: 1/20... Training Step: 228... Training loss: 2.4302... 0.0505 sec/batch
Epoch: 1/20... Training Step: 229... Training loss: 2.3070... 0.0496 sec/batch
Epoch: 1/20... Training Step: 230... Training loss: 2.2711... 0.0499 sec/batch
Epoch: 1/20... Training Step: 231... Training loss: 2.3262... 0.0500 sec/batch
Epoch: 1/20... Training Step: 232... Training loss: 2.2744... 0.0550 sec/batch
Epoch: 1/20... Training Step: 233... Training loss: 2.3565... 0.0428 sec/batch
Epoch: 1/20... Training Step: 234... Training loss: 2.4168... 0.0398 sec/batch
Epoch: 1/20... Training Step: 235... Training loss: 2.4853... 0.0397 sec/batch
Epoch: 1/20... Training Step: 236... Training loss: 2.4871... 0.0479 sec/batch
Epoch: 1/20... Training Step: 237... Training loss: 2.3544... 0.0433 sec/batch
Epoch: 1/20... Training Step: 238... Training loss: 2.4915... 0.0412 sec/batch
Epoch: 1/20... Training Step: 239... Training loss: 2.3658... 0.0383 sec/batch
Epoch: 1/20... Training Step: 240... Training loss: 2.2140... 0.0399 sec/batch
Epoch: 1/20... Training Step: 241... Training loss: 2.3836... 0.0390 sec/batch
Epoch: 1/20... Training Step: 242... Training loss: 2.4011... 0.0436 sec/batch
Epoch: 1/20... Training Step: 243... Training loss: 2.4400... 0.0392 sec/batch
Epoch: 1/20... Training Step: 244... Training loss: 2.4982... 0.0394 sec/batch
Epoch: 1/20... Training Step: 245... Training loss: 2.4169... 0.0397 sec/batch
Epoch: 1/20... Training Step: 246... Training loss: 2.3167... 0.0401 sec/batch
Epoch: 1/20... Training Step: 247... Training loss: 2.3106... 0.0509 sec/batch
Epoch: 1/20... Training Step: 248... Training loss: 2.3461... 0.0498 sec/batch
Epoch: 1/20... Training Step: 249... Training loss: 2.2472... 0.0518 sec/batch
Epoch: 1/20... Training Step: 250... Training loss: 2.3964... 0.0509 sec/batch
Epoch: 1/20... Training Step: 251... Training loss: 2.3711... 0.0552 sec/batch
Epoch: 1/20... Training Step: 252... Training loss: 2.4762... 0.0493 sec/batch
Epoch: 1/20... Training Step: 253... Training loss: 2.2396... 0.0415 sec/batch
Epoch: 1/20... Training Step: 254... Training loss: 2.2817... 0.0411 sec/batch
Epoch: 1/20... Training Step: 255... Training loss: 2.4318... 0.0514 sec/batch
Epoch: 1/20... Training Step: 256... Training loss: 2.3442... 0.0404 sec/batch
Epoch: 1/20... Training Step: 257... Training loss: 2.2951... 0.0404 sec/batch
Epoch: 1/20... Training Step: 258... Training loss: 2.3409... 0.0416 sec/batch
Epoch: 1/20... Training Step: 259... Training loss: 2.2666... 0.0392 sec/batch
Epoch: 1/20... Training Step: 260... Training loss: 2.2844... 0.0390 sec/batch
Epoch: 1/20... Training Step: 261... Training loss: 2.2840... 0.0462 sec/batch
Epoch: 1/20... Training Step: 262... Training loss: 2.3203... 0.0416 sec/batch
Epoch: 1/20... Training Step: 263... Training loss: 2.3109... 0.0397 sec/batch
Epoch: 1/20... Training Step: 264... Training loss: 2.3013... 0.0394 sec/batch
Epoch: 1/20... Training Step: 265... Training loss: 2.2007... 0.0394 sec/batch
Epoch: 1/20... Training Step: 266... Training loss: 2.2434... 0.0501 sec/batch
Epoch: 1/20... Training Step: 267... Training loss: 2.2602... 0.0530 sec/batch
Epoch: 1/20... Training Step: 268... Training loss: 2.3237... 0.0504 sec/batch
Epoch: 1/20... Training Step: 269... Training loss: 2.3135... 0.0516 sec/batch
Epoch: 1/20... Training Step: 270... Training loss: 2.3430... 0.0528 sec/batch
Epoch: 1/20... Training Step: 271... Training loss: 2.3632... 0.0535 sec/batch
Epoch: 1/20... Training Step: 272... Training loss: 2.2917... 0.0396 sec/batch
Epoch: 1/20... Training Step: 273... Training loss: 2.4414... 0.0480 sec/batch
Epoch: 1/20... Training Step: 274... Training loss: 2.2621... 0.0559 sec/batch
Epoch: 1/20... Training Step: 275... Training loss: 2.2999... 0.0555 sec/batch
Epoch: 1/20... Training Step: 276... Training loss: 2.2317... 0.0510 sec/batch
Epoch: 1/20... Training Step: 277... Training loss: 2.2913... 0.0497 sec/batch
Epoch: 1/20... Training Step: 278... Training loss: 2.3455... 0.0462 sec/batch
Epoch: 1/20... Training Step: 279... Training loss: 2.2204... 0.0482 sec/batch
Epoch: 1/20... Training Step: 280... Training loss: 2.2519... 0.0405 sec/batch
Epoch: 1/20... Training Step: 281... Training loss: 2.3257... 0.0418 sec/batch
Epoch: 1/20... Training Step: 282... Training loss: 2.2476... 0.0397 sec/batch
Epoch: 1/20... Training Step: 283... Training loss: 2.1650... 0.0434 sec/batch
Epoch: 1/20... Training Step: 284... Training loss: 2.4043... 0.0506 sec/batch
Epoch: 1/20... Training Step: 285... Training loss: 2.1606... 0.0410 sec/batch
Epoch: 1/20... Training Step: 286... Training loss: 2.1334... 0.0386 sec/batch
Epoch: 1/20... Training Step: 287... Training loss: 2.1413... 0.0411 sec/batch
Epoch: 1/20... Training Step: 288... Training loss: 2.4187... 0.0404 sec/batch
Epoch: 1/20... Training Step: 289... Training loss: 2.2327... 0.0585 sec/batch
Epoch: 1/20... Training Step: 290... Training loss: 2.2750... 0.0398 sec/batch
Epoch: 1/20... Training Step: 291... Training loss: 2.1671... 0.0401 sec/batch
Epoch: 1/20... Training Step: 292... Training loss: 2.2518... 0.0405 sec/batch
Epoch: 1/20... Training Step: 293... Training loss: 2.2809... 0.0420 sec/batch
Epoch: 1/20... Training Step: 294... Training loss: 2.1975... 0.0397 sec/batch
Epoch: 1/20... Training Step: 295... Training loss: 2.3280... 0.0400 sec/batch
Epoch: 1/20... Training Step: 296... Training loss: 2.2281... 0.0393 sec/batch
Epoch: 1/20... Training Step: 297... Training loss: 2.1776... 0.0395 sec/batch
Epoch: 1/20... Training Step: 298... Training loss: 2.2060... 0.0404 sec/batch
Epoch: 1/20... Training Step: 299... Training loss: 2.2105... 0.0413 sec/batch
Epoch: 1/20... Training Step: 300... Training loss: 2.2448... 0.0445 sec/batch
Epoch: 1/20... Training Step: 301... Training loss: 2.2457... 0.0432 sec/batch
Epoch: 1/20... Training Step: 302... Training loss: 2.3069... 0.0401 sec/batch
Epoch: 1/20... Training Step: 303... Training loss: 2.1740... 0.0392 sec/batch
Epoch: 1/20... Training Step: 304... Training loss: 2.1956... 0.0414 sec/batch
Epoch: 1/20... Training Step: 305... Training loss: 2.2816... 0.0434 sec/batch
Epoch: 1/20... Training Step: 306... Training loss: 2.2023... 0.0392 sec/batch
Epoch: 1/20... Training Step: 307... Training loss: 2.3276... 0.0394 sec/batch
Epoch: 1/20... Training Step: 308... Training loss: 2.2285... 0.0468 sec/batch
Epoch: 1/20... Training Step: 309... Training loss: 2.2538... 0.0497 sec/batch
Epoch: 1/20... Training Step: 310... Training loss: 2.2990... 0.0510 sec/batch
Epoch: 1/20... Training Step: 311... Training loss: 2.3662... 0.0519 sec/batch
Epoch: 1/20... Training Step: 312... Training loss: 2.2199... 0.0482 sec/batch
Epoch: 1/20... Training Step: 313... Training loss: 2.2292... 0.0497 sec/batch
Epoch: 1/20... Training Step: 314... Training loss: 2.2255... 0.0433 sec/batch
Epoch: 1/20... Training Step: 315... Training loss: 2.2950... 0.0398 sec/batch
Epoch: 1/20... Training Step: 316... Training loss: 2.3151... 0.0437 sec/batch
Epoch: 1/20... Training Step: 317... Training loss: 2.3161... 0.0398 sec/batch
Epoch: 1/20... Training Step: 318... Training loss: 2.2961... 0.0429 sec/batch
Epoch: 1/20... Training Step: 319... Training loss: 2.2380... 0.0412 sec/batch
Epoch: 1/20... Training Step: 320... Training loss: 2.2424... 0.0394 sec/batch
Epoch: 1/20... Training Step: 321... Training loss: 2.2171... 0.0392 sec/batch
Epoch: 1/20... Training Step: 322... Training loss: 2.1352... 0.0403 sec/batch
Epoch: 1/20... Training Step: 323... Training loss: 2.2779... 0.0478 sec/batch
Epoch: 1/20... Training Step: 324... Training loss: 2.2657... 0.0412 sec/batch
Epoch: 1/20... Training Step: 325... Training loss: 2.2166... 0.0386 sec/batch
Epoch: 1/20... Training Step: 326... Training loss: 2.2360... 0.0414 sec/batch
Epoch: 1/20... Training Step: 327... Training loss: 2.2214... 0.0388 sec/batch
Epoch: 1/20... Training Step: 328... Training loss: 2.2411... 0.0486 sec/batch
Epoch: 1/20... Training Step: 329... Training loss: 2.3859... 0.0415 sec/batch
Epoch: 1/20... Training Step: 330... Training loss: 2.2986... 0.0388 sec/batch
Epoch: 1/20... Training Step: 331... Training loss: 2.1323... 0.0428 sec/batch
Epoch: 1/20... Training Step: 332... Training loss: 2.3335... 0.0430 sec/batch
Epoch: 1/20... Training Step: 333... Training loss: 2.2070... 0.0455 sec/batch
Epoch: 1/20... Training Step: 334... Training loss: 2.2108... 0.0411 sec/batch
Epoch: 1/20... Training Step: 335... Training loss: 2.3034... 0.0398 sec/batch
Epoch: 1/20... Training Step: 336... Training loss: 2.3178... 0.0408 sec/batch
Epoch: 1/20... Training Step: 337... Training loss: 2.2191... 0.0395 sec/batch
Epoch: 1/20... Training Step: 338... Training loss: 2.0898... 0.0451 sec/batch
Epoch: 1/20... Training Step: 339... Training loss: 2.3126... 0.0395 sec/batch
Epoch: 1/20... Training Step: 340... Training loss: 2.2028... 0.0407 sec/batch
Epoch: 1/20... Training Step: 341... Training loss: 2.4569... 0.0389 sec/batch
Epoch: 1/20... Training Step: 342... Training loss: 2.3730... 0.0391 sec/batch
Epoch: 1/20... Training Step: 343... Training loss: 2.4108... 0.0526 sec/batch
Epoch: 1/20... Training Step: 344... Training loss: 2.2810... 0.0434 sec/batch
Epoch: 1/20... Training Step: 345... Training loss: 2.2181... 0.0402 sec/batch
Epoch: 1/20... Training Step: 346... Training loss: 2.2770... 0.0407 sec/batch
Epoch: 1/20... Training Step: 347... Training loss: 2.3131... 0.0423 sec/batch
Epoch: 1/20... Training Step: 348... Training loss: 2.2372... 0.0430 sec/batch
Epoch: 1/20... Training Step: 349... Training loss: 2.2036... 0.0390 sec/batch
Epoch: 1/20... Training Step: 350... Training loss: 2.1902... 0.0392 sec/batch
Epoch: 1/20... Training Step: 351... Training loss: 2.1738... 0.0397 sec/batch
Epoch: 1/20... Training Step: 352... Training loss: 2.2297... 0.0393 sec/batch
Epoch: 1/20... Training Step: 353... Training loss: 2.3508... 0.0500 sec/batch
Epoch: 1/20... Training Step: 354... Training loss: 2.1760... 0.0399 sec/batch
Epoch: 1/20... Training Step: 355... Training loss: 2.2836... 0.0464 sec/batch
Epoch: 1/20... Training Step: 356... Training loss: 2.2355... 0.0523 sec/batch
Epoch: 1/20... Training Step: 357... Training loss: 2.3416... 0.0464 sec/batch
Epoch: 1/20... Training Step: 358... Training loss: 2.2980... 0.0537 sec/batch
Epoch: 1/20... Training Step: 359... Training loss: 2.1753... 0.0499 sec/batch
Epoch: 1/20... Training Step: 360... Training loss: 2.2945... 0.0518 sec/batch
Epoch: 1/20... Training Step: 361... Training loss: 2.2915... 0.0403 sec/batch
Epoch: 1/20... Training Step: 362... Training loss: 2.3182... 0.0509 sec/batch
Epoch: 1/20... Training Step: 363... Training loss: 2.3311... 0.0517 sec/batch
Epoch: 1/20... Training Step: 364... Training loss: 2.1416... 0.0569 sec/batch
Epoch: 1/20... Training Step: 365... Training loss: 2.2105... 0.0500 sec/batch
Epoch: 1/20... Training Step: 366... Training loss: 2.2058... 0.0515 sec/batch
Epoch: 1/20... Training Step: 367... Training loss: 2.2811... 0.0472 sec/batch
Epoch: 1/20... Training Step: 368... Training loss: 2.3141... 0.0484 sec/batch
Epoch: 1/20... Training Step: 369... Training loss: 2.2570... 0.0506 sec/batch
Epoch: 1/20... Training Step: 370... Training loss: 2.1901... 0.0494 sec/batch
Epoch: 1/20... Training Step: 371... Training loss: 2.2157... 0.0586 sec/batch
Epoch: 1/20... Training Step: 372... Training loss: 2.2421... 0.0501 sec/batch
Epoch: 1/20... Training Step: 373... Training loss: 2.2134... 0.0412 sec/batch
Epoch: 1/20... Training Step: 374... Training loss: 2.1108... 0.0434 sec/batch
Epoch: 1/20... Training Step: 375... Training loss: 2.1951... 0.0407 sec/batch
Epoch: 1/20... Training Step: 376... Training loss: 2.2217... 0.0406 sec/batch
Epoch: 1/20... Training Step: 377... Training loss: 2.3062... 0.0440 sec/batch
Epoch: 1/20... Training Step: 378... Training loss: 2.1128... 0.0405 sec/batch
Epoch: 1/20... Training Step: 379... Training loss: 2.2398... 0.0424 sec/batch
Epoch: 1/20... Training Step: 380... Training loss: 2.0980... 0.0394 sec/batch
Epoch: 1/20... Training Step: 381... Training loss: 2.1677... 0.0433 sec/batch
Epoch: 1/20... Training Step: 382... Training loss: 2.1988... 0.0473 sec/batch
Epoch: 1/20... Training Step: 383... Training loss: 2.2845... 0.0394 sec/batch
Epoch: 1/20... Training Step: 384... Training loss: 2.3691... 0.0415 sec/batch
Epoch: 1/20... Training Step: 385... Training loss: 2.2496... 0.0403 sec/batch
Epoch: 1/20... Training Step: 386... Training loss: 2.1700... 0.0415 sec/batch
Epoch: 1/20... Training Step: 387... Training loss: 2.2301... 0.0524 sec/batch
Epoch: 1/20... Training Step: 388... Training loss: 2.2572... 0.0433 sec/batch
Epoch: 1/20... Training Step: 389... Training loss: 2.1446... 0.0408 sec/batch
Epoch: 1/20... Training Step: 390... Training loss: 2.1624... 0.0419 sec/batch
Epoch: 1/20... Training Step: 391... Training loss: 2.1074... 0.0436 sec/batch
Epoch: 1/20... Training Step: 392... Training loss: 2.2176... 0.0453 sec/batch
Epoch: 1/20... Training Step: 393... Training loss: 2.2459... 0.0401 sec/batch
Epoch: 1/20... Training Step: 394... Training loss: 2.1805... 0.0431 sec/batch
Epoch: 1/20... Training Step: 395... Training loss: 2.2284... 0.0415 sec/batch
Epoch: 1/20... Training Step: 396... Training loss: 2.1774... 0.0405 sec/batch
Epoch: 1/20... Training Step: 397... Training loss: 2.1799... 0.0463 sec/batch
Epoch: 1/20... Training Step: 398... Training loss: 2.2862... 0.0414 sec/batch
Epoch: 1/20... Training Step: 399... Training loss: 2.2120... 0.0437 sec/batch
Epoch: 1/20... Training Step: 400... Training loss: 2.1926... 0.0414 sec/batch
Epoch: 1/20... Training Step: 401... Training loss: 2.1670... 0.0486 sec/batch
Epoch: 1/20... Training Step: 402... Training loss: 2.1093... 0.0445 sec/batch
Epoch: 1/20... Training Step: 403... Training loss: 2.1328... 0.0431 sec/batch
Epoch: 1/20... Training Step: 404... Training loss: 2.1220... 0.0404 sec/batch
Epoch: 1/20... Training Step: 405... Training loss: 2.2000... 0.0420 sec/batch
Epoch: 1/20... Training Step: 406... Training loss: 2.2008... 0.0442 sec/batch
Epoch: 1/20... Training Step: 407... Training loss: 2.2384... 0.0489 sec/batch
Epoch: 1/20... Training Step: 408... Training loss: 2.1777... 0.0402 sec/batch
Epoch: 1/20... Training Step: 409... Training loss: 2.2117... 0.0392 sec/batch
Epoch: 1/20... Training Step: 410... Training loss: 2.1997... 0.0386 sec/batch
Epoch: 1/20... Training Step: 411... Training loss: 2.1596... 0.0402 sec/batch
Epoch: 1/20... Training Step: 412... Training loss: 2.2169... 0.0512 sec/batch
Epoch: 1/20... Training Step: 413... Training loss: 2.2584... 0.0441 sec/batch
Epoch: 1/20... Training Step: 414... Training loss: 2.0769... 0.0447 sec/batch
Epoch: 1/20... Training Step: 415... Training loss: 2.1324... 0.0395 sec/batch
Epoch: 1/20... Training Step: 416... Training loss: 2.1100... 0.0416 sec/batch
Epoch: 1/20... Training Step: 417... Training loss: 2.1895... 0.0456 sec/batch
Epoch: 1/20... Training Step: 418... Training loss: 2.0450... 0.0393 sec/batch
Epoch: 1/20... Training Step: 419... Training loss: 2.1401... 0.0412 sec/batch
Epoch: 1/20... Training Step: 420... Training loss: 2.1238... 0.0414 sec/batch
Epoch: 1/20... Training Step: 421... Training loss: 2.2051... 0.0393 sec/batch
Epoch: 1/20... Training Step: 422... Training loss: 2.2109... 0.0525 sec/batch
Epoch: 1/20... Training Step: 423... Training loss: 2.2476... 0.0408 sec/batch
Epoch: 1/20... Training Step: 424... Training loss: 2.1126... 0.0397 sec/batch
Epoch: 1/20... Training Step: 425... Training loss: 2.1674... 0.0389 sec/batch
Epoch: 1/20... Training Step: 426... Training loss: 2.1934... 0.0411 sec/batch
Epoch: 1/20... Training Step: 427... Training loss: 2.1595... 0.0490 sec/batch
Epoch: 1/20... Training Step: 428... Training loss: 2.1457... 0.0429 sec/batch
Epoch: 1/20... Training Step: 429... Training loss: 2.2174... 0.0425 sec/batch
Epoch: 1/20... Training Step: 430... Training loss: 2.1072... 0.0390 sec/batch
Epoch: 1/20... Training Step: 431... Training loss: 2.0471... 0.0459 sec/batch
Epoch: 1/20... Training Step: 432... Training loss: 2.1230... 0.0448 sec/batch
Epoch: 1/20... Training Step: 433... Training loss: 2.0909... 0.0526 sec/batch
Epoch: 1/20... Training Step: 434... Training loss: 2.1498... 0.0504 sec/batch
Epoch: 1/20... Training Step: 435... Training loss: 2.0573... 0.0520 sec/batch
Epoch: 1/20... Training Step: 436... Training loss: 2.1232... 0.0559 sec/batch
Epoch: 1/20... Training Step: 437... Training loss: 2.1720... 0.0508 sec/batch
Epoch: 1/20... Training Step: 438... Training loss: 1.9911... 0.0499 sec/batch
Epoch: 1/20... Training Step: 439... Training loss: 2.0401... 0.0434 sec/batch
Epoch: 1/20... Training Step: 440... Training loss: 2.2213... 0.0492 sec/batch
Epoch: 1/20... Training Step: 441... Training loss: 2.2495... 0.0407 sec/batch
Epoch: 1/20... Training Step: 442... Training loss: 2.1077... 0.0433 sec/batch
Epoch: 1/20... Training Step: 443... Training loss: 2.1152... 0.0397 sec/batch
Epoch: 1/20... Training Step: 444... Training loss: 2.0836... 0.0402 sec/batch
Epoch: 1/20... Training Step: 445... Training loss: 2.1493... 0.0458 sec/batch
Epoch: 1/20... Training Step: 446... Training loss: 2.1735... 0.0436 sec/batch
Epoch: 1/20... Training Step: 447... Training loss: 2.2417... 0.0426 sec/batch
Epoch: 1/20... Training Step: 448... Training loss: 2.0204... 0.0406 sec/batch
Epoch: 1/20... Training Step: 449... Training loss: 2.0622... 0.0436 sec/batch
Epoch: 1/20... Training Step: 450... Training loss: 2.2075... 0.0473 sec/batch
Epoch: 1/20... Training Step: 451... Training loss: 2.1539... 0.0411 sec/batch
Epoch: 1/20... Training Step: 452... Training loss: 2.0160... 0.0406 sec/batch
Epoch: 1/20... Training Step: 453... Training loss: 2.0964... 0.0406 sec/batch
Epoch: 1/20... Training Step: 454... Training loss: 2.1522... 0.0406 sec/batch
Epoch: 1/20... Training Step: 455... Training loss: 2.1533... 0.0452 sec/batch
Epoch: 1/20... Training Step: 456... Training loss: 2.0472... 0.0400 sec/batch
Epoch: 1/20... Training Step: 457... Training loss: 2.1600... 0.0433 sec/batch
Epoch: 1/20... Training Step: 458... Training loss: 2.2103... 0.0427 sec/batch
Epoch: 1/20... Training Step: 459... Training loss: 2.1955... 0.0397 sec/batch
Epoch: 1/20... Training Step: 460... Training loss: 2.1996... 0.0447 sec/batch
Epoch: 1/20... Training Step: 461... Training loss: 2.0479... 0.0431 sec/batch
Epoch: 1/20... Training Step: 462... Training loss: 2.2038... 0.0417 sec/batch
Epoch: 1/20... Training Step: 463... Training loss: 2.1315... 0.0432 sec/batch
Epoch: 1/20... Training Step: 464... Training loss: 2.1141... 0.0431 sec/batch
Epoch: 1/20... Training Step: 465... Training loss: 2.1197... 0.0447 sec/batch
Epoch: 1/20... Training Step: 466... Training loss: 2.0857... 0.0443 sec/batch
Epoch: 1/20... Training Step: 467... Training loss: 2.0741... 0.0432 sec/batch
Epoch: 1/20... Training Step: 468... Training loss: 2.0737... 0.0384 sec/batch
Epoch: 1/20... Training Step: 469... Training loss: 2.0328... 0.0425 sec/batch
Epoch: 1/20... Training Step: 470... Training loss: 2.0858... 0.0441 sec/batch
Epoch: 1/20... Training Step: 471... Training loss: 1.9935... 0.0398 sec/batch
Epoch: 1/20... Training Step: 472... Training loss: 2.2313... 0.0391 sec/batch
Epoch: 1/20... Training Step: 473... Training loss: 2.1420... 0.0393 sec/batch
Epoch: 1/20... Training Step: 474... Training loss: 2.1699... 0.0396 sec/batch
Epoch: 1/20... Training Step: 475... Training loss: 2.1237... 0.0444 sec/batch
Epoch: 1/20... Training Step: 476... Training loss: 2.0073... 0.0393 sec/batch
Epoch: 1/20... Training Step: 477... Training loss: 2.1430... 0.0388 sec/batch
Epoch: 1/20... Training Step: 478... Training loss: 1.9738... 0.0399 sec/batch
Epoch: 1/20... Training Step: 479... Training loss: 2.0115... 0.0416 sec/batch
Epoch: 1/20... Training Step: 480... Training loss: 2.1885... 0.0478 sec/batch
Epoch: 1/20... Training Step: 481... Training loss: 2.0426... 0.0392 sec/batch
Epoch: 1/20... Training Step: 482... Training loss: 2.1134... 0.0430 sec/batch
Epoch: 1/20... Training Step: 483... Training loss: 2.1132... 0.0417 sec/batch
Epoch: 1/20... Training Step: 484... Training loss: 2.0750... 0.0500 sec/batch
Epoch: 1/20... Training Step: 485... Training loss: 2.1551... 0.0486 sec/batch
Epoch: 1/20... Training Step: 486... Training loss: 2.2369... 0.0526 sec/batch
Epoch: 1/20... Training Step: 487... Training loss: 2.1447... 0.0580 sec/batch
Epoch: 1/20... Training Step: 488... Training loss: 2.0890... 0.0476 sec/batch
Epoch: 1/20... Training Step: 489... Training loss: 2.2424... 0.0484 sec/batch
Epoch: 1/20... Training Step: 490... Training loss: 2.1716... 0.0395 sec/batch
Epoch: 1/20... Training Step: 491... Training loss: 2.0808... 0.0417 sec/batch
Epoch: 1/20... Training Step: 492... Training loss: 2.1505... 0.0412 sec/batch
Epoch: 1/20... Training Step: 493... Training loss: 2.2400... 0.0414 sec/batch
Epoch: 1/20... Training Step: 494... Training loss: 2.2053... 0.0533 sec/batch
Epoch: 1/20... Training Step: 495... Training loss: 2.0100... 0.0401 sec/batch
Epoch: 1/20... Training Step: 496... Training loss: 2.0577... 0.0388 sec/batch
Epoch: 1/20... Training Step: 497... Training loss: 1.9923... 0.0409 sec/batch
Epoch: 1/20... Training Step: 498... Training loss: 2.1496... 0.0394 sec/batch
Epoch: 1/20... Training Step: 499... Training loss: 2.0070... 0.0508 sec/batch
Epoch: 1/20... Training Step: 500... Training loss: 2.3695... 0.0412 sec/batch
Epoch: 1/20... Training Step: 501... Training loss: 2.0350... 0.0410 sec/batch
Epoch: 1/20... Training Step: 502... Training loss: 2.1889... 0.0541 sec/batch
Epoch: 1/20... Training Step: 503... Training loss: 2.1577... 0.0507 sec/batch
Epoch: 1/20... Training Step: 504... Training loss: 2.0314... 0.0526 sec/batch
Epoch: 1/20... Training Step: 505... Training loss: 2.0964... 0.0507 sec/batch
Epoch: 1/20... Training Step: 506... Training loss: 2.1088... 0.0517 sec/batch
Epoch: 1/20... Training Step: 507... Training loss: 2.1552... 0.0506 sec/batch
Epoch: 1/20... Training Step: 508... Training loss: 2.1728... 0.0498 sec/batch
Epoch: 1/20... Training Step: 509... Training loss: 2.1278... 0.0396 sec/batch
Epoch: 1/20... Training Step: 510... Training loss: 2.1892... 0.0395 sec/batch
Epoch: 1/20... Training Step: 511... Training loss: 2.0946... 0.0411 sec/batch
Epoch: 1/20... Training Step: 512... Training loss: 2.3321... 0.0396 sec/batch
Epoch: 1/20... Training Step: 513... Training loss: 2.1414... 0.0454 sec/batch
Epoch: 1/20... Training Step: 514... Training loss: 2.0585... 0.0484 sec/batch
Epoch: 1/20... Training Step: 515... Training loss: 2.1505... 0.0516 sec/batch
Epoch: 1/20... Training Step: 516... Training loss: 2.1775... 0.0506 sec/batch
Epoch: 1/20... Training Step: 517... Training loss: 2.1648... 0.0464 sec/batch
Epoch: 1/20... Training Step: 518... Training loss: 2.0225... 0.0502 sec/batch
Epoch: 1/20... Training Step: 519... Training loss: 2.1277... 0.0433 sec/batch
Epoch: 1/20... Training Step: 520... Training loss: 2.0620... 0.0418 sec/batch
Epoch: 1/20... Training Step: 521... Training loss: 1.9713... 0.0477 sec/batch
Epoch: 1/20... Training Step: 522... Training loss: 2.0952... 0.0527 sec/batch
Epoch: 1/20... Training Step: 523... Training loss: 2.1116... 0.0488 sec/batch
Epoch: 1/20... Training Step: 524... Training loss: 2.0876... 0.0499 sec/batch
Epoch: 1/20... Training Step: 525... Training loss: 2.0370... 0.0513 sec/batch
Epoch: 1/20... Training Step: 526... Training loss: 2.0530... 0.0446 sec/batch
Epoch: 1/20... Training Step: 527... Training loss: 2.0211... 0.0567 sec/batch
Epoch: 1/20... Training Step: 528... Training loss: 2.0802... 0.0435 sec/batch
Epoch: 1/20... Training Step: 529... Training loss: 2.0889... 0.0401 sec/batch
Epoch: 1/20... Training Step: 530... Training loss: 2.1800... 0.0397 sec/batch
Epoch: 1/20... Training Step: 531... Training loss: 2.1282... 0.0407 sec/batch
Epoch: 1/20... Training Step: 532... Training loss: 2.0893... 0.0388 sec/batch
Epoch: 1/20... Training Step: 533... Training loss: 2.0156... 0.0488 sec/batch
Epoch: 1/20... Training Step: 534... Training loss: 2.0340... 0.0398 sec/batch
Epoch: 1/20... Training Step: 535... Training loss: 2.0975... 0.0412 sec/batch
Epoch: 1/20... Training Step: 536... Training loss: 2.1489... 0.0434 sec/batch
Epoch: 1/20... Training Step: 537... Training loss: 1.8888... 0.0483 sec/batch
Epoch: 1/20... Training Step: 538... Training loss: 2.0081... 0.0483 sec/batch
Epoch: 1/20... Training Step: 539... Training loss: 2.0262... 0.0518 sec/batch
Epoch: 1/20... Training Step: 540... Training loss: 1.9702... 0.0506 sec/batch
Epoch: 1/20... Training Step: 541... Training loss: 2.0147... 0.0502 sec/batch
Epoch: 1/20... Training Step: 542... Training loss: 2.1044... 0.0492 sec/batch
Epoch: 1/20... Training Step: 543... Training loss: 2.1022... 0.0395 sec/batch
Epoch: 1/20... Training Step: 544... Training loss: 2.0619... 0.0388 sec/batch
Epoch: 1/20... Training Step: 545... Training loss: 2.0686... 0.0395 sec/batch
Epoch: 1/20... Training Step: 546... Training loss: 1.9747... 0.0401 sec/batch
Epoch: 1/20... Training Step: 547... Training loss: 2.0730... 0.0463 sec/batch
Epoch: 1/20... Training Step: 548... Training loss: 2.0692... 0.0414 sec/batch
Epoch: 1/20... Training Step: 549... Training loss: 2.0147... 0.0501 sec/batch
Epoch: 1/20... Training Step: 550... Training loss: 2.0075... 0.0476 sec/batch
Epoch: 1/20... Training Step: 551... Training loss: 2.0424... 0.0476 sec/batch
Epoch: 1/20... Training Step: 552... Training loss: 2.3188... 0.0506 sec/batch
Epoch: 1/20... Training Step: 553... Training loss: 2.0009... 0.0480 sec/batch
Epoch: 1/20... Training Step: 554... Training loss: 2.0757... 0.0413 sec/batch
Epoch: 1/20... Training Step: 555... Training loss: 2.0857... 0.0493 sec/batch
Epoch: 1/20... Training Step: 556... Training loss: 2.1204... 0.0475 sec/batch
Epoch: 1/20... Training Step: 557... Training loss: 2.0363... 0.0512 sec/batch
Epoch: 1/20... Training Step: 558... Training loss: 1.9666... 0.0526 sec/batch
Epoch: 1/20... Training Step: 559... Training loss: 2.1520... 0.0496 sec/batch
Epoch: 1/20... Training Step: 560... Training loss: 2.1197... 0.0416 sec/batch
Epoch: 1/20... Training Step: 561... Training loss: 2.0956... 0.0519 sec/batch
Epoch: 1/20... Training Step: 562... Training loss: 2.1603... 0.0508 sec/batch
Epoch: 1/20... Training Step: 563... Training loss: 1.9994... 0.0517 sec/batch
Epoch: 1/20... Training Step: 564... Training loss: 2.1534... 0.0523 sec/batch
Epoch: 1/20... Training Step: 565... Training loss: 2.1286... 0.0467 sec/batch
Epoch: 1/20... Training Step: 566... Training loss: 2.0175... 0.0442 sec/batch
Epoch: 1/20... Training Step: 567... Training loss: 2.0380... 0.0516 sec/batch
Epoch: 1/20... Training Step: 568... Training loss: 2.1305... 0.0498 sec/batch
Epoch: 1/20... Training Step: 569... Training loss: 2.1178... 0.0500 sec/batch
Epoch: 1/20... Training Step: 570... Training loss: 2.2657... 0.0495 sec/batch
Epoch: 1/20... Training Step: 571... Training loss: 2.1078... 0.0502 sec/batch
Epoch: 1/20... Training Step: 572... Training loss: 2.1776... 0.0443 sec/batch
Epoch: 1/20... Training Step: 573... Training loss: 1.9986... 0.0513 sec/batch
Epoch: 1/20... Training Step: 574... Training loss: 2.0485... 0.0506 sec/batch
Epoch: 1/20... Training Step: 575... Training loss: 1.9899... 0.0471 sec/batch
Epoch: 1/20... Training Step: 576... Training loss: 2.0225... 0.0507 sec/batch
Epoch: 1/20... Training Step: 577... Training loss: 2.1362... 0.0496 sec/batch
Epoch: 1/20... Training Step: 578... Training loss: 2.1342... 0.0406 sec/batch
Epoch: 1/20... Training Step: 579... Training loss: 2.1327... 0.0410 sec/batch
Epoch: 1/20... Training Step: 580... Training loss: 2.2849... 0.0455 sec/batch
Epoch: 1/20... Training Step: 581... Training loss: 2.1709... 0.0510 sec/batch
Epoch: 1/20... Training Step: 582... Training loss: 2.0294... 0.0439 sec/batch
Epoch: 1/20... Training Step: 583... Training loss: 2.1450... 0.0511 sec/batch
Epoch: 1/20... Training Step: 584... Training loss: 2.1431... 0.0554 sec/batch
Epoch: 1/20... Training Step: 585... Training loss: 1.9677... 0.0472 sec/batch
Epoch: 1/20... Training Step: 586... Training loss: 2.1251... 0.0477 sec/batch
Epoch: 1/20... Training Step: 587... Training loss: 1.9326... 0.0450 sec/batch
Epoch: 1/20... Training Step: 588... Training loss: 2.0694... 0.0400 sec/batch
Epoch: 1/20... Training Step: 589... Training loss: 2.3514... 0.0517 sec/batch
Epoch: 1/20... Training Step: 590... Training loss: 2.1964... 0.0507 sec/batch
Epoch: 1/20... Training Step: 591... Training loss: 2.1809... 0.0491 sec/batch
Epoch: 1/20... Training Step: 592... Training loss: 2.0328... 0.0509 sec/batch
Epoch: 1/20... Training Step: 593... Training loss: 2.1438... 0.0521 sec/batch
Epoch: 1/20... Training Step: 594... Training loss: 2.2701... 0.0451 sec/batch
Epoch: 1/20... Training Step: 595... Training loss: 2.3346... 0.0423 sec/batch
Epoch: 1/20... Training Step: 596... Training loss: 2.1661... 0.0455 sec/batch
Epoch: 1/20... Training Step: 597... Training loss: 2.0356... 0.0509 sec/batch
Epoch: 1/20... Training Step: 598... Training loss: 2.1668... 0.0503 sec/batch
Epoch: 1/20... Training Step: 599... Training loss: 2.1341... 0.0528 sec/batch
Epoch: 1/20... Training Step: 600... Training loss: 1.9160... 0.0472 sec/batch
Epoch: 1/20... Training Step: 601... Training loss: 2.0854... 0.0390 sec/batch
Epoch: 1/20... Training Step: 602... Training loss: 2.0589... 0.0420 sec/batch
Epoch: 1/20... Training Step: 603... Training loss: 2.1166... 0.0439 sec/batch
Epoch: 1/20... Training Step: 604... Training loss: 1.9458... 0.0407 sec/batch
Epoch: 1/20... Training Step: 605... Training loss: 2.0053... 0.0402 sec/batch
Epoch: 1/20... Training Step: 606... Training loss: 2.0302... 0.0411 sec/batch
Epoch: 1/20... Training Step: 607... Training loss: 2.1019... 0.0428 sec/batch
Epoch: 1/20... Training Step: 608... Training loss: 2.1146... 0.0425 sec/batch
Epoch: 1/20... Training Step: 609... Training loss: 2.2432... 0.0406 sec/batch
Epoch: 1/20... Training Step: 610... Training loss: 2.0851... 0.0397 sec/batch
Epoch: 1/20... Training Step: 611... Training loss: 2.1746... 0.0433 sec/batch
Epoch: 1/20... Training Step: 612... Training loss: 2.0988... 0.0423 sec/batch
Epoch: 1/20... Training Step: 613... Training loss: 2.0833... 0.0453 sec/batch
Epoch: 1/20... Training Step: 614... Training loss: 2.1642... 0.0393 sec/batch
Epoch: 1/20... Training Step: 615... Training loss: 2.0743... 0.0387 sec/batch
Epoch: 1/20... Training Step: 616... Training loss: 2.0208... 0.0384 sec/batch
Epoch: 1/20... Training Step: 617... Training loss: 1.9904... 0.0401 sec/batch
Epoch: 1/20... Training Step: 618... Training loss: 2.0211... 0.0539 sec/batch
Epoch: 1/20... Training Step: 619... Training loss: 2.0048... 0.0401 sec/batch
Epoch: 1/20... Training Step: 620... Training loss: 1.9215... 0.0394 sec/batch
Epoch: 1/20... Training Step: 621... Training loss: 1.9089... 0.0399 sec/batch
Epoch: 1/20... Training Step: 622... Training loss: 2.0680... 0.0417 sec/batch
Epoch: 1/20... Training Step: 623... Training loss: 2.0676... 0.0448 sec/batch
Epoch: 1/20... Training Step: 624... Training loss: 2.0504... 0.0418 sec/batch
Epoch: 1/20... Training Step: 625... Training loss: 1.9472... 0.0396 sec/batch
Epoch: 1/20... Training Step: 626... Training loss: 2.0465... 0.0397 sec/batch
Epoch: 1/20... Training Step: 627... Training loss: 1.9836... 0.0415 sec/batch
Epoch: 1/20... Training Step: 628... Training loss: 1.9876... 0.0446 sec/batch
Epoch: 1/20... Training Step: 629... Training loss: 2.0572... 0.0421 sec/batch
Epoch: 1/20... Training Step: 630... Training loss: 2.1728... 0.0395 sec/batch
Epoch: 1/20... Training Step: 631... Training loss: 2.0904... 0.0386 sec/batch
Epoch: 1/20... Training Step: 632... Training loss: 1.9556... 0.0411 sec/batch
Epoch: 1/20... Training Step: 633... Training loss: 2.0954... 0.0449 sec/batch
Epoch: 1/20... Training Step: 634... Training loss: 2.0762... 0.0433 sec/batch
Epoch: 1/20... Training Step: 635... Training loss: 2.2621... 0.0399 sec/batch
Epoch: 1/20... Training Step: 636... Training loss: 2.1373... 0.0416 sec/batch
Epoch: 1/20... Training Step: 637... Training loss: 2.0565... 0.0395 sec/batch
Epoch: 1/20... Training Step: 638... Training loss: 2.0485... 0.0450 sec/batch
Epoch: 1/20... Training Step: 639... Training loss: 2.0263... 0.0399 sec/batch
Epoch: 1/20... Training Step: 640... Training loss: 1.9214... 0.0396 sec/batch
Epoch: 1/20... Training Step: 641... Training loss: 2.0618... 0.0394 sec/batch
Epoch: 1/20... Training Step: 642... Training loss: 2.0001... 0.0388 sec/batch
Epoch: 1/20... Training Step: 643... Training loss: 1.9518... 0.0463 sec/batch
Epoch: 1/20... Training Step: 644... Training loss: 1.9400... 0.0462 sec/batch
Epoch: 1/20... Training Step: 645... Training loss: 2.0477... 0.0481 sec/batch
Epoch: 1/20... Training Step: 646... Training loss: 2.2030... 0.0566 sec/batch
Epoch: 1/20... Training Step: 647... Training loss: 2.0052... 0.0609 sec/batch
Epoch: 1/20... Training Step: 648... Training loss: 2.1396... 0.0502 sec/batch
Epoch: 1/20... Training Step: 649... Training loss: 2.0224... 0.0537 sec/batch
Epoch: 1/20... Training Step: 650... Training loss: 1.9944... 0.0402 sec/batch
Epoch: 1/20... Training Step: 651... Training loss: 1.9817... 0.0393 sec/batch
Epoch: 1/20... Training Step: 652... Training loss: 2.1788... 0.0459 sec/batch
Epoch: 1/20... Training Step: 653... Training loss: 2.2073... 0.0590 sec/batch
Epoch: 1/20... Training Step: 654... Training loss: 1.9883... 0.0517 sec/batch
Epoch: 1/20... Training Step: 655... Training loss: 2.0391... 0.0475 sec/batch
Epoch: 1/20... Training Step: 656... Training loss: 2.1008... 0.0453 sec/batch
Epoch: 1/20... Training Step: 657... Training loss: 1.9815... 0.0455 sec/batch
Epoch: 1/20... Training Step: 658... Training loss: 2.1333... 0.0413 sec/batch
Epoch: 1/20... Training Step: 659... Training loss: 2.0831... 0.0421 sec/batch
Epoch: 1/20... Training Step: 660... Training loss: 2.1132... 0.0413 sec/batch
Epoch: 1/20... Training Step: 661... Training loss: 2.0437... 0.0457 sec/batch
Epoch: 1/20... Training Step: 662... Training loss: 2.0985... 0.0579 sec/batch
Epoch: 1/20... Training Step: 663... Training loss: 2.0725... 0.0513 sec/batch
Epoch: 1/20... Training Step: 664... Training loss: 2.2387... 0.0547 sec/batch
Epoch: 1/20... Training Step: 665... Training loss: 2.1500... 0.0513 sec/batch
Epoch: 1/20... Training Step: 666... Training loss: 2.1486... 0.0486 sec/batch
Epoch: 1/20... Training Step: 667... Training loss: 2.1507... 0.0412 sec/batch
Epoch: 1/20... Training Step: 668... Training loss: 1.9191... 0.0404 sec/batch
Epoch: 1/20... Training Step: 669... Training loss: 2.0428... 0.0414 sec/batch
Epoch: 1/20... Training Step: 670... Training loss: 2.0602... 0.0439 sec/batch
Epoch: 1/20... Training Step: 671... Training loss: 2.0854... 0.0503 sec/batch
Epoch: 1/20... Training Step: 672... Training loss: 2.0412... 0.0439 sec/batch
Epoch: 1/20... Training Step: 673... Training loss: 2.1321... 0.0420 sec/batch
Epoch: 1/20... Training Step: 674... Training loss: 2.1735... 0.0430 sec/batch
Epoch: 1/20... Training Step: 675... Training loss: 1.9830... 0.0435 sec/batch
Epoch: 1/20... Training Step: 676... Training loss: 2.0026... 0.0459 sec/batch
Epoch: 1/20... Training Step: 677... Training loss: 2.1350... 0.0414 sec/batch
Epoch: 1/20... Training Step: 678... Training loss: 2.1404... 0.0412 sec/batch
Epoch: 1/20... Training Step: 679... Training loss: 2.0626... 0.0427 sec/batch
Epoch: 1/20... Training Step: 680... Training loss: 2.1824... 0.0418 sec/batch
Epoch: 1/20... Training Step: 681... Training loss: 2.0120... 0.0462 sec/batch
Epoch: 1/20... Training Step: 682... Training loss: 2.0311... 0.0415 sec/batch
Epoch: 1/20... Training Step: 683... Training loss: 2.1067... 0.0423 sec/batch
Epoch: 1/20... Training Step: 684... Training loss: 1.9118... 0.0457 sec/batch
Epoch: 1/20... Training Step: 685... Training loss: 1.9810... 0.0451 sec/batch
Epoch: 1/20... Training Step: 686... Training loss: 1.9963... 0.0498 sec/batch
Epoch: 1/20... Training Step: 687... Training loss: 2.1002... 0.0413 sec/batch
Epoch: 1/20... Training Step: 688... Training loss: 1.9405... 0.0393 sec/batch
Epoch: 1/20... Training Step: 689... Training loss: 1.9829... 0.0386 sec/batch
Epoch: 1/20... Training Step: 690... Training loss: 2.0255... 0.0409 sec/batch
Epoch: 1/20... Training Step: 691... Training loss: 2.1200... 0.0466 sec/batch
Epoch: 1/20... Training Step: 692... Training loss: 2.1032... 0.0406 sec/batch
Epoch: 1/20... Training Step: 693... Training loss: 2.0915... 0.0418 sec/batch
Epoch: 1/20... Training Step: 694... Training loss: 2.1984... 0.0405 sec/batch
Epoch: 1/20... Training Step: 695... Training loss: 1.9501... 0.0414 sec/batch
Epoch: 1/20... Training Step: 696... Training loss: 2.1341... 0.0483 sec/batch
Epoch: 1/20... Training Step: 697... Training loss: 2.1338... 0.0436 sec/batch
Epoch: 1/20... Training Step: 698... Training loss: 2.1270... 0.0428 sec/batch
Epoch: 1/20... Training Step: 699... Training loss: 2.0833... 0.0404 sec/batch
Epoch: 1/20... Training Step: 700... Training loss: 2.0402... 0.0407 sec/batch
Epoch: 1/20... Training Step: 701... Training loss: 2.1140... 0.0507 sec/batch
Epoch: 1/20... Training Step: 702... Training loss: 2.0630... 0.0419 sec/batch
Epoch: 1/20... Training Step: 703... Training loss: 2.1038... 0.0432 sec/batch
Epoch: 1/20... Training Step: 704... Training loss: 2.0197... 0.0410 sec/batch
Epoch: 1/20... Training Step: 705... Training loss: 2.0921... 0.0496 sec/batch
Epoch: 1/20... Training Step: 706... Training loss: 2.1007... 0.0467 sec/batch
Epoch: 1/20... Training Step: 707... Training loss: 2.0523... 0.0434 sec/batch
Epoch: 1/20... Training Step: 708... Training loss: 1.9768... 0.0422 sec/batch
Epoch: 1/20... Training Step: 709... Training loss: 2.1008... 0.0404 sec/batch
Epoch: 1/20... Training Step: 710... Training loss: 2.0708... 0.0402 sec/batch
Epoch: 1/20... Training Step: 711... Training loss: 1.9919... 0.0443 sec/batch
Epoch: 1/20... Training Step: 712... Training loss: 2.0350... 0.0394 sec/batch
Epoch: 1/20... Training Step: 713... Training loss: 2.0588... 0.0394 sec/batch
Epoch: 1/20... Training Step: 714... Training loss: 2.0018... 0.0401 sec/batch
Epoch: 1/20... Training Step: 715... Training loss: 1.9698... 0.0417 sec/batch
Epoch: 1/20... Training Step: 716... Training loss: 1.9646... 0.0536 sec/batch
Epoch: 1/20... Training Step: 717... Training loss: 2.0005... 0.0422 sec/batch
Epoch: 1/20... Training Step: 718... Training loss: 1.8855... 0.0418 sec/batch
Epoch: 1/20... Training Step: 719... Training loss: 1.9702... 0.0404 sec/batch
Epoch: 1/20... Training Step: 720... Training loss: 2.1458... 0.0422 sec/batch
Epoch: 1/20... Training Step: 721... Training loss: 2.0250... 0.0449 sec/batch
Epoch: 1/20... Training Step: 722... Training loss: 1.9191... 0.0438 sec/batch
Epoch: 1/20... Training Step: 723... Training loss: 2.0606... 0.0396 sec/batch
Epoch: 1/20... Training Step: 724... Training loss: 2.0541... 0.0408 sec/batch
Epoch: 1/20... Training Step: 725... Training loss: 1.9416... 0.0395 sec/batch
Epoch: 1/20... Training Step: 726... Training loss: 2.0272... 0.0557 sec/batch
Epoch: 1/20... Training Step: 727... Training loss: 2.0098... 0.0429 sec/batch
Epoch: 1/20... Training Step: 728... Training loss: 2.0276... 0.0406 sec/batch
Epoch: 1/20... Training Step: 729... Training loss: 1.9157... 0.0417 sec/batch
Epoch: 1/20... Training Step: 730... Training loss: 1.9708... 0.0400 sec/batch
Epoch: 1/20... Training Step: 731... Training loss: 2.0600... 0.0451 sec/batch
Epoch: 1/20... Training Step: 732... Training loss: 1.9483... 0.0399 sec/batch
Epoch: 1/20... Training Step: 733... Training loss: 2.0486... 0.0406 sec/batch
Epoch: 1/20... Training Step: 734... Training loss: 1.9404... 0.0391 sec/batch
Epoch: 1/20... Training Step: 735... Training loss: 1.8714... 0.0399 sec/batch
Epoch: 1/20... Training Step: 736... Training loss: 2.1776... 0.0453 sec/batch
Epoch: 1/20... Training Step: 737... Training loss: 2.0903... 0.0388 sec/batch
Epoch: 1/20... Training Step: 738... Training loss: 2.0269... 0.0389 sec/batch
Epoch: 1/20... Training Step: 739... Training loss: 2.1576... 0.0399 sec/batch
Epoch: 1/20... Training Step: 740... Training loss: 1.9759... 0.0421 sec/batch
Epoch: 1/20... Training Step: 741... Training loss: 2.0453... 0.0471 sec/batch
Epoch: 1/20... Training Step: 742... Training loss: 1.8148... 0.0383 sec/batch
Epoch: 1/20... Training Step: 743... Training loss: 1.9231... 0.0415 sec/batch
Epoch: 1/20... Training Step: 744... Training loss: 1.9876... 0.0423 sec/batch
Epoch: 1/20... Training Step: 745... Training loss: 2.1227... 0.0382 sec/batch
Epoch: 1/20... Training Step: 746... Training loss: 2.1351... 0.0504 sec/batch
Epoch: 1/20... Training Step: 747... Training loss: 2.0857... 0.0425 sec/batch
Epoch: 1/20... Training Step: 748... Training loss: 1.9967... 0.0388 sec/batch
Epoch: 1/20... Training Step: 749... Training loss: 1.9545... 0.0428 sec/batch
Epoch: 1/20... Training Step: 750... Training loss: 2.0868... 0.0389 sec/batch
Epoch: 1/20... Training Step: 751... Training loss: 2.0415... 0.0520 sec/batch
Epoch: 1/20... Training Step: 752... Training loss: 2.3169... 0.0406 sec/batch
Epoch: 1/20... Training Step: 753... Training loss: 2.1118... 0.0379 sec/batch
Epoch: 1/20... Training Step: 754... Training loss: 1.9207... 0.0410 sec/batch
Epoch: 1/20... Training Step: 755... Training loss: 1.9795... 0.0401 sec/batch
Epoch: 1/20... Training Step: 756... Training loss: 1.9642... 0.0479 sec/batch
Epoch: 1/20... Training Step: 757... Training loss: 1.9637... 0.0414 sec/batch
Epoch: 1/20... Training Step: 758... Training loss: 1.9523... 0.0408 sec/batch
Epoch: 1/20... Training Step: 759... Training loss: 2.0643... 0.0415 sec/batch
Epoch: 1/20... Training Step: 760... Training loss: 1.9554... 0.0435 sec/batch
Epoch: 1/20... Training Step: 761... Training loss: 1.9611... 0.0595 sec/batch
Epoch: 1/20... Training Step: 762... Training loss: 2.0781... 0.0392 sec/batch
Epoch: 1/20... Training Step: 763... Training loss: 2.0810... 0.0402 sec/batch
Epoch: 1/20... Training Step: 764... Training loss: 2.0124... 0.0396 sec/batch
Epoch: 1/20... Training Step: 765... Training loss: 2.0656... 0.0400 sec/batch
Epoch: 1/20... Training Step: 766... Training loss: 2.1050... 0.0584 sec/batch
Epoch: 1/20... Training Step: 767... Training loss: 2.0010... 0.0382 sec/batch
Epoch: 1/20... Training Step: 768... Training loss: 2.0392... 0.0388 sec/batch
Epoch: 1/20... Training Step: 769... Training loss: 2.1018... 0.0403 sec/batch
Epoch: 1/20... Training Step: 770... Training loss: 2.0436... 0.0420 sec/batch
Epoch: 1/20... Training Step: 771... Training loss: 1.9636... 0.0562 sec/batch
Epoch: 1/20... Training Step: 772... Training loss: 1.9998... 0.0440 sec/batch
Epoch: 1/20... Training Step: 773... Training loss: 2.0459... 0.0421 sec/batch
Epoch: 1/20... Training Step: 774... Training loss: 2.1705... 0.0412 sec/batch
Epoch: 1/20... Training Step: 775... Training loss: 2.0069... 0.0442 sec/batch
Epoch: 1/20... Training Step: 776... Training loss: 2.0894... 0.0442 sec/batch
Epoch: 1/20... Training Step: 777... Training loss: 2.0617... 0.0394 sec/batch
Epoch: 1/20... Training Step: 778... Training loss: 2.1491... 0.0416 sec/batch
Epoch: 1/20... Training Step: 779... Training loss: 2.0873... 0.0431 sec/batch
Epoch: 1/20... Training Step: 780... Training loss: 2.0784... 0.0452 sec/batch
Epoch: 1/20... Training Step: 781... Training loss: 1.9680... 0.0464 sec/batch
Epoch: 1/20... Training Step: 782... Training loss: 2.0825... 0.0398 sec/batch
Epoch: 1/20... Training Step: 783... Training loss: 1.9626... 0.0419 sec/batch
Epoch: 1/20... Training Step: 784... Training loss: 2.0078... 0.0384 sec/batch
Epoch: 1/20... Training Step: 785... Training loss: 2.1136... 0.0396 sec/batch
Epoch: 1/20... Training Step: 786... Training loss: 2.1251... 0.0476 sec/batch
Epoch: 1/20... Training Step: 787... Training loss: 2.0020... 0.0408 sec/batch
Epoch: 1/20... Training Step: 788... Training loss: 1.8420... 0.0400 sec/batch
Epoch: 1/20... Training Step: 789... Training loss: 2.0470... 0.0438 sec/batch
Epoch: 1/20... Training Step: 790... Training loss: 2.0053... 0.0420 sec/batch
Epoch: 1/20... Training Step: 791... Training loss: 2.0201... 0.0458 sec/batch
Epoch: 1/20... Training Step: 792... Training loss: 1.8596... 0.0397 sec/batch
Epoch: 1/20... Training Step: 793... Training loss: 2.0285... 0.0395 sec/batch
Epoch: 1/20... Training Step: 794... Training loss: 2.0397... 0.0382 sec/batch
Epoch: 1/20... Training Step: 795... Training loss: 1.9150... 0.0409 sec/batch
Epoch: 1/20... Training Step: 796... Training loss: 1.9139... 0.0548 sec/batch
Epoch: 1/20... Training Step: 797... Training loss: 2.0040... 0.0408 sec/batch
Epoch: 1/20... Training Step: 798... Training loss: 1.9045... 0.0404 sec/batch
Epoch: 1/20... Training Step: 799... Training loss: 1.9714... 0.0416 sec/batch
Epoch: 1/20... Training Step: 800... Training loss: 1.9946... 0.0405 sec/batch
Epoch: 1/20... Training Step: 801... Training loss: 2.0010... 0.0403 sec/batch
Epoch: 1/20... Training Step: 802... Training loss: 1.8985... 0.0393 sec/batch
Epoch: 1/20... Training Step: 803... Training loss: 2.0302... 0.0411 sec/batch
Epoch: 1/20... Training Step: 804... Training loss: 2.0804... 0.0425 sec/batch
Epoch: 1/20... Training Step: 805... Training loss: 2.1253... 0.0469 sec/batch
Epoch: 1/20... Training Step: 806... Training loss: 2.0871... 0.0413 sec/batch
Epoch: 1/20... Training Step: 807... Training loss: 1.9365... 0.0406 sec/batch
Epoch: 1/20... Training Step: 808... Training loss: 1.9179... 0.0384 sec/batch
Epoch: 1/20... Training Step: 809... Training loss: 1.9898... 0.0406 sec/batch
Epoch: 1/20... Training Step: 810... Training loss: 2.0002... 0.0491 sec/batch
Epoch: 1/20... Training Step: 811... Training loss: 2.0058... 0.0435 sec/batch
Epoch: 1/20... Training Step: 812... Training loss: 2.0085... 0.0395 sec/batch
Epoch: 1/20... Training Step: 813... Training loss: 1.9537... 0.0410 sec/batch
Epoch: 1/20... Training Step: 814... Training loss: 2.0315... 0.0406 sec/batch
Epoch: 1/20... Training Step: 815... Training loss: 2.0062... 0.0504 sec/batch
Epoch: 1/20... Training Step: 816... Training loss: 1.9005... 0.0460 sec/batch
Epoch: 1/20... Training Step: 817... Training loss: 1.9103... 0.0411 sec/batch
Epoch: 1/20... Training Step: 818... Training loss: 1.9103... 0.0407 sec/batch
Epoch: 1/20... Training Step: 819... Training loss: 1.9586... 0.0403 sec/batch
Epoch: 1/20... Training Step: 820... Training loss: 2.1126... 0.0456 sec/batch
Epoch: 1/20... Training Step: 821... Training loss: 1.9602... 0.0420 sec/batch
Epoch: 1/20... Training Step: 822... Training loss: 2.1012... 0.0420 sec/batch
Epoch: 1/20... Training Step: 823... Training loss: 1.9334... 0.0428 sec/batch
Epoch: 1/20... Training Step: 824... Training loss: 1.9355... 0.0432 sec/batch
Epoch: 1/20... Training Step: 825... Training loss: 1.9316... 0.0528 sec/batch
Epoch: 1/20... Training Step: 826... Training loss: 1.9634... 0.0411 sec/batch
Epoch: 1/20... Training Step: 827... Training loss: 1.9829... 0.0399 sec/batch
Epoch: 1/20... Training Step: 828... Training loss: 2.0168... 0.0397 sec/batch
Epoch: 1/20... Training Step: 829... Training loss: 1.9017... 0.0409 sec/batch
Epoch: 1/20... Training Step: 830... Training loss: 1.9633... 0.0434 sec/batch
Epoch: 1/20... Training Step: 831... Training loss: 2.0313... 0.0407 sec/batch
Epoch: 1/20... Training Step: 832... Training loss: 1.9188... 0.0395 sec/batch
Epoch: 1/20... Training Step: 833... Training loss: 1.8985... 0.0409 sec/batch
Epoch: 1/20... Training Step: 834... Training loss: 2.0089... 0.0417 sec/batch
Epoch: 1/20... Training Step: 835... Training loss: 2.0476... 0.0510 sec/batch
Epoch: 1/20... Training Step: 836... Training loss: 1.8722... 0.0442 sec/batch
Epoch: 1/20... Training Step: 837... Training loss: 1.9410... 0.0389 sec/batch
Epoch: 1/20... Training Step: 838... Training loss: 1.8813... 0.0407 sec/batch
Epoch: 1/20... Training Step: 839... Training loss: 2.0475... 0.0408 sec/batch
Epoch: 1/20... Training Step: 840... Training loss: 1.9735... 0.0480 sec/batch
Epoch: 1/20... Training Step: 841... Training loss: 2.0227... 0.0398 sec/batch
Epoch: 1/20... Training Step: 842... Training loss: 1.9957... 0.0408 sec/batch
Epoch: 1/20... Training Step: 843... Training loss: 2.0448... 0.0412 sec/batch
Epoch: 1/20... Training Step: 844... Training loss: 1.9654... 0.0401 sec/batch
Epoch: 1/20... Training Step: 845... Training loss: 1.8861... 0.0455 sec/batch
Epoch: 1/20... Training Step: 846... Training loss: 1.9603... 0.0404 sec/batch
Epoch: 1/20... Training Step: 847... Training loss: 2.0725... 0.0412 sec/batch
Epoch: 1/20... Training Step: 848... Training loss: 1.9867... 0.0399 sec/batch
Epoch: 1/20... Training Step: 849... Training loss: 1.9241... 0.0395 sec/batch
Epoch: 1/20... Training Step: 850... Training loss: 2.1166... 0.0557 sec/batch
Epoch: 1/20... Training Step: 851... Training loss: 2.1097... 0.0402 sec/batch
Epoch: 1/20... Training Step: 852... Training loss: 1.8849... 0.0396 sec/batch
Epoch: 1/20... Training Step: 853... Training loss: 2.0012... 0.0399 sec/batch
Epoch: 1/20... Training Step: 854... Training loss: 1.8939... 0.0422 sec/batch
Epoch: 1/20... Training Step: 855... Training loss: 2.1147... 0.0503 sec/batch
Epoch: 1/20... Training Step: 856... Training loss: 2.0856... 0.0479 sec/batch
Epoch: 1/20... Training Step: 857... Training loss: 2.1462... 0.0511 sec/batch
Epoch: 1/20... Training Step: 858... Training loss: 1.9292... 0.0436 sec/batch
Epoch: 1/20... Training Step: 859... Training loss: 2.0287... 0.0535 sec/batch
Epoch: 1/20... Training Step: 860... Training loss: 2.1103... 0.0452 sec/batch
Epoch: 1/20... Training Step: 861... Training loss: 1.9767... 0.0398 sec/batch
Epoch: 1/20... Training Step: 862... Training loss: 2.0234... 0.0402 sec/batch
Epoch: 1/20... Training Step: 863... Training loss: 2.1295... 0.0401 sec/batch
Epoch: 1/20... Training Step: 864... Training loss: 2.0188... 0.0400 sec/batch
Epoch: 1/20... Training Step: 865... Training loss: 2.0704... 0.0450 sec/batch
Epoch: 1/20... Training Step: 866... Training loss: 2.1528... 0.0408 sec/batch
Epoch: 1/20... Training Step: 867... Training loss: 2.0754... 0.0442 sec/batch
Epoch: 1/20... Training Step: 868... Training loss: 2.0934... 0.0412 sec/batch
Epoch: 1/20... Training Step: 869... Training loss: 2.0376... 0.0428 sec/batch
Epoch: 1/20... Training Step: 870... Training loss: 2.0858... 0.0470 sec/batch
Epoch: 1/20... Training Step: 871... Training loss: 1.9920... 0.0453 sec/batch
Epoch: 1/20... Training Step: 872... Training loss: 1.9865... 0.0420 sec/batch
Epoch: 1/20... Training Step: 873... Training loss: 1.9999... 0.0424 sec/batch
Epoch: 1/20... Training Step: 874... Training loss: 1.9130... 0.0506 sec/batch
Epoch: 1/20... Training Step: 875... Training loss: 2.0204... 0.0536 sec/batch
Epoch: 1/20... Training Step: 876... Training loss: 1.9587... 0.0439 sec/batch
Epoch: 1/20... Training Step: 877... Training loss: 1.9072... 0.0427 sec/batch
Epoch: 1/20... Training Step: 878... Training loss: 2.0882... 0.0401 sec/batch
Epoch: 1/20... Training Step: 879... Training loss: 1.9228... 0.0414 sec/batch
Epoch: 1/20... Training Step: 880... Training loss: 1.9219... 0.0449 sec/batch
Epoch: 1/20... Training Step: 881... Training loss: 2.1909... 0.0423 sec/batch
Epoch: 1/20... Training Step: 882... Training loss: 2.1554... 0.0398 sec/batch
Epoch: 1/20... Training Step: 883... Training loss: 2.0188... 0.0406 sec/batch
Epoch: 1/20... Training Step: 884... Training loss: 2.0199... 0.0421 sec/batch
Epoch: 1/20... Training Step: 885... Training loss: 1.9328... 0.0492 sec/batch
Epoch: 1/20... Training Step: 886... Training loss: 1.9621... 0.0406 sec/batch
Epoch: 1/20... Training Step: 887... Training loss: 2.0400... 0.0416 sec/batch
Epoch: 1/20... Training Step: 888... Training loss: 1.9835... 0.0423 sec/batch
Epoch: 1/20... Training Step: 889... Training loss: 1.9449... 0.0435 sec/batch
Epoch: 1/20... Training Step: 890... Training loss: 1.9960... 0.0454 sec/batch
Epoch: 1/20... Training Step: 891... Training loss: 1.9269... 0.0423 sec/batch
Epoch: 1/20... Training Step: 892... Training loss: 2.0412... 0.0394 sec/batch
Epoch: 1/20... Training Step: 893... Training loss: 2.1302... 0.0394 sec/batch
Epoch: 1/20... Training Step: 894... Training loss: 2.0086... 0.0403 sec/batch
Epoch: 1/20... Training Step: 895... Training loss: 1.9090... 0.0419 sec/batch
Epoch: 1/20... Training Step: 896... Training loss: 1.9495... 0.0411 sec/batch
Epoch: 1/20... Training Step: 897... Training loss: 2.0726... 0.0413 sec/batch
Epoch: 1/20... Training Step: 898... Training loss: 1.9241... 0.0429 sec/batch
Epoch: 1/20... Training Step: 899... Training loss: 2.0842... 0.0405 sec/batch
Epoch: 1/20... Training Step: 900... Training loss: 1.9184... 0.0461 sec/batch
Epoch: 1/20... Training Step: 901... Training loss: 1.8832... 0.0428 sec/batch
Epoch: 1/20... Training Step: 902... Training loss: 1.8981... 0.0432 sec/batch
Epoch: 1/20... Training Step: 903... Training loss: 1.9609... 0.0396 sec/batch
Epoch: 1/20... Training Step: 904... Training loss: 1.9345... 0.0420 sec/batch
Epoch: 1/20... Training Step: 905... Training loss: 1.9508... 0.0450 sec/batch
Epoch: 1/20... Training Step: 906... Training loss: 2.0723... 0.0398 sec/batch
Epoch: 1/20... Training Step: 907... Training loss: 2.1266... 0.0415 sec/batch
Epoch: 1/20... Training Step: 908... Training loss: 2.1244... 0.0384 sec/batch
Epoch: 1/20... Training Step: 909... Training loss: 2.0918... 0.0413 sec/batch
Epoch: 1/20... Training Step: 910... Training loss: 2.0378... 0.0475 sec/batch
Epoch: 1/20... Training Step: 911... Training loss: 1.9615... 0.0420 sec/batch
Epoch: 1/20... Training Step: 912... Training loss: 1.9304... 0.0430 sec/batch
Epoch: 1/20... Training Step: 913... Training loss: 2.0712... 0.0414 sec/batch
Epoch: 1/20... Training Step: 914... Training loss: 2.0079... 0.0428 sec/batch
Epoch: 1/20... Training Step: 915... Training loss: 1.9557... 0.0477 sec/batch
Epoch: 1/20... Training Step: 916... Training loss: 1.9871... 0.0419 sec/batch
Epoch: 1/20... Training Step: 917... Training loss: 2.0664... 0.0517 sec/batch
Epoch: 1/20... Training Step: 918... Training loss: 1.8373... 0.0524 sec/batch
Epoch: 1/20... Training Step: 919... Training loss: 1.8766... 0.0472 sec/batch
Epoch: 1/20... Training Step: 920... Training loss: 1.9491... 0.0580 sec/batch
Epoch: 1/20... Training Step: 921... Training loss: 1.8895... 0.0473 sec/batch
Epoch: 1/20... Training Step: 922... Training loss: 2.0022... 0.0433 sec/batch
Epoch: 1/20... Training Step: 923... Training loss: 2.0223... 0.0429 sec/batch
Epoch: 1/20... Training Step: 924... Training loss: 2.0071... 0.0449 sec/batch
Epoch: 1/20... Training Step: 925... Training loss: 1.9783... 0.0492 sec/batch
Epoch: 1/20... Training Step: 926... Training loss: 1.9992... 0.0493 sec/batch
Epoch: 1/20... Training Step: 927... Training loss: 2.0023... 0.0512 sec/batch
Epoch: 1/20... Training Step: 928... Training loss: 1.9062... 0.0500 sec/batch
Epoch: 1/20... Training Step: 929... Training loss: 1.9306... 0.0478 sec/batch
Epoch: 1/20... Training Step: 930... Training loss: 2.0394... 0.0406 sec/batch
Epoch: 1/20... Training Step: 931... Training loss: 2.0366... 0.0406 sec/batch
Epoch: 1/20... Training Step: 932... Training loss: 1.8858... 0.0389 sec/batch
Epoch: 1/20... Training Step: 933... Training loss: 1.9458... 0.0407 sec/batch
Epoch: 1/20... Training Step: 934... Training loss: 1.9269... 0.0384 sec/batch
Epoch: 1/20... Training Step: 935... Training loss: 1.9714... 0.0489 sec/batch
Epoch: 1/20... Training Step: 936... Training loss: 2.0330... 0.0396 sec/batch
Epoch: 1/20... Training Step: 937... Training loss: 1.8323... 0.0391 sec/batch
Epoch: 1/20... Training Step: 938... Training loss: 1.9439... 0.0418 sec/batch
Epoch: 1/20... Training Step: 939... Training loss: 1.8956... 0.0402 sec/batch
Epoch: 1/20... Training Step: 940... Training loss: 1.9844... 0.0546 sec/batch
Epoch: 1/20... Training Step: 941... Training loss: 1.8885... 0.0399 sec/batch
Epoch: 1/20... Training Step: 942... Training loss: 1.8733... 0.0409 sec/batch
Epoch: 1/20... Training Step: 943... Training loss: 1.8696... 0.0388 sec/batch
Epoch: 1/20... Training Step: 944... Training loss: 2.0192... 0.0391 sec/batch
Epoch: 1/20... Training Step: 945... Training loss: 1.9258... 0.0559 sec/batch
Epoch: 1/20... Training Step: 946... Training loss: 1.9405... 0.0411 sec/batch
Epoch: 1/20... Training Step: 947... Training loss: 1.8180... 0.0417 sec/batch
Epoch: 1/20... Training Step: 948... Training loss: 1.9496... 0.0395 sec/batch
Epoch: 1/20... Training Step: 949... Training loss: 1.9915... 0.0412 sec/batch
Epoch: 1/20... Training Step: 950... Training loss: 1.9801... 0.0640 sec/batch
Epoch: 1/20... Training Step: 951... Training loss: 1.9684... 0.0426 sec/batch
Epoch: 1/20... Training Step: 952... Training loss: 2.0420... 0.0392 sec/batch
Epoch: 1/20... Training Step: 953... Training loss: 1.9656... 0.0398 sec/batch
Epoch: 1/20... Training Step: 954... Training loss: 1.9584... 0.0441 sec/batch
Epoch: 1/20... Training Step: 955... Training loss: 2.0189... 0.0442 sec/batch
Epoch: 1/20... Training Step: 956... Training loss: 1.8598... 0.0417 sec/batch
Epoch: 1/20... Training Step: 957... Training loss: 2.1042... 0.0414 sec/batch
Epoch: 1/20... Training Step: 958... Training loss: 2.0455... 0.0406 sec/batch
Epoch: 1/20... Training Step: 959... Training loss: 1.9173... 0.0411 sec/batch
Epoch: 1/20... Training Step: 960... Training loss: 2.0495... 0.0461 sec/batch
Epoch: 1/20... Training Step: 961... Training loss: 1.8798... 0.0430 sec/batch
Epoch: 1/20... Training Step: 962... Training loss: 1.8529... 0.0430 sec/batch
Epoch: 1/20... Training Step: 963... Training loss: 1.8066... 0.0433 sec/batch
Epoch: 1/20... Training Step: 964... Training loss: 1.9318... 0.0430 sec/batch
Epoch: 1/20... Training Step: 965... Training loss: 1.9204... 0.0537 sec/batch
Epoch: 1/20... Training Step: 966... Training loss: 2.0883... 0.0410 sec/batch
Epoch: 1/20... Training Step: 967... Training loss: 2.0128... 0.0413 sec/batch
Epoch: 1/20... Training Step: 968... Training loss: 1.9032... 0.0408 sec/batch
Epoch: 1/20... Training Step: 969... Training loss: 1.9054... 0.0446 sec/batch
Epoch: 1/20... Training Step: 970... Training loss: 1.8245... 0.0479 sec/batch
Epoch: 1/20... Training Step: 971... Training loss: 1.9525... 0.0396 sec/batch
Epoch: 1/20... Training Step: 972... Training loss: 2.0588... 0.0393 sec/batch
Epoch: 1/20... Training Step: 973... Training loss: 1.8971... 0.0409 sec/batch
Epoch: 1/20... Training Step: 974... Training loss: 1.9189... 0.0389 sec/batch
Epoch: 1/20... Training Step: 975... Training loss: 1.9806... 0.0456 sec/batch
Epoch: 1/20... Training Step: 976... Training loss: 2.0521... 0.0404 sec/batch
Epoch: 1/20... Training Step: 977... Training loss: 2.2147... 0.0399 sec/batch
Epoch: 1/20... Training Step: 978... Training loss: 1.9502... 0.0388 sec/batch
Epoch: 1/20... Training Step: 979... Training loss: 1.9848... 0.0395 sec/batch
Epoch: 1/20... Training Step: 980... Training loss: 1.8617... 0.0483 sec/batch
Epoch: 1/20... Training Step: 981... Training loss: 1.9785... 0.0424 sec/batch
Epoch: 1/20... Training Step: 982... Training loss: 1.8584... 0.0448 sec/batch
Epoch: 1/20... Training Step: 983... Training loss: 1.9979... 0.0399 sec/batch
Epoch: 1/20... Training Step: 984... Training loss: 1.9058... 0.0397 sec/batch
Epoch: 1/20... Training Step: 985... Training loss: 1.9219... 0.0542 sec/batch
Epoch: 1/20... Training Step: 986... Training loss: 1.8616... 0.0409 sec/batch
Epoch: 1/20... Training Step: 987... Training loss: 1.8971... 0.0421 sec/batch
Epoch: 1/20... Training Step: 988... Training loss: 1.8733... 0.0407 sec/batch
Epoch: 1/20... Training Step: 989... Training loss: 1.9502... 0.0391 sec/batch
Epoch: 1/20... Training Step: 990... Training loss: 1.9020... 0.0504 sec/batch
Epoch: 1/20... Training Step: 991... Training loss: 1.8262... 0.0398 sec/batch
Epoch: 1/20... Training Step: 992... Training loss: 2.0025... 0.0492 sec/batch
Epoch: 1/20... Training Step: 993... Training loss: 2.0687... 0.0391 sec/batch
Epoch: 1/20... Training Step: 994... Training loss: 2.0013... 0.0418 sec/batch
Epoch: 1/20... Training Step: 995... Training loss: 1.9905... 0.0464 sec/batch
Epoch: 1/20... Training Step: 996... Training loss: 2.0160... 0.0431 sec/batch
Epoch: 1/20... Training Step: 997... Training loss: 1.8590... 0.0393 sec/batch
Epoch: 1/20... Training Step: 998... Training loss: 2.0421... 0.0418 sec/batch
Epoch: 1/20... Training Step: 999... Training loss: 1.9054... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1000... Training loss: 1.9161... 0.0576 sec/batch
Epoch: 1/20... Training Step: 1001... Training loss: 1.9091... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1002... Training loss: 1.8201... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1003... Training loss: 1.9486... 0.0531 sec/batch
Epoch: 1/20... Training Step: 1004... Training loss: 1.8281... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1005... Training loss: 1.7991... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1006... Training loss: 1.9020... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1007... Training loss: 1.9458... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1008... Training loss: 2.0325... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1009... Training loss: 1.9866... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1010... Training loss: 2.1024... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1011... Training loss: 1.9401... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1012... Training loss: 2.0293... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1013... Training loss: 2.0045... 0.0557 sec/batch
Epoch: 1/20... Training Step: 1014... Training loss: 2.0853... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1015... Training loss: 1.8219... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1016... Training loss: 2.0191... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1017... Training loss: 2.0582... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1018... Training loss: 1.9863... 0.0446 sec/batch
Epoch: 1/20... Training Step: 1019... Training loss: 2.0487... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1020... Training loss: 1.7901... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1021... Training loss: 1.8883... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1022... Training loss: 1.9357... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1023... Training loss: 1.8595... 0.0474 sec/batch
Epoch: 1/20... Training Step: 1024... Training loss: 1.9469... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1025... Training loss: 2.0284... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1026... Training loss: 2.0277... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1027... Training loss: 1.9829... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1028... Training loss: 1.8548... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1029... Training loss: 1.9768... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1030... Training loss: 1.9469... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1031... Training loss: 1.8981... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1032... Training loss: 1.8760... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1033... Training loss: 1.8671... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1034... Training loss: 1.9419... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1035... Training loss: 2.0101... 0.0381 sec/batch
Epoch: 1/20... Training Step: 1036... Training loss: 2.0430... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1037... Training loss: 1.8079... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1038... Training loss: 1.8966... 0.0545 sec/batch
Epoch: 1/20... Training Step: 1039... Training loss: 2.0059... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1040... Training loss: 1.9194... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1041... Training loss: 1.9322... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1042... Training loss: 1.8750... 0.0476 sec/batch
Epoch: 1/20... Training Step: 1043... Training loss: 2.0038... 0.0512 sec/batch
Epoch: 1/20... Training Step: 1044... Training loss: 2.0157... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1045... Training loss: 2.0452... 0.0384 sec/batch
Epoch: 1/20... Training Step: 1046... Training loss: 1.9091... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1047... Training loss: 1.8972... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1048... Training loss: 1.8603... 0.0498 sec/batch
Epoch: 1/20... Training Step: 1049... Training loss: 1.8172... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1050... Training loss: 2.0379... 0.0442 sec/batch
Epoch: 1/20... Training Step: 1051... Training loss: 1.9718... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1052... Training loss: 1.9597... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1053... Training loss: 1.8446... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1054... Training loss: 1.8889... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1055... Training loss: 1.9074... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1056... Training loss: 1.8771... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1057... Training loss: 1.7856... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1058... Training loss: 1.9632... 0.0526 sec/batch
Epoch: 1/20... Training Step: 1059... Training loss: 2.0898... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1060... Training loss: 1.9276... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1061... Training loss: 1.9661... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1062... Training loss: 1.9251... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1063... Training loss: 1.8955... 0.0603 sec/batch
Epoch: 1/20... Training Step: 1064... Training loss: 1.9954... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1065... Training loss: 1.8985... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1066... Training loss: 2.0657... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1067... Training loss: 1.8908... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1068... Training loss: 2.0641... 0.0444 sec/batch
Epoch: 1/20... Training Step: 1069... Training loss: 1.8861... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1070... Training loss: 1.9116... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1071... Training loss: 1.9282... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1072... Training loss: 1.9156... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1073... Training loss: 1.7980... 0.0481 sec/batch
Epoch: 1/20... Training Step: 1074... Training loss: 1.8650... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1075... Training loss: 1.8732... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1076... Training loss: 1.8747... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1077... Training loss: 1.9095... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1078... Training loss: 1.9962... 0.0483 sec/batch
Epoch: 1/20... Training Step: 1079... Training loss: 1.8813... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1080... Training loss: 1.7988... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1081... Training loss: 1.8815... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1082... Training loss: 1.8326... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1083... Training loss: 2.0105... 0.0504 sec/batch
Epoch: 1/20... Training Step: 1084... Training loss: 1.7019... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1085... Training loss: 1.8187... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1086... Training loss: 1.9066... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1087... Training loss: 1.8672... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1088... Training loss: 1.7318... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1089... Training loss: 1.9317... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1090... Training loss: 2.0976... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1091... Training loss: 1.9891... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1092... Training loss: 2.1088... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1093... Training loss: 2.0094... 0.0510 sec/batch
Epoch: 1/20... Training Step: 1094... Training loss: 1.9406... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1095... Training loss: 1.8897... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1096... Training loss: 1.9656... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1097... Training loss: 1.8584... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1098... Training loss: 1.8327... 0.0522 sec/batch
Epoch: 1/20... Training Step: 1099... Training loss: 2.0180... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1100... Training loss: 1.9478... 0.0387 sec/batch
Epoch: 1/20... Training Step: 1101... Training loss: 1.9012... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1102... Training loss: 1.8654... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1103... Training loss: 1.7439... 0.0545 sec/batch
Epoch: 1/20... Training Step: 1104... Training loss: 1.9980... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1105... Training loss: 1.8980... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1106... Training loss: 1.9127... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1107... Training loss: 1.9281... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1108... Training loss: 1.9811... 0.0484 sec/batch
Epoch: 1/20... Training Step: 1109... Training loss: 1.8339... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1110... Training loss: 1.8931... 0.0587 sec/batch
Epoch: 1/20... Training Step: 1111... Training loss: 1.9947... 0.0573 sec/batch
Epoch: 1/20... Training Step: 1112... Training loss: 1.9736... 0.0635 sec/batch
Epoch: 1/20... Training Step: 1113... Training loss: 1.8508... 0.0477 sec/batch
Epoch: 1/20... Training Step: 1114... Training loss: 1.9941... 0.0604 sec/batch
Epoch: 1/20... Training Step: 1115... Training loss: 1.9036... 0.0702 sec/batch
Epoch: 1/20... Training Step: 1116... Training loss: 2.0877... 0.0551 sec/batch
Epoch: 1/20... Training Step: 1117... Training loss: 1.9410... 0.0643 sec/batch
Epoch: 1/20... Training Step: 1118... Training loss: 2.0685... 0.0539 sec/batch
Epoch: 1/20... Training Step: 1119... Training loss: 1.8951... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1120... Training loss: 1.7747... 0.0548 sec/batch
Epoch: 1/20... Training Step: 1121... Training loss: 1.9543... 0.0604 sec/batch
Epoch: 1/20... Training Step: 1122... Training loss: 1.8820... 0.0559 sec/batch
Epoch: 1/20... Training Step: 1123... Training loss: 1.8764... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1124... Training loss: 1.9082... 0.0524 sec/batch
Epoch: 1/20... Training Step: 1125... Training loss: 1.9979... 0.0840 sec/batch
Epoch: 1/20... Training Step: 1126... Training loss: 1.8866... 0.0545 sec/batch
Epoch: 1/20... Training Step: 1127... Training loss: 1.9587... 0.0630 sec/batch
Epoch: 1/20... Training Step: 1128... Training loss: 1.8228... 0.0654 sec/batch
Epoch: 1/20... Training Step: 1129... Training loss: 1.9167... 0.0469 sec/batch
Epoch: 1/20... Training Step: 1130... Training loss: 1.8494... 0.0530 sec/batch
Epoch: 1/20... Training Step: 1131... Training loss: 1.8965... 0.0635 sec/batch
Epoch: 1/20... Training Step: 1132... Training loss: 1.7033... 0.0661 sec/batch
Epoch: 1/20... Training Step: 1133... Training loss: 1.8071... 0.0684 sec/batch
Epoch: 1/20... Training Step: 1134... Training loss: 2.0058... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1135... Training loss: 1.8294... 0.0677 sec/batch
Epoch: 1/20... Training Step: 1136... Training loss: 2.0558... 0.0610 sec/batch
Epoch: 1/20... Training Step: 1137... Training loss: 1.9543... 0.0747 sec/batch
Epoch: 1/20... Training Step: 1138... Training loss: 2.0190... 0.0767 sec/batch
Epoch: 1/20... Training Step: 1139... Training loss: 2.0393... 0.0513 sec/batch
Epoch: 1/20... Training Step: 1140... Training loss: 2.0896... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1141... Training loss: 1.8485... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1142... Training loss: 1.8370... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1143... Training loss: 1.8195... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1144... Training loss: 2.1518... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1145... Training loss: 2.0631... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1146... Training loss: 1.9320... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1147... Training loss: 2.0024... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1148... Training loss: 1.9980... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1149... Training loss: 2.0934... 0.0494 sec/batch
Epoch: 1/20... Training Step: 1150... Training loss: 1.9397... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1151... Training loss: 1.9289... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1152... Training loss: 1.9526... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1153... Training loss: 2.0247... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1154... Training loss: 1.9415... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1155... Training loss: 1.9368... 0.0384 sec/batch
Epoch: 1/20... Training Step: 1156... Training loss: 1.8502... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1157... Training loss: 1.9267... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1158... Training loss: 1.8390... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1159... Training loss: 1.9067... 0.0442 sec/batch
Epoch: 1/20... Training Step: 1160... Training loss: 1.7245... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1161... Training loss: 1.9581... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1162... Training loss: 1.9990... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1163... Training loss: 1.9902... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1164... Training loss: 2.0191... 0.0478 sec/batch
Epoch: 1/20... Training Step: 1165... Training loss: 1.9224... 0.0533 sec/batch
Epoch: 1/20... Training Step: 1166... Training loss: 1.9710... 0.0540 sec/batch
Epoch: 1/20... Training Step: 1167... Training loss: 1.9240... 0.0555 sec/batch
Epoch: 1/20... Training Step: 1168... Training loss: 1.9900... 0.0592 sec/batch
Epoch: 1/20... Training Step: 1169... Training loss: 2.0161... 0.0741 sec/batch
Epoch: 1/20... Training Step: 1170... Training loss: 1.9544... 0.0513 sec/batch
Epoch: 1/20... Training Step: 1171... Training loss: 1.7711... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1172... Training loss: 1.9545... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1173... Training loss: 1.8770... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1174... Training loss: 1.8705... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1175... Training loss: 1.9972... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1176... Training loss: 1.8850... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1177... Training loss: 1.9085... 0.0592 sec/batch
Epoch: 1/20... Training Step: 1178... Training loss: 1.7606... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1179... Training loss: 1.7514... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1180... Training loss: 1.9859... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1181... Training loss: 1.9162... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1182... Training loss: 1.9708... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1183... Training loss: 1.9105... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1184... Training loss: 2.0185... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1185... Training loss: 1.9745... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1186... Training loss: 1.8346... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1187... Training loss: 1.9932... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1188... Training loss: 1.9697... 0.0384 sec/batch
Epoch: 1/20... Training Step: 1189... Training loss: 1.8737... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1190... Training loss: 1.9207... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1191... Training loss: 1.8701... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1192... Training loss: 1.9227... 0.0452 sec/batch
Epoch: 1/20... Training Step: 1193... Training loss: 1.9396... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1194... Training loss: 1.9873... 0.0387 sec/batch
Epoch: 1/20... Training Step: 1195... Training loss: 2.1155... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1196... Training loss: 1.8940... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1197... Training loss: 1.8357... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1198... Training loss: 1.8982... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1199... Training loss: 1.9141... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1200... Training loss: 1.8794... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1201... Training loss: 1.8785... 0.0562 sec/batch
Epoch: 1/20... Training Step: 1202... Training loss: 1.9563... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1203... Training loss: 1.9125... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1204... Training loss: 1.8558... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1205... Training loss: 1.9525... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1206... Training loss: 1.8001... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1207... Training loss: 1.8165... 0.0439 sec/batch
Epoch: 1/20... Training Step: 1208... Training loss: 1.9860... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1209... Training loss: 2.0489... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1210... Training loss: 2.0065... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1211... Training loss: 1.9337... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1212... Training loss: 1.9667... 0.0526 sec/batch
Epoch: 1/20... Training Step: 1213... Training loss: 2.1027... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1214... Training loss: 1.8427... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1215... Training loss: 1.8323... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1216... Training loss: 2.0372... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1217... Training loss: 2.0322... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1218... Training loss: 2.0443... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1219... Training loss: 1.8912... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1220... Training loss: 1.9077... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1221... Training loss: 2.0029... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1222... Training loss: 1.9720... 0.0551 sec/batch
Epoch: 1/20... Training Step: 1223... Training loss: 1.8189... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1224... Training loss: 1.8218... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1225... Training loss: 2.0584... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1226... Training loss: 1.9220... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1227... Training loss: 2.0554... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1228... Training loss: 1.9578... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1229... Training loss: 1.8577... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1230... Training loss: 2.0265... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1231... Training loss: 1.7750... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1232... Training loss: 2.0390... 0.0475 sec/batch
Epoch: 1/20... Training Step: 1233... Training loss: 1.8931... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1234... Training loss: 2.0280... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1235... Training loss: 1.8300... 0.0687 sec/batch
Epoch: 1/20... Training Step: 1236... Training loss: 1.8730... 0.0503 sec/batch
Epoch: 1/20... Training Step: 1237... Training loss: 1.9229... 0.0594 sec/batch
Epoch: 1/20... Training Step: 1238... Training loss: 1.6854... 0.0637 sec/batch
Epoch: 1/20... Training Step: 1239... Training loss: 1.7871... 0.0675 sec/batch
Epoch: 1/20... Training Step: 1240... Training loss: 1.8009... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1241... Training loss: 1.8090... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1242... Training loss: 1.8930... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1243... Training loss: 1.7689... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1244... Training loss: 1.9586... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1245... Training loss: 1.8414... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1246... Training loss: 1.9452... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1247... Training loss: 1.9907... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1248... Training loss: 1.9165... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1249... Training loss: 1.9347... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1250... Training loss: 1.7763... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1251... Training loss: 1.9557... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1252... Training loss: 1.7691... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1253... Training loss: 1.7511... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1254... Training loss: 1.9221... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1255... Training loss: 1.8538... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1256... Training loss: 1.9762... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1257... Training loss: 1.9538... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1258... Training loss: 1.8653... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1259... Training loss: 1.9128... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1260... Training loss: 1.8792... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1261... Training loss: 1.9787... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1262... Training loss: 1.9210... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1263... Training loss: 1.8741... 0.0661 sec/batch
Epoch: 1/20... Training Step: 1264... Training loss: 1.7204... 0.0585 sec/batch
Epoch: 1/20... Training Step: 1265... Training loss: 1.8346... 0.0647 sec/batch
Epoch: 1/20... Training Step: 1266... Training loss: 1.7993... 0.0683 sec/batch
Epoch: 1/20... Training Step: 1267... Training loss: 2.0495... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1268... Training loss: 1.8142... 0.0459 sec/batch
Epoch: 1/20... Training Step: 1269... Training loss: 1.8750... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1270... Training loss: 2.0261... 0.0642 sec/batch
Epoch: 1/20... Training Step: 1271... Training loss: 1.8551... 0.0619 sec/batch
Epoch: 1/20... Training Step: 1272... Training loss: 1.9303... 0.0747 sec/batch
Epoch: 1/20... Training Step: 1273... Training loss: 1.8898... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1274... Training loss: 1.7872... 0.0586 sec/batch
Epoch: 1/20... Training Step: 1275... Training loss: 1.9120... 0.1022 sec/batch
Epoch: 1/20... Training Step: 1276... Training loss: 1.9445... 0.0626 sec/batch
Epoch: 1/20... Training Step: 1277... Training loss: 1.9596... 0.0506 sec/batch
Epoch: 1/20... Training Step: 1278... Training loss: 2.0168... 0.0458 sec/batch
Epoch: 1/20... Training Step: 1279... Training loss: 2.0114... 0.0583 sec/batch
Epoch: 1/20... Training Step: 1280... Training loss: 1.8990... 0.0534 sec/batch
Epoch: 1/20... Training Step: 1281... Training loss: 1.8448... 0.0662 sec/batch
Epoch: 1/20... Training Step: 1282... Training loss: 2.0132... 0.0746 sec/batch
Epoch: 1/20... Training Step: 1283... Training loss: 1.8503... 0.0530 sec/batch
Epoch: 1/20... Training Step: 1284... Training loss: 2.0011... 0.0502 sec/batch
Epoch: 1/20... Training Step: 1285... Training loss: 2.0663... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1286... Training loss: 1.9117... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1287... Training loss: 2.0224... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1288... Training loss: 1.9026... 0.0469 sec/batch
Epoch: 1/20... Training Step: 1289... Training loss: 1.8937... 0.0496 sec/batch
Epoch: 1/20... Training Step: 1290... Training loss: 2.0680... 0.0526 sec/batch
Epoch: 1/20... Training Step: 1291... Training loss: 2.0252... 0.0545 sec/batch
Epoch: 1/20... Training Step: 1292... Training loss: 2.0294... 0.0519 sec/batch
Epoch: 1/20... Training Step: 1293... Training loss: 1.9253... 0.0476 sec/batch
Epoch: 1/20... Training Step: 1294... Training loss: 1.9100... 0.0460 sec/batch
Epoch: 1/20... Training Step: 1295... Training loss: 1.9961... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1296... Training loss: 2.0027... 0.0439 sec/batch
Epoch: 1/20... Training Step: 1297... Training loss: 1.8719... 0.0457 sec/batch
Epoch: 1/20... Training Step: 1298... Training loss: 1.8221... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1299... Training loss: 1.9106... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1300... Training loss: 1.8505... 0.0470 sec/batch
Epoch: 1/20... Training Step: 1301... Training loss: 1.9123... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1302... Training loss: 1.9762... 0.0543 sec/batch
Epoch: 1/20... Training Step: 1303... Training loss: 1.8812... 0.0495 sec/batch
Epoch: 1/20... Training Step: 1304... Training loss: 2.0275... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1305... Training loss: 1.9832... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1306... Training loss: 2.0738... 0.0450 sec/batch
Epoch: 1/20... Training Step: 1307... Training loss: 2.1744... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1308... Training loss: 1.9973... 0.0502 sec/batch
Epoch: 1/20... Training Step: 1309... Training loss: 2.0170... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1310... Training loss: 1.9827... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1311... Training loss: 1.8249... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1312... Training loss: 1.9164... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1313... Training loss: 1.8405... 0.0489 sec/batch
Epoch: 1/20... Training Step: 1314... Training loss: 2.0051... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1315... Training loss: 1.9932... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1316... Training loss: 1.9294... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1317... Training loss: 1.9139... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1318... Training loss: 1.8729... 0.0482 sec/batch
Epoch: 1/20... Training Step: 1319... Training loss: 1.8571... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1320... Training loss: 1.9731... 0.0508 sec/batch
Epoch: 1/20... Training Step: 1321... Training loss: 1.9637... 0.0563 sec/batch
Epoch: 1/20... Training Step: 1322... Training loss: 1.9805... 0.0506 sec/batch
Epoch: 1/20... Training Step: 1323... Training loss: 1.9723... 0.0553 sec/batch
Epoch: 1/20... Training Step: 1324... Training loss: 1.9256... 0.0583 sec/batch
Epoch: 1/20... Training Step: 1325... Training loss: 1.9714... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1326... Training loss: 1.9133... 0.0477 sec/batch
Epoch: 1/20... Training Step: 1327... Training loss: 1.9903... 0.0476 sec/batch
Epoch: 1/20... Training Step: 1328... Training loss: 1.8796... 0.0464 sec/batch
Epoch: 1/20... Training Step: 1329... Training loss: 1.9660... 0.0460 sec/batch
Epoch: 1/20... Training Step: 1330... Training loss: 1.9701... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1331... Training loss: 1.9764... 0.0450 sec/batch
Epoch: 1/20... Training Step: 1332... Training loss: 1.9101... 0.0479 sec/batch
Epoch: 1/20... Training Step: 1333... Training loss: 1.8698... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1334... Training loss: 1.9320... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1335... Training loss: 1.8271... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1336... Training loss: 1.9580... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1337... Training loss: 1.8189... 0.0502 sec/batch
Epoch: 1/20... Training Step: 1338... Training loss: 1.9160... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1339... Training loss: 1.9429... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1340... Training loss: 1.9807... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1341... Training loss: 1.7769... 0.0565 sec/batch
Epoch: 1/20... Training Step: 1342... Training loss: 1.8475... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1343... Training loss: 1.9856... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1344... Training loss: 1.9857... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1345... Training loss: 1.9956... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1346... Training loss: 1.9957... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1347... Training loss: 1.9121... 0.0538 sec/batch
Epoch: 1/20... Training Step: 1348... Training loss: 1.9358... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1349... Training loss: 1.9443... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1350... Training loss: 1.7949... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1351... Training loss: 1.7363... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1352... Training loss: 1.7849... 0.0548 sec/batch
Epoch: 1/20... Training Step: 1353... Training loss: 1.8788... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1354... Training loss: 1.9270... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1355... Training loss: 1.8158... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1356... Training loss: 1.8666... 0.0457 sec/batch
Epoch: 1/20... Training Step: 1357... Training loss: 1.9218... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1358... Training loss: 1.7563... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1359... Training loss: 2.0677... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1360... Training loss: 1.8280... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1361... Training loss: 1.8831... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1362... Training loss: 1.7938... 0.0549 sec/batch
Epoch: 1/20... Training Step: 1363... Training loss: 1.7062... 0.0501 sec/batch
Epoch: 1/20... Training Step: 1364... Training loss: 1.8009... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1365... Training loss: 1.8256... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1366... Training loss: 1.8941... 0.0520 sec/batch
Epoch: 1/20... Training Step: 1367... Training loss: 1.8391... 0.0501 sec/batch
Epoch: 1/20... Training Step: 1368... Training loss: 1.9079... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1369... Training loss: 2.0257... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1370... Training loss: 1.9143... 0.0439 sec/batch
Epoch: 1/20... Training Step: 1371... Training loss: 1.9548... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1372... Training loss: 1.7929... 0.0517 sec/batch
Epoch: 1/20... Training Step: 1373... Training loss: 1.8273... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1374... Training loss: 1.9699... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1375... Training loss: 1.8332... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1376... Training loss: 1.8829... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1377... Training loss: 1.9267... 0.0498 sec/batch
Epoch: 1/20... Training Step: 1378... Training loss: 1.9329... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1379... Training loss: 1.8437... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1380... Training loss: 1.8939... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1381... Training loss: 2.0153... 0.0387 sec/batch
Epoch: 1/20... Training Step: 1382... Training loss: 1.9737... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1383... Training loss: 2.0407... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1384... Training loss: 1.8654... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1385... Training loss: 1.9188... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1386... Training loss: 1.9995... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1387... Training loss: 1.9188... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1388... Training loss: 1.8642... 0.0474 sec/batch
Epoch: 1/20... Training Step: 1389... Training loss: 1.9358... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1390... Training loss: 1.9442... 0.0484 sec/batch
Epoch: 1/20... Training Step: 1391... Training loss: 1.9710... 0.0541 sec/batch
Epoch: 1/20... Training Step: 1392... Training loss: 1.9463... 0.0479 sec/batch
Epoch: 1/20... Training Step: 1393... Training loss: 1.9625... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1394... Training loss: 1.8495... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1395... Training loss: 1.8875... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1396... Training loss: 1.8383... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1397... Training loss: 1.9138... 0.0537 sec/batch
Epoch: 1/20... Training Step: 1398... Training loss: 1.8552... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1399... Training loss: 2.0084... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1400... Training loss: 2.0341... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1401... Training loss: 1.7580... 0.0588 sec/batch
Epoch: 1/20... Training Step: 1402... Training loss: 2.0016... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1403... Training loss: 1.9597... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1404... Training loss: 2.0124... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1405... Training loss: 1.9802... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1406... Training loss: 1.8251... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1407... Training loss: 1.9496... 0.0486 sec/batch
Epoch: 1/20... Training Step: 1408... Training loss: 1.9815... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1409... Training loss: 1.9018... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1410... Training loss: 1.9623... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1411... Training loss: 1.8626... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1412... Training loss: 1.9478... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1413... Training loss: 1.8130... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1414... Training loss: 2.2621... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1415... Training loss: 1.9360... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1416... Training loss: 1.9074... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1417... Training loss: 2.0320... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1418... Training loss: 1.9261... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1419... Training loss: 1.8952... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1420... Training loss: 1.8824... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1421... Training loss: 2.0098... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1422... Training loss: 1.8984... 0.0467 sec/batch
Epoch: 1/20... Training Step: 1423... Training loss: 1.9830... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1424... Training loss: 1.8422... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1425... Training loss: 1.8177... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1426... Training loss: 1.9032... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1427... Training loss: 1.7705... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1428... Training loss: 1.9964... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1429... Training loss: 1.8992... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1430... Training loss: 1.8996... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1431... Training loss: 1.8731... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1432... Training loss: 1.9457... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1433... Training loss: 2.0337... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1434... Training loss: 1.8292... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1435... Training loss: 1.9160... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1436... Training loss: 1.8593... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1437... Training loss: 1.8976... 0.0571 sec/batch
Epoch: 1/20... Training Step: 1438... Training loss: 1.9223... 0.0460 sec/batch
Epoch: 1/20... Training Step: 1439... Training loss: 2.0078... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1440... Training loss: 1.8812... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1441... Training loss: 1.8122... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1442... Training loss: 1.9679... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1443... Training loss: 1.9591... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1444... Training loss: 1.9178... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1445... Training loss: 1.9514... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1446... Training loss: 1.9988... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1447... Training loss: 2.1668... 0.0541 sec/batch
Epoch: 1/20... Training Step: 1448... Training loss: 1.8369... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1449... Training loss: 1.9691... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1450... Training loss: 1.9114... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1451... Training loss: 1.8606... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1452... Training loss: 1.8635... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1453... Training loss: 2.0339... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1454... Training loss: 1.9881... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1455... Training loss: 2.0241... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1456... Training loss: 1.8239... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1457... Training loss: 1.8504... 0.0498 sec/batch
Epoch: 1/20... Training Step: 1458... Training loss: 1.9043... 0.0550 sec/batch
Epoch: 1/20... Training Step: 1459... Training loss: 1.8101... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1460... Training loss: 1.9004... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1461... Training loss: 1.8693... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1462... Training loss: 1.8140... 0.0522 sec/batch
Epoch: 1/20... Training Step: 1463... Training loss: 1.8840... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1464... Training loss: 1.8989... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1465... Training loss: 1.9586... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1466... Training loss: 1.8874... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1467... Training loss: 1.8829... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1468... Training loss: 1.8606... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1469... Training loss: 1.8571... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1470... Training loss: 1.8554... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1471... Training loss: 1.9150... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1472... Training loss: 2.0027... 0.0510 sec/batch
Epoch: 1/20... Training Step: 1473... Training loss: 1.9283... 0.0467 sec/batch
Epoch: 1/20... Training Step: 1474... Training loss: 2.0892... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1475... Training loss: 1.8796... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1476... Training loss: 1.7712... 0.0515 sec/batch
Epoch: 1/20... Training Step: 1477... Training loss: 1.9699... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1478... Training loss: 1.9567... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1479... Training loss: 1.9756... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1480... Training loss: 1.9616... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1481... Training loss: 1.9019... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1482... Training loss: 1.8833... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1483... Training loss: 1.8813... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1484... Training loss: 1.9475... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1485... Training loss: 1.8666... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1486... Training loss: 1.9901... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1487... Training loss: 1.9398... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1488... Training loss: 1.9158... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1489... Training loss: 1.9791... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1490... Training loss: 1.9438... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1491... Training loss: 1.9889... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1492... Training loss: 1.8848... 0.0472 sec/batch
Epoch: 1/20... Training Step: 1493... Training loss: 2.0902... 0.0458 sec/batch
Epoch: 1/20... Training Step: 1494... Training loss: 1.7997... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1495... Training loss: 1.8650... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1496... Training loss: 1.8639... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1497... Training loss: 1.9147... 0.0581 sec/batch
Epoch: 1/20... Training Step: 1498... Training loss: 1.9252... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1499... Training loss: 1.9480... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1500... Training loss: 1.9310... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1501... Training loss: 1.9062... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1502... Training loss: 1.8779... 0.0517 sec/batch
Epoch: 1/20... Training Step: 1503... Training loss: 1.9642... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1504... Training loss: 2.1131... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1505... Training loss: 1.8425... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1506... Training loss: 1.8078... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1507... Training loss: 1.9778... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1508... Training loss: 1.7769... 0.0449 sec/batch
Epoch: 1/20... Training Step: 1509... Training loss: 1.9095... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1510... Training loss: 1.8966... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1511... Training loss: 1.8686... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1512... Training loss: 1.7274... 0.0391 sec/batch
Epoch: 1/20... Training Step: 1513... Training loss: 1.8104... 0.0492 sec/batch
Epoch: 1/20... Training Step: 1514... Training loss: 1.8395... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1515... Training loss: 1.8388... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1516... Training loss: 1.9476... 0.0444 sec/batch
Epoch: 1/20... Training Step: 1517... Training loss: 1.9236... 0.0564 sec/batch
Epoch: 1/20... Training Step: 1518... Training loss: 1.9309... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1519... Training loss: 1.9797... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1520... Training loss: 1.8599... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1521... Training loss: 1.8768... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1522... Training loss: 2.0102... 0.0457 sec/batch
Epoch: 1/20... Training Step: 1523... Training loss: 1.9534... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1524... Training loss: 2.0476... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1525... Training loss: 1.9902... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1526... Training loss: 1.9217... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1527... Training loss: 1.7056... 0.0485 sec/batch
Epoch: 1/20... Training Step: 1528... Training loss: 1.9353... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1529... Training loss: 1.9342... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1530... Training loss: 1.9322... 0.0558 sec/batch
Epoch: 1/20... Training Step: 1531... Training loss: 1.9793... 0.0686 sec/batch
Epoch: 1/20... Training Step: 1532... Training loss: 2.0468... 0.0626 sec/batch
Epoch: 1/20... Training Step: 1533... Training loss: 1.8763... 0.0668 sec/batch
Epoch: 1/20... Training Step: 1534... Training loss: 1.9376... 0.0544 sec/batch
Epoch: 1/20... Training Step: 1535... Training loss: 1.9115... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1536... Training loss: 1.7726... 0.0510 sec/batch
Epoch: 1/20... Training Step: 1537... Training loss: 1.8188... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1538... Training loss: 1.8992... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1539... Training loss: 2.0380... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1540... Training loss: 1.9682... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1541... Training loss: 1.9927... 0.0465 sec/batch
Epoch: 1/20... Training Step: 1542... Training loss: 1.8867... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1543... Training loss: 1.8724... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1544... Training loss: 1.9843... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1545... Training loss: 1.7377... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1546... Training loss: 1.9668... 0.0497 sec/batch
Epoch: 1/20... Training Step: 1547... Training loss: 1.8706... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1548... Training loss: 1.6931... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1549... Training loss: 1.6611... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1550... Training loss: 1.8671... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1551... Training loss: 1.8225... 0.0580 sec/batch
Epoch: 1/20... Training Step: 1552... Training loss: 1.8630... 0.0481 sec/batch
Epoch: 1/20... Training Step: 1553... Training loss: 2.0261... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1554... Training loss: 1.9177... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1555... Training loss: 1.9236... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1556... Training loss: 1.9685... 0.0550 sec/batch
Epoch: 1/20... Training Step: 1557... Training loss: 2.0634... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1558... Training loss: 1.9117... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1559... Training loss: 1.9440... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1560... Training loss: 1.9130... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1561... Training loss: 1.7722... 0.0464 sec/batch
Epoch: 1/20... Training Step: 1562... Training loss: 1.9678... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1563... Training loss: 1.8631... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1564... Training loss: 1.9103... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1565... Training loss: 1.9416... 0.0462 sec/batch
Epoch: 1/20... Training Step: 1566... Training loss: 2.1258... 0.0552 sec/batch
Epoch: 1/20... Training Step: 1567... Training loss: 1.9625... 0.0439 sec/batch
Epoch: 1/20... Training Step: 1568... Training loss: 1.8000... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1569... Training loss: 1.8396... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1570... Training loss: 1.8806... 0.0519 sec/batch
Epoch: 1/20... Training Step: 1571... Training loss: 1.8675... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1572... Training loss: 1.8470... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1573... Training loss: 1.8498... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1574... Training loss: 1.7829... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1575... Training loss: 1.9615... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1576... Training loss: 1.9026... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1577... Training loss: 1.8054... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1578... Training loss: 2.0210... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1579... Training loss: 1.9094... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1580... Training loss: 1.8936... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1581... Training loss: 1.9959... 0.0482 sec/batch
Epoch: 1/20... Training Step: 1582... Training loss: 1.9728... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1583... Training loss: 1.9299... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1584... Training loss: 1.8524... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1585... Training loss: 1.8749... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1586... Training loss: 1.8470... 0.0495 sec/batch
Epoch: 1/20... Training Step: 1587... Training loss: 1.8002... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1588... Training loss: 1.8692... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1589... Training loss: 1.9429... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1590... Training loss: 1.8768... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1591... Training loss: 1.8398... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1592... Training loss: 1.9562... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1593... Training loss: 1.9128... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1594... Training loss: 1.9370... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1595... Training loss: 1.9329... 0.0430 sec/batch
Epoch: 1/20... Training Step: 1596... Training loss: 1.8854... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1597... Training loss: 1.7690... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1598... Training loss: 1.9295... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1599... Training loss: 1.9079... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1600... Training loss: 1.8773... 0.0382 sec/batch
Epoch: 1/20... Training Step: 1601... Training loss: 1.9872... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1602... Training loss: 1.9040... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1603... Training loss: 2.0206... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1604... Training loss: 2.0416... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1605... Training loss: 2.0803... 0.0464 sec/batch
Epoch: 1/20... Training Step: 1606... Training loss: 1.9059... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1607... Training loss: 2.0837... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1608... Training loss: 2.1277... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1609... Training loss: 2.0102... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1610... Training loss: 1.9450... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1611... Training loss: 2.1051... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1612... Training loss: 1.9424... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1613... Training loss: 1.9584... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1614... Training loss: 1.9371... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1615... Training loss: 2.0253... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1616... Training loss: 1.8324... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1617... Training loss: 1.9641... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1618... Training loss: 2.0373... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1619... Training loss: 2.0086... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1620... Training loss: 1.9401... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1621... Training loss: 2.0060... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1622... Training loss: 1.9353... 0.0383 sec/batch
Epoch: 1/20... Training Step: 1623... Training loss: 1.8322... 0.0387 sec/batch
Epoch: 1/20... Training Step: 1624... Training loss: 1.8584... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1625... Training loss: 1.9614... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1626... Training loss: 1.8934... 0.0557 sec/batch
Epoch: 1/20... Training Step: 1627... Training loss: 1.7580... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1628... Training loss: 1.9446... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1629... Training loss: 1.8776... 0.0441 sec/batch
Epoch: 1/20... Training Step: 1630... Training loss: 1.9348... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1631... Training loss: 1.9033... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1632... Training loss: 1.7664... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1633... Training loss: 1.8907... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1634... Training loss: 1.7888... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1635... Training loss: 1.9074... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1636... Training loss: 1.9871... 0.0505 sec/batch
Epoch: 1/20... Training Step: 1637... Training loss: 1.9629... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1638... Training loss: 1.8138... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1639... Training loss: 1.9230... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1640... Training loss: 1.9364... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1641... Training loss: 1.7421... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1642... Training loss: 1.8637... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1643... Training loss: 1.7836... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1644... Training loss: 1.8256... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1645... Training loss: 1.8842... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1646... Training loss: 1.8050... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1647... Training loss: 1.8149... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1648... Training loss: 1.9695... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1649... Training loss: 1.7752... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1650... Training loss: 1.8372... 0.0482 sec/batch
Epoch: 1/20... Training Step: 1651... Training loss: 1.8682... 0.0472 sec/batch
Epoch: 1/20... Training Step: 1652... Training loss: 2.0440... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1653... Training loss: 1.9352... 0.0442 sec/batch
Epoch: 1/20... Training Step: 1654... Training loss: 1.8208... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1655... Training loss: 1.7874... 0.0489 sec/batch
Epoch: 1/20... Training Step: 1656... Training loss: 1.7985... 0.0427 sec/batch
Epoch: 1/20... Training Step: 1657... Training loss: 1.8929... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1658... Training loss: 1.9182... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1659... Training loss: 1.9819... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1660... Training loss: 1.7986... 0.0515 sec/batch
Epoch: 1/20... Training Step: 1661... Training loss: 1.9474... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1662... Training loss: 1.7688... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1663... Training loss: 1.9326... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1664... Training loss: 1.9724... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1665... Training loss: 2.1704... 0.0519 sec/batch
Epoch: 1/20... Training Step: 1666... Training loss: 1.9570... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1667... Training loss: 1.9009... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1668... Training loss: 2.0357... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1669... Training loss: 1.9019... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1670... Training loss: 1.9496... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1671... Training loss: 1.8492... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1672... Training loss: 1.9189... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1673... Training loss: 1.9667... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1674... Training loss: 1.8897... 0.0386 sec/batch
Epoch: 1/20... Training Step: 1675... Training loss: 2.0417... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1676... Training loss: 1.8799... 0.0467 sec/batch
Epoch: 1/20... Training Step: 1677... Training loss: 1.9825... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1678... Training loss: 1.9828... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1679... Training loss: 1.9424... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1680... Training loss: 1.8064... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1681... Training loss: 1.9524... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1682... Training loss: 1.8715... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1683... Training loss: 1.9531... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1684... Training loss: 1.8415... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1685... Training loss: 1.8268... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1686... Training loss: 1.8600... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1687... Training loss: 1.9082... 0.0386 sec/batch
Epoch: 1/20... Training Step: 1688... Training loss: 2.0216... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1689... Training loss: 1.8956... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1690... Training loss: 1.8904... 0.0479 sec/batch
Epoch: 1/20... Training Step: 1691... Training loss: 1.8531... 0.0437 sec/batch
Epoch: 1/20... Training Step: 1692... Training loss: 1.8508... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1693... Training loss: 1.8041... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1694... Training loss: 1.8311... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1695... Training loss: 1.9397... 0.0503 sec/batch
Epoch: 1/20... Training Step: 1696... Training loss: 1.7993... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1697... Training loss: 1.8166... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1698... Training loss: 1.8607... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1699... Training loss: 1.9004... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1700... Training loss: 1.8964... 0.0498 sec/batch
Epoch: 1/20... Training Step: 1701... Training loss: 1.7898... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1702... Training loss: 1.8678... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1703... Training loss: 1.8363... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1704... Training loss: 1.9133... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1705... Training loss: 1.9241... 0.0477 sec/batch
Epoch: 1/20... Training Step: 1706... Training loss: 2.0068... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1707... Training loss: 1.9066... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1708... Training loss: 1.8387... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1709... Training loss: 1.9174... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1710... Training loss: 1.8614... 0.0575 sec/batch
Epoch: 1/20... Training Step: 1711... Training loss: 1.8837... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1712... Training loss: 1.9413... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1713... Training loss: 1.7487... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1714... Training loss: 1.9895... 0.0425 sec/batch
Epoch: 1/20... Training Step: 1715... Training loss: 2.0244... 0.0529 sec/batch
Epoch: 1/20... Training Step: 1716... Training loss: 2.0081... 0.0392 sec/batch
Epoch: 1/20... Training Step: 1717... Training loss: 1.7871... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1718... Training loss: 1.9207... 0.0381 sec/batch
Epoch: 1/20... Training Step: 1719... Training loss: 1.8954... 0.0458 sec/batch
Epoch: 1/20... Training Step: 1720... Training loss: 1.8796... 0.0502 sec/batch
Epoch: 1/20... Training Step: 1721... Training loss: 1.8685... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1722... Training loss: 1.8906... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1723... Training loss: 1.7833... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1724... Training loss: 1.7574... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1725... Training loss: 1.8245... 0.0513 sec/batch
Epoch: 1/20... Training Step: 1726... Training loss: 1.8595... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1727... Training loss: 1.7643... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1728... Training loss: 1.9122... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1729... Training loss: 1.8324... 0.0490 sec/batch
Epoch: 1/20... Training Step: 1730... Training loss: 1.7814... 0.0464 sec/batch
Epoch: 1/20... Training Step: 1731... Training loss: 1.9045... 0.0441 sec/batch
Epoch: 1/20... Training Step: 1732... Training loss: 1.9574... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1733... Training loss: 1.9896... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1734... Training loss: 2.0340... 0.0514 sec/batch
Epoch: 1/20... Training Step: 1735... Training loss: 1.9365... 0.0536 sec/batch
Epoch: 1/20... Training Step: 1736... Training loss: 1.9548... 0.0529 sec/batch
Epoch: 1/20... Training Step: 1737... Training loss: 1.8637... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1738... Training loss: 1.9193... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1739... Training loss: 1.8692... 0.0457 sec/batch
Epoch: 1/20... Training Step: 1740... Training loss: 1.9266... 0.0564 sec/batch
Epoch: 1/20... Training Step: 1741... Training loss: 1.8403... 0.0473 sec/batch
Epoch: 1/20... Training Step: 1742... Training loss: 1.8186... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1743... Training loss: 2.0206... 0.0539 sec/batch
Epoch: 1/20... Training Step: 1744... Training loss: 1.7536... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1745... Training loss: 1.8629... 0.0519 sec/batch
Epoch: 1/20... Training Step: 1746... Training loss: 1.7203... 0.0502 sec/batch
Epoch: 1/20... Training Step: 1747... Training loss: 1.8727... 0.0525 sec/batch
Epoch: 1/20... Training Step: 1748... Training loss: 1.9488... 0.0572 sec/batch
Epoch: 1/20... Training Step: 1749... Training loss: 1.7766... 0.0490 sec/batch
Epoch: 1/20... Training Step: 1750... Training loss: 1.9335... 0.0526 sec/batch
Epoch: 1/20... Training Step: 1751... Training loss: 1.9091... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1752... Training loss: 1.8063... 0.0520 sec/batch
Epoch: 1/20... Training Step: 1753... Training loss: 1.7882... 0.0468 sec/batch
Epoch: 1/20... Training Step: 1754... Training loss: 1.8673... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1755... Training loss: 1.8656... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1756... Training loss: 1.7358... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1757... Training loss: 1.8277... 0.0545 sec/batch
Epoch: 1/20... Training Step: 1758... Training loss: 1.8255... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1759... Training loss: 1.9391... 0.0413 sec/batch
Epoch: 1/20... Training Step: 1760... Training loss: 1.7957... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1761... Training loss: 1.9893... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1762... Training loss: 1.9358... 0.0538 sec/batch
Epoch: 1/20... Training Step: 1763... Training loss: 1.8000... 0.0478 sec/batch
Epoch: 1/20... Training Step: 1764... Training loss: 1.8779... 0.0400 sec/batch
Epoch: 1/20... Training Step: 1765... Training loss: 1.8400... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1766... Training loss: 1.9678... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1767... Training loss: 1.7507... 0.0534 sec/batch
Epoch: 1/20... Training Step: 1768... Training loss: 1.8333... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1769... Training loss: 1.9699... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1770... Training loss: 1.8657... 0.0393 sec/batch
Epoch: 1/20... Training Step: 1771... Training loss: 1.7247... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1772... Training loss: 1.7895... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1773... Training loss: 1.8184... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1774... Training loss: 2.0077... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1775... Training loss: 1.9842... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1776... Training loss: 1.7409... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1777... Training loss: 1.9170... 0.0473 sec/batch
Epoch: 1/20... Training Step: 1778... Training loss: 1.7504... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1779... Training loss: 1.8615... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1780... Training loss: 1.9477... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1781... Training loss: 1.9209... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1782... Training loss: 1.8150... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1783... Training loss: 1.8937... 0.0472 sec/batch
Epoch: 1/20... Training Step: 1784... Training loss: 1.7600... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1785... Training loss: 1.8312... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1786... Training loss: 1.8583... 0.0444 sec/batch
Epoch: 1/20... Training Step: 1787... Training loss: 1.9465... 0.0498 sec/batch
Epoch: 1/20... Training Step: 1788... Training loss: 1.8564... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1789... Training loss: 1.8987... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1790... Training loss: 1.8160... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1791... Training loss: 1.9308... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1792... Training loss: 2.0700... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1793... Training loss: 1.8164... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1794... Training loss: 1.7118... 0.0396 sec/batch
Epoch: 1/20... Training Step: 1795... Training loss: 1.6621... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1796... Training loss: 1.8656... 0.0399 sec/batch
Epoch: 1/20... Training Step: 1797... Training loss: 1.7216... 0.0477 sec/batch
Epoch: 1/20... Training Step: 1798... Training loss: 1.7725... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1799... Training loss: 1.7648... 0.0426 sec/batch
Epoch: 1/20... Training Step: 1800... Training loss: 1.8781... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1801... Training loss: 1.8934... 0.0490 sec/batch
Epoch: 1/20... Training Step: 1802... Training loss: 1.8667... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1803... Training loss: 1.8608... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1804... Training loss: 1.7361... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1805... Training loss: 1.7277... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1806... Training loss: 1.8081... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1807... Training loss: 1.7926... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1808... Training loss: 1.8327... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1809... Training loss: 1.7475... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1810... Training loss: 1.6128... 0.0387 sec/batch
Epoch: 1/20... Training Step: 1811... Training loss: 1.7636... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1812... Training loss: 1.8989... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1813... Training loss: 1.8076... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1814... Training loss: 1.9087... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1815... Training loss: 1.7926... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1816... Training loss: 1.9463... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1817... Training loss: 1.8427... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1818... Training loss: 1.7570... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1819... Training loss: 1.8139... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1820... Training loss: 1.8648... 0.0402 sec/batch
Epoch: 1/20... Training Step: 1821... Training loss: 1.8889... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1822... Training loss: 1.9709... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1823... Training loss: 1.7849... 0.0444 sec/batch
Epoch: 1/20... Training Step: 1824... Training loss: 1.8521... 0.0382 sec/batch
Epoch: 1/20... Training Step: 1825... Training loss: 1.9022... 0.0388 sec/batch
Epoch: 1/20... Training Step: 1826... Training loss: 1.7731... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1827... Training loss: 1.8877... 0.0511 sec/batch
Epoch: 1/20... Training Step: 1828... Training loss: 1.8210... 0.0484 sec/batch
Epoch: 1/20... Training Step: 1829... Training loss: 1.9419... 0.0408 sec/batch
Epoch: 1/20... Training Step: 1830... Training loss: 1.8449... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1831... Training loss: 1.9190... 0.0501 sec/batch
Epoch: 1/20... Training Step: 1832... Training loss: 1.8902... 0.0488 sec/batch
Epoch: 1/20... Training Step: 1833... Training loss: 1.8651... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1834... Training loss: 1.9022... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1835... Training loss: 1.8221... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1836... Training loss: 1.8797... 0.0480 sec/batch
Epoch: 1/20... Training Step: 1837... Training loss: 1.8950... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1838... Training loss: 1.8903... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1839... Training loss: 1.7431... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1840... Training loss: 1.8420... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1841... Training loss: 1.7254... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1842... Training loss: 1.9845... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1843... Training loss: 1.6929... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1844... Training loss: 1.6838... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1845... Training loss: 1.8212... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1846... Training loss: 1.9326... 0.0444 sec/batch
Epoch: 1/20... Training Step: 1847... Training loss: 1.9102... 0.0492 sec/batch
Epoch: 1/20... Training Step: 1848... Training loss: 1.8292... 0.0464 sec/batch
Epoch: 1/20... Training Step: 1849... Training loss: 1.7752... 0.0441 sec/batch
Epoch: 1/20... Training Step: 1850... Training loss: 2.0350... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1851... Training loss: 1.8261... 0.0482 sec/batch
Epoch: 1/20... Training Step: 1852... Training loss: 1.9479... 0.0561 sec/batch
Epoch: 1/20... Training Step: 1853... Training loss: 2.0483... 0.0395 sec/batch
Epoch: 1/20... Training Step: 1854... Training loss: 1.9959... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1855... Training loss: 1.8854... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1856... Training loss: 1.6962... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1857... Training loss: 2.1049... 0.0531 sec/batch
Epoch: 1/20... Training Step: 1858... Training loss: 1.9853... 0.0421 sec/batch
Epoch: 1/20... Training Step: 1859... Training loss: 1.8139... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1860... Training loss: 1.9059... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1861... Training loss: 1.9253... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1862... Training loss: 1.8520... 0.0529 sec/batch
Epoch: 1/20... Training Step: 1863... Training loss: 1.8759... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1864... Training loss: 1.9985... 0.0389 sec/batch
Epoch: 1/20... Training Step: 1865... Training loss: 1.7756... 0.0424 sec/batch
Epoch: 1/20... Training Step: 1866... Training loss: 1.7448... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1867... Training loss: 1.8898... 0.0435 sec/batch
Epoch: 1/20... Training Step: 1868... Training loss: 1.8403... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1869... Training loss: 1.8906... 0.0466 sec/batch
Epoch: 1/20... Training Step: 1870... Training loss: 1.8588... 0.0450 sec/batch
Epoch: 1/20... Training Step: 1871... Training loss: 1.8856... 0.0438 sec/batch
Epoch: 1/20... Training Step: 1872... Training loss: 1.6928... 0.0506 sec/batch
Epoch: 1/20... Training Step: 1873... Training loss: 1.7723... 0.0468 sec/batch
Epoch: 1/20... Training Step: 1874... Training loss: 1.8125... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1875... Training loss: 1.8564... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1876... Training loss: 1.9563... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1877... Training loss: 1.8216... 0.0562 sec/batch
Epoch: 1/20... Training Step: 1878... Training loss: 1.8421... 0.0410 sec/batch
Epoch: 1/20... Training Step: 1879... Training loss: 1.8181... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1880... Training loss: 2.0237... 0.0403 sec/batch
Epoch: 1/20... Training Step: 1881... Training loss: 1.9626... 0.0479 sec/batch
Epoch: 1/20... Training Step: 1882... Training loss: 1.7263... 0.0513 sec/batch
Epoch: 1/20... Training Step: 1883... Training loss: 1.9699... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1884... Training loss: 1.9955... 0.0390 sec/batch
Epoch: 1/20... Training Step: 1885... Training loss: 1.6934... 0.0394 sec/batch
Epoch: 1/20... Training Step: 1886... Training loss: 1.8767... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1887... Training loss: 1.9660... 0.0452 sec/batch
Epoch: 1/20... Training Step: 1888... Training loss: 1.8587... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1889... Training loss: 1.8291... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1890... Training loss: 1.8790... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1891... Training loss: 1.9502... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1892... Training loss: 1.9307... 0.0510 sec/batch
Epoch: 1/20... Training Step: 1893... Training loss: 1.8504... 0.0465 sec/batch
Epoch: 1/20... Training Step: 1894... Training loss: 1.8561... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1895... Training loss: 1.7630... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1896... Training loss: 1.8118... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1897... Training loss: 1.8970... 0.0515 sec/batch
Epoch: 1/20... Training Step: 1898... Training loss: 1.9294... 0.0404 sec/batch
Epoch: 1/20... Training Step: 1899... Training loss: 1.8225... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1900... Training loss: 1.7470... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1901... Training loss: 1.9410... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1902... Training loss: 1.9223... 0.0564 sec/batch
Epoch: 1/20... Training Step: 1903... Training loss: 1.8343... 0.0644 sec/batch
Epoch: 1/20... Training Step: 1904... Training loss: 1.6468... 0.0629 sec/batch
Epoch: 1/20... Training Step: 1905... Training loss: 1.7257... 0.0479 sec/batch
Epoch: 1/20... Training Step: 1906... Training loss: 1.9627... 0.0450 sec/batch
Epoch: 1/20... Training Step: 1907... Training loss: 1.8896... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1908... Training loss: 2.0066... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1909... Training loss: 1.8928... 0.0423 sec/batch
Epoch: 1/20... Training Step: 1910... Training loss: 1.8036... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1911... Training loss: 1.7925... 0.0546 sec/batch
Epoch: 1/20... Training Step: 1912... Training loss: 1.8266... 0.0434 sec/batch
Epoch: 1/20... Training Step: 1913... Training loss: 1.8172... 0.0432 sec/batch
Epoch: 1/20... Training Step: 1914... Training loss: 1.8659... 0.0462 sec/batch
Epoch: 1/20... Training Step: 1915... Training loss: 1.8203... 0.0501 sec/batch
Epoch: 1/20... Training Step: 1916... Training loss: 1.8728... 0.0544 sec/batch
Epoch: 1/20... Training Step: 1917... Training loss: 1.7977... 0.0412 sec/batch
Epoch: 1/20... Training Step: 1918... Training loss: 2.0304... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1919... Training loss: 1.7881... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1920... Training loss: 1.8239... 0.0520 sec/batch
Epoch: 1/20... Training Step: 1921... Training loss: 1.9075... 0.0440 sec/batch
Epoch: 1/20... Training Step: 1922... Training loss: 1.6758... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1923... Training loss: 1.7412... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1924... Training loss: 1.7517... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1925... Training loss: 1.8555... 0.0422 sec/batch
Epoch: 1/20... Training Step: 1926... Training loss: 1.7345... 0.0510 sec/batch
Epoch: 1/20... Training Step: 1927... Training loss: 1.7730... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1928... Training loss: 1.9322... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1929... Training loss: 1.8849... 0.0443 sec/batch
Epoch: 1/20... Training Step: 1930... Training loss: 1.9355... 0.0488 sec/batch
Epoch: 1/20... Training Step: 1931... Training loss: 1.8315... 0.0472 sec/batch
Epoch: 1/20... Training Step: 1932... Training loss: 1.7589... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1933... Training loss: 1.8005... 0.0401 sec/batch
Epoch: 1/20... Training Step: 1934... Training loss: 2.0368... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1935... Training loss: 1.8203... 0.0428 sec/batch
Epoch: 1/20... Training Step: 1936... Training loss: 1.9454... 0.0451 sec/batch
Epoch: 1/20... Training Step: 1937... Training loss: 1.9900... 0.0455 sec/batch
Epoch: 1/20... Training Step: 1938... Training loss: 2.1516... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1939... Training loss: 1.9354... 0.0467 sec/batch
Epoch: 1/20... Training Step: 1940... Training loss: 2.1602... 0.0458 sec/batch
Epoch: 1/20... Training Step: 1941... Training loss: 1.8715... 0.0641 sec/batch
Epoch: 1/20... Training Step: 1942... Training loss: 1.8124... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1943... Training loss: 1.9071... 0.0398 sec/batch
Epoch: 1/20... Training Step: 1944... Training loss: 1.8065... 0.0405 sec/batch
Epoch: 1/20... Training Step: 1945... Training loss: 1.9907... 0.0447 sec/batch
Epoch: 1/20... Training Step: 1946... Training loss: 1.7465... 0.0550 sec/batch
Epoch: 1/20... Training Step: 1947... Training loss: 1.8864... 0.0397 sec/batch
Epoch: 1/20... Training Step: 1948... Training loss: 1.8107... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1949... Training loss: 1.8905... 0.0604 sec/batch
Epoch: 1/20... Training Step: 1950... Training loss: 1.8195... 0.0641 sec/batch
Epoch: 1/20... Training Step: 1951... Training loss: 1.8564... 0.0577 sec/batch
Epoch: 1/20... Training Step: 1952... Training loss: 1.9126... 0.0792 sec/batch
Epoch: 1/20... Training Step: 1953... Training loss: 1.8935... 0.0954 sec/batch
Epoch: 1/20... Training Step: 1954... Training loss: 1.7660... 0.0552 sec/batch
Epoch: 1/20... Training Step: 1955... Training loss: 2.0325... 0.0649 sec/batch
Epoch: 1/20... Training Step: 1956... Training loss: 1.9542... 0.0698 sec/batch
Epoch: 1/20... Training Step: 1957... Training loss: 1.8498... 0.0705 sec/batch
Epoch: 1/20... Training Step: 1958... Training loss: 1.8053... 0.0465 sec/batch
Epoch: 1/20... Training Step: 1959... Training loss: 1.8663... 0.0441 sec/batch
Epoch: 1/20... Training Step: 1960... Training loss: 1.8725... 0.0406 sec/batch
Epoch: 1/20... Training Step: 1961... Training loss: 1.7477... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1962... Training loss: 1.8891... 0.0433 sec/batch
Epoch: 1/20... Training Step: 1963... Training loss: 1.6917... 0.0454 sec/batch
Epoch: 1/20... Training Step: 1964... Training loss: 1.6373... 0.0457 sec/batch
Epoch: 1/20... Training Step: 1965... Training loss: 1.8024... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1966... Training loss: 1.6760... 0.0417 sec/batch
Epoch: 1/20... Training Step: 1967... Training loss: 1.9366... 0.0495 sec/batch
Epoch: 1/20... Training Step: 1968... Training loss: 1.9402... 0.0580 sec/batch
Epoch: 1/20... Training Step: 1969... Training loss: 2.0351... 0.0456 sec/batch
Epoch: 1/20... Training Step: 1970... Training loss: 1.9073... 0.0493 sec/batch
Epoch: 1/20... Training Step: 1971... Training loss: 1.8364... 0.0453 sec/batch
Epoch: 1/20... Training Step: 1972... Training loss: 1.8848... 0.0572 sec/batch
Epoch: 1/20... Training Step: 1973... Training loss: 1.8234... 0.0504 sec/batch
Epoch: 1/20... Training Step: 1974... Training loss: 1.9274... 0.0419 sec/batch
Epoch: 1/20... Training Step: 1975... Training loss: 1.9295... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1976... Training loss: 1.7568... 0.0445 sec/batch
Epoch: 1/20... Training Step: 1977... Training loss: 1.9185... 0.0630 sec/batch
Epoch: 1/20... Training Step: 1978... Training loss: 1.8088... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1979... Training loss: 1.8925... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1980... Training loss: 1.8252... 0.0441 sec/batch
Epoch: 1/20... Training Step: 1981... Training loss: 1.9095... 0.0503 sec/batch
Epoch: 1/20... Training Step: 1982... Training loss: 1.7736... 0.0560 sec/batch
Epoch: 1/20... Training Step: 1983... Training loss: 1.8468... 0.0436 sec/batch
Epoch: 1/20... Training Step: 1984... Training loss: 2.0933... 0.0416 sec/batch
Epoch: 1/20... Training Step: 1985... Training loss: 2.0599... 0.0429 sec/batch
Epoch: 1/20... Training Step: 1986... Training loss: 1.8329... 0.0431 sec/batch
Epoch: 1/20... Training Step: 1987... Training loss: 1.8038... 0.0448 sec/batch
Epoch: 1/20... Training Step: 1988... Training loss: 1.8580... 0.0420 sec/batch
Epoch: 1/20... Training Step: 1989... Training loss: 1.7760... 0.0414 sec/batch
Epoch: 1/20... Training Step: 1990... Training loss: 1.7915... 0.0407 sec/batch
Epoch: 1/20... Training Step: 1991... Training loss: 1.7381... 0.0463 sec/batch
Epoch: 1/20... Training Step: 1992... Training loss: 1.8519... 0.0592 sec/batch
Epoch: 1/20... Training Step: 1993... Training loss: 2.1765... 0.0409 sec/batch
Epoch: 1/20... Training Step: 1994... Training loss: 2.0639... 0.0411 sec/batch
Epoch: 1/20... Training Step: 1995... Training loss: 1.8770... 0.0415 sec/batch
Epoch: 1/20... Training Step: 1996... Training loss: 1.9510... 0.0450 sec/batch
Epoch: 1/20... Training Step: 1997... Training loss: 1.8154... 0.0544 sec/batch
Epoch: 1/20... Training Step: 1998... Training loss: 1.8156... 0.0418 sec/batch
Epoch: 1/20... Training Step: 1999... Training loss: 1.8177... 0.0569 sec/batch
Epoch: 1/20... Training Step: 2000... Training loss: 1.8396... 0.0473 sec/batch
Epoch: 1/20... Training Step: 2001... Training loss: 1.7578... 0.0462 sec/batch
Epoch: 1/20... Training Step: 2002... Training loss: 1.7041... 0.0625 sec/batch
Epoch: 1/20... Training Step: 2003... Training loss: 1.7150... 0.0626 sec/batch
Epoch: 1/20... Training Step: 2004... Training loss: 1.7388... 0.0571 sec/batch
Epoch: 1/20... Training Step: 2005... Training loss: 1.8461... 0.0514 sec/batch
Epoch: 1/20... Training Step: 2006... Training loss: 1.7403... 0.0455 sec/batch
Epoch: 1/20... Training Step: 2007... Training loss: 1.8694... 0.0543 sec/batch
Epoch: 1/20... Training Step: 2008... Training loss: 1.8288... 0.0547 sec/batch
Epoch: 1/20... Training Step: 2009... Training loss: 1.7128... 0.0622 sec/batch
Epoch: 1/20... Training Step: 2010... Training loss: 1.8750... 0.0519 sec/batch
Epoch: 1/20... Training Step: 2011... Training loss: 1.7793... 0.0450 sec/batch
Epoch: 1/20... Training Step: 2012... Training loss: 1.7799... 0.0458 sec/batch
Epoch: 1/20... Training Step: 2013... Training loss: 1.7877... 0.0565 sec/batch
Epoch: 1/20... Training Step: 2014... Training loss: 1.7784... 0.0444 sec/batch
Epoch: 1/20... Training Step: 2015... Training loss: 1.8834... 0.0425 sec/batch
Epoch: 1/20... Training Step: 2016... Training loss: 1.7557... 0.0484 sec/batch
Epoch: 1/20... Training Step: 2017... Training loss: 1.7263... 0.0589 sec/batch
Epoch: 1/20... Training Step: 2018... Training loss: 1.6937... 0.0448 sec/batch
Epoch: 1/20... Training Step: 2019... Training loss: 1.7471... 0.0440 sec/batch
Epoch: 1/20... Training Step: 2020... Training loss: 1.7625... 0.0443 sec/batch
Epoch: 1/20... Training Step: 2021... Training loss: 1.8730... 0.0611 sec/batch
Epoch: 1/20... Training Step: 2022... Training loss: 1.7798... 0.0535 sec/batch
Epoch: 1/20... Training Step: 2023... Training loss: 1.9943... 0.0458 sec/batch
Epoch: 1/20... Training Step: 2024... Training loss: 1.8152... 0.0494 sec/batch
Epoch: 1/20... Training Step: 2025... Training loss: 1.7850... 0.0462 sec/batch
Epoch: 1/20... Training Step: 2026... Training loss: 1.7866... 0.0487 sec/batch
Epoch: 1/20... Training Step: 2027... Training loss: 1.8862... 0.0609 sec/batch
Epoch: 1/20... Training Step: 2028... Training loss: 1.6829... 0.0469 sec/batch
Epoch: 1/20... Training Step: 2029... Training loss: 1.9016... 0.0459 sec/batch
Epoch: 1/20... Training Step: 2030... Training loss: 1.7131... 0.0722 sec/batch
Epoch: 1/20... Training Step: 2031... Training loss: 1.7910... 0.0927 sec/batch
Epoch: 1/20... Training Step: 2032... Training loss: 1.8245... 0.0653 sec/batch
Epoch: 1/20... Training Step: 2033... Training loss: 1.8381... 0.0562 sec/batch
Epoch: 1/20... Training Step: 2034... Training loss: 1.8911... 0.0836 sec/batch
Epoch: 1/20... Training Step: 2035... Training loss: 1.8131... 0.0529 sec/batch
Epoch: 1/20... Training Step: 2036... Training loss: 1.8031... 0.0661 sec/batch
Epoch: 1/20... Training Step: 2037... Training loss: 1.7406... 0.0605 sec/batch
Epoch: 1/20... Training Step: 2038... Training loss: 1.8652... 0.0705 sec/batch
Epoch: 1/20... Training Step: 2039... Training loss: 1.8632... 0.0672 sec/batch
Epoch: 1/20... Training Step: 2040... Training loss: 1.9327... 0.0783 sec/batch
Epoch: 1/20... Training Step: 2041... Training loss: 1.7589... 0.0517 sec/batch
Epoch: 1/20... Training Step: 2042... Training loss: 1.8654... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2043... Training loss: 1.8892... 0.0499 sec/batch
Epoch: 1/20... Training Step: 2044... Training loss: 1.8274... 0.0547 sec/batch
Epoch: 1/20... Training Step: 2045... Training loss: 1.9152... 0.0954 sec/batch
Epoch: 1/20... Training Step: 2046... Training loss: 1.8301... 0.0613 sec/batch
Epoch: 1/20... Training Step: 2047... Training loss: 1.9187... 0.0657 sec/batch
Epoch: 1/20... Training Step: 2048... Training loss: 1.7876... 0.0695 sec/batch
Epoch: 1/20... Training Step: 2049... Training loss: 1.9332... 0.0796 sec/batch
Epoch: 1/20... Training Step: 2050... Training loss: 1.9773... 0.0760 sec/batch
Epoch: 1/20... Training Step: 2051... Training loss: 1.9284... 0.0938 sec/batch
Epoch: 1/20... Training Step: 2052... Training loss: 1.8460... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2053... Training loss: 1.7497... 0.0684 sec/batch
Epoch: 1/20... Training Step: 2054... Training loss: 1.7400... 0.0854 sec/batch
Epoch: 1/20... Training Step: 2055... Training loss: 1.7278... 0.0776 sec/batch
Epoch: 1/20... Training Step: 2056... Training loss: 1.8636... 0.0827 sec/batch
Epoch: 1/20... Training Step: 2057... Training loss: 1.7706... 0.0883 sec/batch
Epoch: 1/20... Training Step: 2058... Training loss: 1.8081... 0.0539 sec/batch
Epoch: 1/20... Training Step: 2059... Training loss: 1.7942... 0.0704 sec/batch
Epoch: 1/20... Training Step: 2060... Training loss: 1.8172... 0.0513 sec/batch
Epoch: 1/20... Training Step: 2061... Training loss: 1.8596... 0.0521 sec/batch
Epoch: 1/20... Training Step: 2062... Training loss: 1.8451... 0.0474 sec/batch
Epoch: 1/20... Training Step: 2063... Training loss: 1.7994... 0.0482 sec/batch
Epoch: 1/20... Training Step: 2064... Training loss: 1.8519... 0.0582 sec/batch
Epoch: 1/20... Training Step: 2065... Training loss: 1.8086... 0.0522 sec/batch
Epoch: 1/20... Training Step: 2066... Training loss: 1.8974... 0.0463 sec/batch
Epoch: 1/20... Training Step: 2067... Training loss: 1.8368... 0.0431 sec/batch
Epoch: 1/20... Training Step: 2068... Training loss: 1.9279... 0.0550 sec/batch
Epoch: 1/20... Training Step: 2069... Training loss: 1.9227... 0.0507 sec/batch
Epoch: 1/20... Training Step: 2070... Training loss: 1.7410... 0.0458 sec/batch
Epoch: 1/20... Training Step: 2071... Training loss: 1.8990... 0.0431 sec/batch
Epoch: 1/20... Training Step: 2072... Training loss: 1.9719... 0.0522 sec/batch
Epoch: 1/20... Training Step: 2073... Training loss: 1.7706... 0.0535 sec/batch
Epoch: 1/20... Training Step: 2074... Training loss: 1.9442... 0.0575 sec/batch
Epoch: 1/20... Training Step: 2075... Training loss: 1.8207... 0.0499 sec/batch
Epoch: 1/20... Training Step: 2076... Training loss: 1.8407... 0.0471 sec/batch
Epoch: 1/20... Training Step: 2077... Training loss: 1.7816... 0.0454 sec/batch
Epoch: 1/20... Training Step: 2078... Training loss: 1.8634... 0.0467 sec/batch
Epoch: 1/20... Training Step: 2079... Training loss: 1.9007... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2080... Training loss: 1.7810... 0.0577 sec/batch
Epoch: 1/20... Training Step: 2081... Training loss: 1.8517... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2082... Training loss: 1.7931... 0.0597 sec/batch
Epoch: 1/20... Training Step: 2083... Training loss: 1.8974... 0.0419 sec/batch
Epoch: 1/20... Training Step: 2084... Training loss: 1.8412... 0.0506 sec/batch
Epoch: 1/20... Training Step: 2085... Training loss: 1.7557... 0.0528 sec/batch
Epoch: 1/20... Training Step: 2086... Training loss: 1.7646... 0.0491 sec/batch
Epoch: 1/20... Training Step: 2087... Training loss: 1.8531... 0.0477 sec/batch
Epoch: 1/20... Training Step: 2088... Training loss: 2.0016... 0.0430 sec/batch
Epoch: 1/20... Training Step: 2089... Training loss: 1.8178... 0.0608 sec/batch
Epoch: 1/20... Training Step: 2090... Training loss: 1.7889... 0.0580 sec/batch
Epoch: 1/20... Training Step: 2091... Training loss: 1.8918... 0.0427 sec/batch
Epoch: 1/20... Training Step: 2092... Training loss: 1.7882... 0.0485 sec/batch
Epoch: 1/20... Training Step: 2093... Training loss: 1.8204... 0.0457 sec/batch
Epoch: 1/20... Training Step: 2094... Training loss: 1.8709... 0.0521 sec/batch
Epoch: 1/20... Training Step: 2095... Training loss: 1.8162... 0.0628 sec/batch
Epoch: 1/20... Training Step: 2096... Training loss: 1.9083... 0.0422 sec/batch
Epoch: 1/20... Training Step: 2097... Training loss: 1.9021... 0.0404 sec/batch
Epoch: 1/20... Training Step: 2098... Training loss: 1.7927... 0.0414 sec/batch
Epoch: 1/20... Training Step: 2099... Training loss: 1.7486... 0.0472 sec/batch
Epoch: 1/20... Training Step: 2100... Training loss: 1.8440... 0.0534 sec/batch
Epoch: 1/20... Training Step: 2101... Training loss: 1.7175... 0.0425 sec/batch
Epoch: 1/20... Training Step: 2102... Training loss: 1.8160... 0.0410 sec/batch
Epoch: 1/20... Training Step: 2103... Training loss: 2.0109... 0.0491 sec/batch
Epoch: 1/20... Training Step: 2104... Training loss: 1.8889... 0.0598 sec/batch
Epoch: 1/20... Training Step: 2105... Training loss: 1.8864... 0.0522 sec/batch
Epoch: 1/20... Training Step: 2106... Training loss: 1.8874... 0.0433 sec/batch
Epoch: 1/20... Training Step: 2107... Training loss: 1.7736... 0.0431 sec/batch
Epoch: 1/20... Training Step: 2108... Training loss: 1.9426... 0.0454 sec/batch
Epoch: 1/20... Training Step: 2109... Training loss: 1.8394... 0.0500 sec/batch
Epoch: 1/20... Training Step: 2110... Training loss: 1.7925... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2111... Training loss: 1.7411... 0.0430 sec/batch
Epoch: 1/20... Training Step: 2112... Training loss: 1.8661... 0.0433 sec/batch
Epoch: 1/20... Training Step: 2113... Training loss: 1.8250... 0.0398 sec/batch
Epoch: 1/20... Training Step: 2114... Training loss: 1.7702... 0.0445 sec/batch
Epoch: 1/20... Training Step: 2115... Training loss: 1.8137... 0.0521 sec/batch
Epoch: 1/20... Training Step: 2116... Training loss: 1.7258... 0.0433 sec/batch
Epoch: 1/20... Training Step: 2117... Training loss: 1.8202... 0.0436 sec/batch
Epoch: 1/20... Training Step: 2118... Training loss: 1.8235... 0.0420 sec/batch
Epoch: 1/20... Training Step: 2119... Training loss: 1.7126... 0.0405 sec/batch
Epoch: 1/20... Training Step: 2120... Training loss: 1.6951... 0.0537 sec/batch
Epoch: 1/20... Training Step: 2121... Training loss: 1.8596... 0.0463 sec/batch
Epoch: 1/20... Training Step: 2122... Training loss: 1.9728... 0.0424 sec/batch
Epoch: 1/20... Training Step: 2123... Training loss: 1.7572... 0.0452 sec/batch
Epoch: 1/20... Training Step: 2124... Training loss: 1.8903... 0.0408 sec/batch
Epoch: 1/20... Training Step: 2125... Training loss: 1.8883... 0.0492 sec/batch
Epoch: 1/20... Training Step: 2126... Training loss: 1.8725... 0.0410 sec/batch
Epoch: 1/20... Training Step: 2127... Training loss: 1.9051... 0.0420 sec/batch
Epoch: 1/20... Training Step: 2128... Training loss: 1.7427... 0.0407 sec/batch
Epoch: 1/20... Training Step: 2129... Training loss: 1.7062... 0.0402 sec/batch
Epoch: 1/20... Training Step: 2130... Training loss: 1.8981... 0.0425 sec/batch
Epoch: 1/20... Training Step: 2131... Training loss: 1.8269... 0.0462 sec/batch
Epoch: 1/20... Training Step: 2132... Training loss: 1.7857... 0.0395 sec/batch
Epoch: 1/20... Training Step: 2133... Training loss: 1.9360... 0.0407 sec/batch
Epoch: 1/20... Training Step: 2134... Training loss: 1.8618... 0.0397 sec/batch
Epoch: 1/20... Training Step: 2135... Training loss: 1.8257... 0.0439 sec/batch
Epoch: 1/20... Training Step: 2136... Training loss: 1.7996... 0.0451 sec/batch
Epoch: 1/20... Training Step: 2137... Training loss: 1.8406... 0.0434 sec/batch
Epoch: 1/20... Training Step: 2138... Training loss: 1.8257... 0.0457 sec/batch
Epoch: 1/20... Training Step: 2139... Training loss: 1.8273... 0.0452 sec/batch
Epoch: 1/20... Training Step: 2140... Training loss: 1.9236... 0.0464 sec/batch
Epoch: 1/20... Training Step: 2141... Training loss: 1.7897... 0.0419 sec/batch
Epoch: 1/20... Training Step: 2142... Training loss: 1.8619... 0.0402 sec/batch
Epoch: 1/20... Training Step: 2143... Training loss: 1.9337... 0.0415 sec/batch
Epoch: 1/20... Training Step: 2144... Training loss: 1.8194... 0.0432 sec/batch
Epoch: 1/20... Training Step: 2145... Training loss: 1.9452... 0.0479 sec/batch
Epoch: 1/20... Training Step: 2146... Training loss: 1.8365... 0.0475 sec/batch
Epoch: 1/20... Training Step: 2147... Training loss: 1.7006... 0.0403 sec/batch
Epoch: 1/20... Training Step: 2148... Training loss: 2.1631... 0.0403 sec/batch
Epoch: 1/20... Training Step: 2149... Training loss: 1.8896... 0.0409 sec/batch
Epoch: 1/20... Training Step: 2150... Training loss: 1.7780... 0.0524 sec/batch
Epoch: 1/20... Training Step: 2151... Training loss: 2.0168... 0.0471 sec/batch
Epoch: 1/20... Training Step: 2152... Training loss: 1.8372... 0.0394 sec/batch
Epoch: 1/20... Training Step: 2153... Training loss: 1.6935... 0.0405 sec/batch
Epoch: 1/20... Training Step: 2154... Training loss: 1.8534... 0.0442 sec/batch
Epoch: 1/20... Training Step: 2155... Training loss: 1.8536... 0.0412 sec/batch
Epoch: 1/20... Training Step: 2156... Training loss: 1.6290... 0.0455 sec/batch
Epoch: 1/20... Training Step: 2157... Training loss: 1.8707... 0.0429 sec/batch
Epoch: 1/20... Training Step: 2158... Training loss: 1.7839... 0.0397 sec/batch
Epoch: 1/20... Training Step: 2159... Training loss: 1.9341... 0.0406 sec/batch
Epoch: 1/20... Training Step: 2160... Training loss: 1.8651... 0.0510 sec/batch
Epoch: 1/20... Training Step: 2161... Training loss: 1.8397... 0.0429 sec/batch
Epoch: 1/20... Training Step: 2162... Training loss: 1.7740... 0.0404 sec/batch
Epoch: 1/20... Training Step: 2163... Training loss: 1.7320... 0.0430 sec/batch
Epoch: 1/20... Training Step: 2164... Training loss: 1.8314... 0.0410 sec/batch
Epoch: 1/20... Training Step: 2165... Training loss: 1.9341... 0.0467 sec/batch
Epoch: 1/20... Training Step: 2166... Training loss: 1.7550... 0.0442 sec/batch
Epoch: 1/20... Training Step: 2167... Training loss: 1.7552... 0.0409 sec/batch
Epoch: 1/20... Training Step: 2168... Training loss: 1.7108... 0.0404 sec/batch
Epoch: 1/20... Training Step: 2169... Training loss: 1.8611... 0.0419 sec/batch
Epoch: 1/20... Training Step: 2170... Training loss: 1.7648... 0.0447 sec/batch
Epoch: 1/20... Training Step: 2171... Training loss: 1.7006... 0.0424 sec/batch
Epoch: 1/20... Training Step: 2172... Training loss: 1.8630... 0.0414 sec/batch
Epoch: 1/20... Training Step: 2173... Training loss: 1.9385... 0.0400 sec/batch
Epoch: 1/20... Training Step: 2174... Training loss: 1.7884... 0.0420 sec/batch
Epoch: 1/20... Training Step: 2175... Training loss: 1.9053... 0.0442 sec/batch
Epoch: 1/20... Training Step: 2176... Training loss: 1.8824... 0.0475 sec/batch
Epoch: 1/20... Training Step: 2177... Training loss: 1.8519... 0.0403 sec/batch
Epoch: 1/20... Training Step: 2178... Training loss: 1.9383... 0.0467 sec/batch
Epoch: 1/20... Training Step: 2179... Training loss: 1.8980... 0.0433 sec/batch
Epoch: 1/20... Training Step: 2180... Training loss: 1.9447... 0.0493 sec/batch
Epoch: 1/20... Training Step: 2181... Training loss: 1.7830... 0.0439 sec/batch
Epoch: 1/20... Training Step: 2182... Training loss: 2.1195... 0.0444 sec/batch
Epoch: 1/20... Training Step: 2183... Training loss: 1.7447... 0.0399 sec/batch
Epoch: 1/20... Training Step: 2184... Training loss: 1.9302... 0.0503 sec/batch
Epoch: 1/20... Training Step: 2185... Training loss: 1.9541... 0.0441 sec/batch
Epoch: 1/20... Training Step: 2186... Training loss: 1.8494... 0.0448 sec/batch
Epoch: 1/20... Training Step: 2187... Training loss: 1.8755... 0.0417 sec/batch
Epoch: 1/20... Training Step: 2188... Training loss: 1.7187... 0.0399 sec/batch
Epoch: 1/20... Training Step: 2189... Training loss: 1.8228... 0.0511 sec/batch
Epoch: 1/20... Training Step: 2190... Training loss: 1.8009... 0.0524 sec/batch
Epoch: 1/20... Training Step: 2191... Training loss: 2.0105... 0.0408 sec/batch
Epoch: 1/20... Training Step: 2192... Training loss: 1.8196... 0.0636 sec/batch
Epoch: 1/20... Training Step: 2193... Training loss: 1.7949... 0.0523 sec/batch
Epoch: 1/20... Training Step: 2194... Training loss: 1.7218... 0.0686 sec/batch
Epoch: 1/20... Training Step: 2195... Training loss: 1.8780... 0.0700 sec/batch
Epoch: 1/20... Training Step: 2196... Training loss: 1.9737... 0.0554 sec/batch
Epoch: 1/20... Training Step: 2197... Training loss: 1.8288... 0.0825 sec/batch
Epoch: 1/20... Training Step: 2198... Training loss: 1.8913... 0.0561 sec/batch
Epoch: 1/20... Training Step: 2199... Training loss: 1.7531... 0.0602 sec/batch
Epoch: 1/20... Training Step: 2200... Training loss: 1.8084... 0.0548 sec/batch
Epoch: 1/20... Training Step: 2201... Training loss: 1.9372... 0.0773 sec/batch
Epoch: 1/20... Training Step: 2202... Training loss: 1.9769... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2203... Training loss: 1.9667... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2204... Training loss: 1.6924... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2205... Training loss: 1.8600... 0.0769 sec/batch
Epoch: 1/20... Training Step: 2206... Training loss: 1.8451... 0.0746 sec/batch
Epoch: 1/20... Training Step: 2207... Training loss: 1.9027... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2208... Training loss: 1.8808... 0.0605 sec/batch
Epoch: 1/20... Training Step: 2209... Training loss: 1.8736... 0.0727 sec/batch
Epoch: 1/20... Training Step: 2210... Training loss: 1.8991... 0.0727 sec/batch
Epoch: 1/20... Training Step: 2211... Training loss: 1.9865... 0.0604 sec/batch
Epoch: 1/20... Training Step: 2212... Training loss: 1.7275... 0.0638 sec/batch
Epoch: 1/20... Training Step: 2213... Training loss: 1.8435... 0.0658 sec/batch
Epoch: 1/20... Training Step: 2214... Training loss: 1.9355... 0.0820 sec/batch
Epoch: 1/20... Training Step: 2215... Training loss: 1.8942... 0.0640 sec/batch
Epoch: 1/20... Training Step: 2216... Training loss: 2.2002... 0.0539 sec/batch
Epoch: 1/20... Training Step: 2217... Training loss: 1.9203... 0.0752 sec/batch
Epoch: 1/20... Training Step: 2218... Training loss: 1.7895... 0.0687 sec/batch
Epoch: 1/20... Training Step: 2219... Training loss: 1.8176... 0.0666 sec/batch
Epoch: 1/20... Training Step: 2220... Training loss: 1.9146... 0.0869 sec/batch
Epoch: 1/20... Training Step: 2221... Training loss: 1.8122... 0.0620 sec/batch
Epoch: 1/20... Training Step: 2222... Training loss: 1.8433... 0.0608 sec/batch
Epoch: 1/20... Training Step: 2223... Training loss: 1.7704... 0.0754 sec/batch
Epoch: 1/20... Training Step: 2224... Training loss: 1.7135... 0.0596 sec/batch
Epoch: 1/20... Training Step: 2225... Training loss: 2.0454... 0.0584 sec/batch
Epoch: 1/20... Training Step: 2226... Training loss: 1.7347... 0.0735 sec/batch
Epoch: 1/20... Training Step: 2227... Training loss: 1.7676... 0.0711 sec/batch
Epoch: 1/20... Training Step: 2228... Training loss: 1.8557... 0.0607 sec/batch
Epoch: 1/20... Training Step: 2229... Training loss: 1.8186... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2230... Training loss: 1.8387... 0.0734 sec/batch
Epoch: 1/20... Training Step: 2231... Training loss: 1.8932... 0.0735 sec/batch
Epoch: 1/20... Training Step: 2232... Training loss: 1.7999... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2233... Training loss: 1.8403... 0.0638 sec/batch
Epoch: 1/20... Training Step: 2234... Training loss: 1.8436... 0.0863 sec/batch
Epoch: 1/20... Training Step: 2235... Training loss: 1.7393... 0.0808 sec/batch
Epoch: 1/20... Training Step: 2236... Training loss: 1.6642... 0.0655 sec/batch
Epoch: 1/20... Training Step: 2237... Training loss: 1.8297... 0.0745 sec/batch
Epoch: 1/20... Training Step: 2238... Training loss: 1.7047... 0.0579 sec/batch
Epoch: 1/20... Training Step: 2239... Training loss: 1.9016... 0.0600 sec/batch
Epoch: 1/20... Training Step: 2240... Training loss: 2.0764... 0.0778 sec/batch
Epoch: 1/20... Training Step: 2241... Training loss: 1.9247... 0.0856 sec/batch
Epoch: 1/20... Training Step: 2242... Training loss: 1.8285... 0.0647 sec/batch
Epoch: 1/20... Training Step: 2243... Training loss: 1.9343... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2244... Training loss: 1.8885... 0.0855 sec/batch
Epoch: 1/20... Training Step: 2245... Training loss: 2.1716... 0.0552 sec/batch
Epoch: 1/20... Training Step: 2246... Training loss: 1.9378... 0.0652 sec/batch
Epoch: 1/20... Training Step: 2247... Training loss: 1.7762... 0.0792 sec/batch
Epoch: 1/20... Training Step: 2248... Training loss: 1.8751... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2249... Training loss: 1.7365... 0.0638 sec/batch
Epoch: 1/20... Training Step: 2250... Training loss: 1.9186... 0.0937 sec/batch
Epoch: 1/20... Training Step: 2251... Training loss: 1.7484... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2252... Training loss: 1.8923... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2253... Training loss: 1.6584... 0.0936 sec/batch
Epoch: 1/20... Training Step: 2254... Training loss: 1.7632... 0.0680 sec/batch
Epoch: 1/20... Training Step: 2255... Training loss: 1.7902... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2256... Training loss: 1.7415... 0.0705 sec/batch
Epoch: 1/20... Training Step: 2257... Training loss: 1.8597... 0.0775 sec/batch
Epoch: 1/20... Training Step: 2258... Training loss: 1.7425... 0.0653 sec/batch
Epoch: 1/20... Training Step: 2259... Training loss: 1.7131... 0.0688 sec/batch
Epoch: 1/20... Training Step: 2260... Training loss: 2.0134... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2261... Training loss: 1.9978... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2262... Training loss: 1.6481... 0.0779 sec/batch
Epoch: 1/20... Training Step: 2263... Training loss: 1.7240... 0.0669 sec/batch
Epoch: 1/20... Training Step: 2264... Training loss: 1.8494... 0.0840 sec/batch
Epoch: 1/20... Training Step: 2265... Training loss: 1.7440... 0.0634 sec/batch
Epoch: 1/20... Training Step: 2266... Training loss: 1.7769... 0.0928 sec/batch
Epoch: 1/20... Training Step: 2267... Training loss: 1.7777... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2268... Training loss: 1.8674... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2269... Training loss: 1.8537... 0.0740 sec/batch
Epoch: 1/20... Training Step: 2270... Training loss: 1.7387... 0.0637 sec/batch
Epoch: 1/20... Training Step: 2271... Training loss: 1.8328... 0.0621 sec/batch
Epoch: 1/20... Training Step: 2272... Training loss: 1.9134... 0.0729 sec/batch
Epoch: 1/20... Training Step: 2273... Training loss: 1.7066... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2274... Training loss: 1.6836... 0.0538 sec/batch
Epoch: 1/20... Training Step: 2275... Training loss: 1.6669... 0.0734 sec/batch
Epoch: 1/20... Training Step: 2276... Training loss: 1.7082... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2277... Training loss: 1.7368... 0.0670 sec/batch
Epoch: 1/20... Training Step: 2278... Training loss: 1.8657... 0.0532 sec/batch
Epoch: 1/20... Training Step: 2279... Training loss: 1.8920... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2280... Training loss: 1.8043... 0.0870 sec/batch
Epoch: 1/20... Training Step: 2281... Training loss: 1.7581... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2282... Training loss: 1.7977... 0.0594 sec/batch
Epoch: 1/20... Training Step: 2283... Training loss: 1.8363... 0.0804 sec/batch
Epoch: 1/20... Training Step: 2284... Training loss: 1.6891... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2285... Training loss: 1.7492... 0.0620 sec/batch
Epoch: 1/20... Training Step: 2286... Training loss: 1.7255... 0.0946 sec/batch
Epoch: 1/20... Training Step: 2287... Training loss: 1.7202... 0.0594 sec/batch
Epoch: 1/20... Training Step: 2288... Training loss: 1.7721... 0.0597 sec/batch
Epoch: 1/20... Training Step: 2289... Training loss: 1.8119... 0.0932 sec/batch
Epoch: 1/20... Training Step: 2290... Training loss: 1.8399... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2291... Training loss: 1.7023... 0.0566 sec/batch
Epoch: 1/20... Training Step: 2292... Training loss: 1.8417... 0.0620 sec/batch
Epoch: 1/20... Training Step: 2293... Training loss: 1.7325... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2294... Training loss: 1.7741... 0.0603 sec/batch
Epoch: 1/20... Training Step: 2295... Training loss: 1.8354... 0.0774 sec/batch
Epoch: 1/20... Training Step: 2296... Training loss: 1.7266... 0.0610 sec/batch
Epoch: 1/20... Training Step: 2297... Training loss: 1.7934... 0.0600 sec/batch
Epoch: 1/20... Training Step: 2298... Training loss: 1.6954... 0.0631 sec/batch
Epoch: 1/20... Training Step: 2299... Training loss: 1.9191... 0.0701 sec/batch
Epoch: 1/20... Training Step: 2300... Training loss: 1.6938... 0.0695 sec/batch
Epoch: 1/20... Training Step: 2301... Training loss: 1.9097... 0.0558 sec/batch
Epoch: 1/20... Training Step: 2302... Training loss: 1.9006... 0.0614 sec/batch
Epoch: 1/20... Training Step: 2303... Training loss: 1.7715... 0.0772 sec/batch
Epoch: 1/20... Training Step: 2304... Training loss: 1.8168... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2305... Training loss: 1.8352... 0.0554 sec/batch
Epoch: 1/20... Training Step: 2306... Training loss: 1.9379... 0.0599 sec/batch
Epoch: 1/20... Training Step: 2307... Training loss: 1.8408... 0.0627 sec/batch
Epoch: 1/20... Training Step: 2308... Training loss: 1.9460... 0.0658 sec/batch
Epoch: 1/20... Training Step: 2309... Training loss: 1.7827... 0.0624 sec/batch
Epoch: 1/20... Training Step: 2310... Training loss: 1.8506... 0.0599 sec/batch
Epoch: 1/20... Training Step: 2311... Training loss: 1.7729... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2312... Training loss: 1.7362... 0.0716 sec/batch
Epoch: 1/20... Training Step: 2313... Training loss: 1.8932... 0.0541 sec/batch
Epoch: 1/20... Training Step: 2314... Training loss: 1.8667... 0.0618 sec/batch
Epoch: 1/20... Training Step: 2315... Training loss: 1.7829... 0.0824 sec/batch
Epoch: 1/20... Training Step: 2316... Training loss: 1.7830... 0.0768 sec/batch
Epoch: 1/20... Training Step: 2317... Training loss: 1.6694... 0.0596 sec/batch
Epoch: 1/20... Training Step: 2318... Training loss: 1.8430... 0.0660 sec/batch
Epoch: 1/20... Training Step: 2319... Training loss: 1.8121... 0.0635 sec/batch
Epoch: 1/20... Training Step: 2320... Training loss: 1.7895... 0.0700 sec/batch
Epoch: 1/20... Training Step: 2321... Training loss: 1.7652... 0.0516 sec/batch
Epoch: 1/20... Training Step: 2322... Training loss: 1.6807... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2323... Training loss: 1.7306... 0.0720 sec/batch
Epoch: 1/20... Training Step: 2324... Training loss: 1.9413... 0.0712 sec/batch
Epoch: 1/20... Training Step: 2325... Training loss: 1.9946... 0.0541 sec/batch
Epoch: 1/20... Training Step: 2326... Training loss: 1.8930... 0.0653 sec/batch
Epoch: 1/20... Training Step: 2327... Training loss: 1.8723... 0.0653 sec/batch
Epoch: 1/20... Training Step: 2328... Training loss: 1.8268... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2329... Training loss: 1.9380... 0.0612 sec/batch
Epoch: 1/20... Training Step: 2330... Training loss: 1.7476... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2331... Training loss: 1.6811... 0.0635 sec/batch
Epoch: 1/20... Training Step: 2332... Training loss: 1.9367... 0.0651 sec/batch
Epoch: 1/20... Training Step: 2333... Training loss: 1.7042... 0.0589 sec/batch
Epoch: 1/20... Training Step: 2334... Training loss: 1.7223... 0.0959 sec/batch
Epoch: 1/20... Training Step: 2335... Training loss: 1.7390... 0.0715 sec/batch
Epoch: 1/20... Training Step: 2336... Training loss: 1.8478... 0.0639 sec/batch
Epoch: 1/20... Training Step: 2337... Training loss: 2.0163... 0.0660 sec/batch
Epoch: 1/20... Training Step: 2338... Training loss: 2.0439... 0.0759 sec/batch
Epoch: 1/20... Training Step: 2339... Training loss: 1.9058... 0.0712 sec/batch
Epoch: 1/20... Training Step: 2340... Training loss: 1.9516... 0.0561 sec/batch
Epoch: 1/20... Training Step: 2341... Training loss: 1.8012... 0.0642 sec/batch
Epoch: 1/20... Training Step: 2342... Training loss: 1.8176... 0.0972 sec/batch
Epoch: 1/20... Training Step: 2343... Training loss: 1.7731... 0.0791 sec/batch
Epoch: 1/20... Training Step: 2344... Training loss: 1.7731... 0.0651 sec/batch
Epoch: 1/20... Training Step: 2345... Training loss: 1.7901... 0.0637 sec/batch
Epoch: 1/20... Training Step: 2346... Training loss: 1.6408... 0.0672 sec/batch
Epoch: 1/20... Training Step: 2347... Training loss: 1.8318... 0.0680 sec/batch
Epoch: 1/20... Training Step: 2348... Training loss: 1.7455... 0.0591 sec/batch
Epoch: 1/20... Training Step: 2349... Training loss: 1.8651... 0.0779 sec/batch
Epoch: 1/20... Training Step: 2350... Training loss: 1.9115... 0.0760 sec/batch
Epoch: 1/20... Training Step: 2351... Training loss: 1.9539... 0.0624 sec/batch
Epoch: 1/20... Training Step: 2352... Training loss: 1.9362... 0.0660 sec/batch
Epoch: 1/20... Training Step: 2353... Training loss: 1.9445... 0.0893 sec/batch
Epoch: 1/20... Training Step: 2354... Training loss: 1.9282... 0.0687 sec/batch
Epoch: 1/20... Training Step: 2355... Training loss: 1.8014... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2356... Training loss: 1.8142... 0.0722 sec/batch
Epoch: 1/20... Training Step: 2357... Training loss: 1.7585... 0.0724 sec/batch
Epoch: 1/20... Training Step: 2358... Training loss: 1.7910... 0.0660 sec/batch
Epoch: 1/20... Training Step: 2359... Training loss: 1.7830... 0.0996 sec/batch
Epoch: 1/20... Training Step: 2360... Training loss: 1.9607... 0.0695 sec/batch
Epoch: 1/20... Training Step: 2361... Training loss: 1.8092... 0.0709 sec/batch
Epoch: 1/20... Training Step: 2362... Training loss: 1.9352... 0.0736 sec/batch
Epoch: 1/20... Training Step: 2363... Training loss: 1.8251... 0.0690 sec/batch
Epoch: 1/20... Training Step: 2364... Training loss: 1.9751... 0.0588 sec/batch
Epoch: 1/20... Training Step: 2365... Training loss: 1.7127... 0.0738 sec/batch
Epoch: 1/20... Training Step: 2366... Training loss: 1.7626... 0.0793 sec/batch
Epoch: 1/20... Training Step: 2367... Training loss: 1.8406... 0.0566 sec/batch
Epoch: 1/20... Training Step: 2368... Training loss: 1.7910... 0.0799 sec/batch
Epoch: 1/20... Training Step: 2369... Training loss: 1.8674... 0.0721 sec/batch
Epoch: 1/20... Training Step: 2370... Training loss: 1.7265... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2371... Training loss: 1.9143... 0.0834 sec/batch
Epoch: 1/20... Training Step: 2372... Training loss: 1.8103... 0.0671 sec/batch
Epoch: 1/20... Training Step: 2373... Training loss: 1.7311... 0.0651 sec/batch
Epoch: 1/20... Training Step: 2374... Training loss: 1.8205... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2375... Training loss: 1.8241... 0.0704 sec/batch
Epoch: 1/20... Training Step: 2376... Training loss: 1.9231... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2377... Training loss: 2.0093... 0.0602 sec/batch
Epoch: 1/20... Training Step: 2378... Training loss: 1.8885... 0.0739 sec/batch
Epoch: 1/20... Training Step: 2379... Training loss: 1.7994... 0.0542 sec/batch
Epoch: 1/20... Training Step: 2380... Training loss: 1.8931... 0.0626 sec/batch
Epoch: 1/20... Training Step: 2381... Training loss: 1.8130... 0.0734 sec/batch
Epoch: 1/20... Training Step: 2382... Training loss: 1.9581... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2383... Training loss: 1.9187... 0.0502 sec/batch
Epoch: 1/20... Training Step: 2384... Training loss: 1.8546... 0.0565 sec/batch
Epoch: 1/20... Training Step: 2385... Training loss: 1.9040... 0.0817 sec/batch
Epoch: 1/20... Training Step: 2386... Training loss: 1.7629... 0.0533 sec/batch
Epoch: 1/20... Training Step: 2387... Training loss: 1.6981... 0.0585 sec/batch
Epoch: 1/20... Training Step: 2388... Training loss: 1.9144... 0.0930 sec/batch
Epoch: 1/20... Training Step: 2389... Training loss: 1.8862... 0.0702 sec/batch
Epoch: 1/20... Training Step: 2390... Training loss: 1.8654... 0.0539 sec/batch
Epoch: 1/20... Training Step: 2391... Training loss: 1.9185... 0.0572 sec/batch
Epoch: 1/20... Training Step: 2392... Training loss: 1.8410... 0.0669 sec/batch
Epoch: 1/20... Training Step: 2393... Training loss: 2.0210... 0.0725 sec/batch
Epoch: 1/20... Training Step: 2394... Training loss: 1.9933... 0.0629 sec/batch
Epoch: 1/20... Training Step: 2395... Training loss: 1.9016... 0.0569 sec/batch
Epoch: 1/20... Training Step: 2396... Training loss: 1.8357... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2397... Training loss: 1.8744... 0.0644 sec/batch
Epoch: 1/20... Training Step: 2398... Training loss: 1.7626... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2399... Training loss: 1.8771... 0.0547 sec/batch
Epoch: 1/20... Training Step: 2400... Training loss: 1.9155... 0.0737 sec/batch
Epoch: 1/20... Training Step: 2401... Training loss: 1.8135... 0.0500 sec/batch
Epoch: 1/20... Training Step: 2402... Training loss: 1.7393... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2403... Training loss: 1.7670... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2404... Training loss: 1.8539... 0.0713 sec/batch
Epoch: 1/20... Training Step: 2405... Training loss: 1.7363... 0.0509 sec/batch
Epoch: 1/20... Training Step: 2406... Training loss: 1.8605... 0.0535 sec/batch
Epoch: 1/20... Training Step: 2407... Training loss: 1.7792... 0.0517 sec/batch
Epoch: 1/20... Training Step: 2408... Training loss: 1.8020... 0.0699 sec/batch
Epoch: 1/20... Training Step: 2409... Training loss: 1.7028... 0.0564 sec/batch
Epoch: 1/20... Training Step: 2410... Training loss: 1.8096... 0.0546 sec/batch
Epoch: 1/20... Training Step: 2411... Training loss: 1.8021... 0.0692 sec/batch
Epoch: 1/20... Training Step: 2412... Training loss: 1.8129... 0.0849 sec/batch
Epoch: 1/20... Training Step: 2413... Training loss: 1.8997... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2414... Training loss: 1.6729... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2415... Training loss: 1.7506... 0.0900 sec/batch
Epoch: 1/20... Training Step: 2416... Training loss: 1.7117... 0.0572 sec/batch
Epoch: 1/20... Training Step: 2417... Training loss: 1.6999... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2418... Training loss: 1.7550... 0.0758 sec/batch
Epoch: 1/20... Training Step: 2419... Training loss: 1.8543... 0.0722 sec/batch
Epoch: 1/20... Training Step: 2420... Training loss: 1.9533... 0.0566 sec/batch
Epoch: 1/20... Training Step: 2421... Training loss: 1.9434... 0.0687 sec/batch
Epoch: 1/20... Training Step: 2422... Training loss: 1.8618... 0.0651 sec/batch
Epoch: 1/20... Training Step: 2423... Training loss: 1.8233... 0.0537 sec/batch
Epoch: 1/20... Training Step: 2424... Training loss: 1.8026... 0.0647 sec/batch
Epoch: 1/20... Training Step: 2425... Training loss: 1.7541... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2426... Training loss: 1.8247... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2427... Training loss: 1.8016... 0.0558 sec/batch
Epoch: 1/20... Training Step: 2428... Training loss: 1.8881... 0.0690 sec/batch
Epoch: 1/20... Training Step: 2429... Training loss: 1.8740... 0.0685 sec/batch
Epoch: 1/20... Training Step: 2430... Training loss: 1.7982... 0.0644 sec/batch
Epoch: 1/20... Training Step: 2431... Training loss: 1.8352... 0.0608 sec/batch
Epoch: 1/20... Training Step: 2432... Training loss: 1.9319... 0.0860 sec/batch
Epoch: 1/20... Training Step: 2433... Training loss: 1.9408... 0.0698 sec/batch
Epoch: 1/20... Training Step: 2434... Training loss: 1.8433... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2435... Training loss: 1.6914... 0.0614 sec/batch
Epoch: 1/20... Training Step: 2436... Training loss: 1.8365... 0.0926 sec/batch
Epoch: 1/20... Training Step: 2437... Training loss: 1.7889... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2438... Training loss: 1.8662... 0.0697 sec/batch
Epoch: 1/20... Training Step: 2439... Training loss: 1.8755... 0.0602 sec/batch
Epoch: 1/20... Training Step: 2440... Training loss: 1.8349... 0.1008 sec/batch
Epoch: 1/20... Training Step: 2441... Training loss: 1.7438... 0.0647 sec/batch
Epoch: 1/20... Training Step: 2442... Training loss: 1.7946... 0.0561 sec/batch
Epoch: 1/20... Training Step: 2443... Training loss: 1.7918... 0.0636 sec/batch
Epoch: 1/20... Training Step: 2444... Training loss: 1.9490... 0.0790 sec/batch
Epoch: 1/20... Training Step: 2445... Training loss: 1.8118... 0.0701 sec/batch
Epoch: 1/20... Training Step: 2446... Training loss: 1.7741... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2447... Training loss: 1.8659... 0.0643 sec/batch
Epoch: 1/20... Training Step: 2448... Training loss: 1.7694... 0.0695 sec/batch
Epoch: 1/20... Training Step: 2449... Training loss: 1.8058... 0.0711 sec/batch
Epoch: 1/20... Training Step: 2450... Training loss: 2.0083... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2451... Training loss: 1.9616... 0.0669 sec/batch
Epoch: 1/20... Training Step: 2452... Training loss: 1.8487... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2453... Training loss: 1.9050... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2454... Training loss: 1.7707... 0.0769 sec/batch
Epoch: 1/20... Training Step: 2455... Training loss: 1.8204... 0.0636 sec/batch
Epoch: 1/20... Training Step: 2456... Training loss: 1.8834... 0.0692 sec/batch
Epoch: 1/20... Training Step: 2457... Training loss: 1.9662... 0.0591 sec/batch
Epoch: 1/20... Training Step: 2458... Training loss: 1.9024... 0.0837 sec/batch
Epoch: 1/20... Training Step: 2459... Training loss: 2.0154... 0.0721 sec/batch
Epoch: 1/20... Training Step: 2460... Training loss: 1.8795... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2461... Training loss: 1.7215... 0.0594 sec/batch
Epoch: 1/20... Training Step: 2462... Training loss: 1.8675... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2463... Training loss: 1.7307... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2464... Training loss: 1.6580... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2465... Training loss: 1.9004... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2466... Training loss: 1.8197... 0.0871 sec/batch
Epoch: 1/20... Training Step: 2467... Training loss: 1.8598... 0.0750 sec/batch
Epoch: 1/20... Training Step: 2468... Training loss: 1.8103... 0.0582 sec/batch
Epoch: 1/20... Training Step: 2469... Training loss: 1.8083... 0.0656 sec/batch
Epoch: 1/20... Training Step: 2470... Training loss: 1.8798... 0.0683 sec/batch
Epoch: 1/20... Training Step: 2471... Training loss: 1.7449... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2472... Training loss: 1.8378... 0.0654 sec/batch
Epoch: 1/20... Training Step: 2473... Training loss: 1.6758... 0.0579 sec/batch
Epoch: 1/20... Training Step: 2474... Training loss: 1.7593... 0.0756 sec/batch
Epoch: 1/20... Training Step: 2475... Training loss: 1.7069... 0.0601 sec/batch
Epoch: 1/20... Training Step: 2476... Training loss: 1.7675... 0.0751 sec/batch
Epoch: 1/20... Training Step: 2477... Training loss: 1.7224... 0.0867 sec/batch
Epoch: 1/20... Training Step: 2478... Training loss: 1.8221... 0.0524 sec/batch
Epoch: 1/20... Training Step: 2479... Training loss: 1.8038... 0.0552 sec/batch
Epoch: 1/20... Training Step: 2480... Training loss: 1.8581... 0.0666 sec/batch
Epoch: 1/20... Training Step: 2481... Training loss: 1.7353... 0.0915 sec/batch
Epoch: 1/20... Training Step: 2482... Training loss: 1.8214... 0.0688 sec/batch
Epoch: 1/20... Training Step: 2483... Training loss: 1.9925... 0.0634 sec/batch
Epoch: 1/20... Training Step: 2484... Training loss: 1.7988... 0.0915 sec/batch
Epoch: 1/20... Training Step: 2485... Training loss: 1.7783... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2486... Training loss: 1.6982... 0.0730 sec/batch
Epoch: 1/20... Training Step: 2487... Training loss: 1.7642... 0.0808 sec/batch
Epoch: 1/20... Training Step: 2488... Training loss: 1.7206... 0.0666 sec/batch
Epoch: 1/20... Training Step: 2489... Training loss: 1.9461... 0.0696 sec/batch
Epoch: 1/20... Training Step: 2490... Training loss: 1.8695... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2491... Training loss: 1.8844... 0.0748 sec/batch
Epoch: 1/20... Training Step: 2492... Training loss: 1.8906... 0.0635 sec/batch
Epoch: 1/20... Training Step: 2493... Training loss: 1.8295... 0.0558 sec/batch
Epoch: 1/20... Training Step: 2494... Training loss: 1.8165... 0.0629 sec/batch
Epoch: 1/20... Training Step: 2495... Training loss: 1.8330... 0.0633 sec/batch
Epoch: 1/20... Training Step: 2496... Training loss: 1.7786... 0.0677 sec/batch
Epoch: 1/20... Training Step: 2497... Training loss: 1.8146... 0.0774 sec/batch
Epoch: 1/20... Training Step: 2498... Training loss: 1.8325... 0.0552 sec/batch
Epoch: 1/20... Training Step: 2499... Training loss: 1.8795... 0.0480 sec/batch
Epoch: 1/20... Training Step: 2500... Training loss: 1.7497... 0.0685 sec/batch
Epoch: 1/20... Training Step: 2501... Training loss: 1.7549... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2502... Training loss: 1.7824... 0.0524 sec/batch
Epoch: 1/20... Training Step: 2503... Training loss: 1.7541... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2504... Training loss: 1.7906... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2505... Training loss: 1.9208... 0.0723 sec/batch
Epoch: 1/20... Training Step: 2506... Training loss: 1.8522... 0.0639 sec/batch
Epoch: 1/20... Training Step: 2507... Training loss: 1.7358... 0.0571 sec/batch
Epoch: 1/20... Training Step: 2508... Training loss: 1.8820... 0.0632 sec/batch
Epoch: 1/20... Training Step: 2509... Training loss: 1.8971... 0.0708 sec/batch
Epoch: 1/20... Training Step: 2510... Training loss: 1.9292... 0.0575 sec/batch
Epoch: 1/20... Training Step: 2511... Training loss: 1.8110... 0.0653 sec/batch
Epoch: 1/20... Training Step: 2512... Training loss: 1.7047... 0.0761 sec/batch
Epoch: 1/20... Training Step: 2513... Training loss: 1.7731... 0.0589 sec/batch
Epoch: 1/20... Training Step: 2514... Training loss: 1.8447... 0.0571 sec/batch
Epoch: 1/20... Training Step: 2515... Training loss: 1.7282... 0.0637 sec/batch
Epoch: 1/20... Training Step: 2516... Training loss: 1.8367... 0.0602 sec/batch
Epoch: 1/20... Training Step: 2517... Training loss: 1.8060... 0.0977 sec/batch
Epoch: 1/20... Training Step: 2518... Training loss: 1.8648... 0.0501 sec/batch
Epoch: 1/20... Training Step: 2519... Training loss: 1.8842... 0.0560 sec/batch
Epoch: 1/20... Training Step: 2520... Training loss: 1.8839... 0.0861 sec/batch
Epoch: 1/20... Training Step: 2521... Training loss: 1.7991... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2522... Training loss: 1.7670... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2523... Training loss: 1.7915... 0.0695 sec/batch
Epoch: 1/20... Training Step: 2524... Training loss: 1.9165... 0.0694 sec/batch
Epoch: 1/20... Training Step: 2525... Training loss: 1.8923... 0.0610 sec/batch
Epoch: 1/20... Training Step: 2526... Training loss: 1.7425... 0.0813 sec/batch
Epoch: 1/20... Training Step: 2527... Training loss: 1.9810... 0.0715 sec/batch
Epoch: 1/20... Training Step: 2528... Training loss: 1.7649... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2529... Training loss: 1.9003... 0.0634 sec/batch
Epoch: 1/20... Training Step: 2530... Training loss: 1.7847... 0.0685 sec/batch
Epoch: 1/20... Training Step: 2531... Training loss: 1.7652... 0.0750 sec/batch
Epoch: 1/20... Training Step: 2532... Training loss: 1.7325... 0.0631 sec/batch
Epoch: 1/20... Training Step: 2533... Training loss: 1.8932... 0.0916 sec/batch
Epoch: 1/20... Training Step: 2534... Training loss: 1.6783... 0.0704 sec/batch
Epoch: 1/20... Training Step: 2535... Training loss: 1.8227... 0.0640 sec/batch
Epoch: 1/20... Training Step: 2536... Training loss: 1.8474... 0.0749 sec/batch
Epoch: 1/20... Training Step: 2537... Training loss: 1.8043... 0.0614 sec/batch
Epoch: 1/20... Training Step: 2538... Training loss: 1.9469... 0.0606 sec/batch
Epoch: 1/20... Training Step: 2539... Training loss: 1.8828... 0.0761 sec/batch
Epoch: 1/20... Training Step: 2540... Training loss: 1.8582... 0.0825 sec/batch
Epoch: 1/20... Training Step: 2541... Training loss: 1.8328... 0.0580 sec/batch
Epoch: 1/20... Training Step: 2542... Training loss: 1.8989... 0.0767 sec/batch
Epoch: 1/20... Training Step: 2543... Training loss: 1.6991... 0.0778 sec/batch
Epoch: 1/20... Training Step: 2544... Training loss: 1.9163... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2545... Training loss: 1.7451... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2546... Training loss: 1.7849... 0.0598 sec/batch
Epoch: 1/20... Training Step: 2547... Training loss: 1.8979... 0.0626 sec/batch
Epoch: 1/20... Training Step: 2548... Training loss: 1.8039... 0.0684 sec/batch
Epoch: 1/20... Training Step: 2549... Training loss: 1.7962... 0.0900 sec/batch
Epoch: 1/20... Training Step: 2550... Training loss: 1.8357... 0.0808 sec/batch
Epoch: 1/20... Training Step: 2551... Training loss: 1.7006... 0.0522 sec/batch
Epoch: 1/20... Training Step: 2552... Training loss: 1.9572... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2553... Training loss: 1.6913... 0.0803 sec/batch
Epoch: 1/20... Training Step: 2554... Training loss: 1.7000... 0.0720 sec/batch
Epoch: 1/20... Training Step: 2555... Training loss: 1.8265... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2556... Training loss: 1.8434... 0.0708 sec/batch
Epoch: 1/20... Training Step: 2557... Training loss: 1.8260... 0.0906 sec/batch
Epoch: 1/20... Training Step: 2558... Training loss: 1.7913... 0.0711 sec/batch
Epoch: 1/20... Training Step: 2559... Training loss: 1.8787... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2560... Training loss: 1.7025... 0.0622 sec/batch
Epoch: 1/20... Training Step: 2561... Training loss: 1.6217... 0.0782 sec/batch
Epoch: 1/20... Training Step: 2562... Training loss: 1.7630... 0.0651 sec/batch
Epoch: 1/20... Training Step: 2563... Training loss: 1.8176... 0.0759 sec/batch
Epoch: 1/20... Training Step: 2564... Training loss: 1.6503... 0.0773 sec/batch
Epoch: 1/20... Training Step: 2565... Training loss: 1.6036... 0.0641 sec/batch
Epoch: 1/20... Training Step: 2566... Training loss: 1.8554... 0.0753 sec/batch
Epoch: 1/20... Training Step: 2567... Training loss: 1.7612... 0.0702 sec/batch
Epoch: 1/20... Training Step: 2568... Training loss: 1.8777... 0.0574 sec/batch
Epoch: 1/20... Training Step: 2569... Training loss: 1.6898... 0.0835 sec/batch
Epoch: 1/20... Training Step: 2570... Training loss: 1.7622... 0.0742 sec/batch
Epoch: 1/20... Training Step: 2571... Training loss: 1.7555... 0.0633 sec/batch
Epoch: 1/20... Training Step: 2572... Training loss: 1.8479... 0.0721 sec/batch
Epoch: 1/20... Training Step: 2573... Training loss: 1.8279... 0.0666 sec/batch
Epoch: 1/20... Training Step: 2574... Training loss: 1.8015... 0.0670 sec/batch
Epoch: 1/20... Training Step: 2575... Training loss: 1.7766... 0.0823 sec/batch
Epoch: 1/20... Training Step: 2576... Training loss: 1.7669... 0.0699 sec/batch
Epoch: 1/20... Training Step: 2577... Training loss: 1.9080... 0.0770 sec/batch
Epoch: 1/20... Training Step: 2578... Training loss: 1.8976... 0.0867 sec/batch
Epoch: 1/20... Training Step: 2579... Training loss: 1.7912... 0.0740 sec/batch
Epoch: 1/20... Training Step: 2580... Training loss: 1.8859... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2581... Training loss: 1.7509... 0.0789 sec/batch
Epoch: 1/20... Training Step: 2582... Training loss: 1.9750... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2583... Training loss: 1.8004... 0.0680 sec/batch
Epoch: 1/20... Training Step: 2584... Training loss: 1.7931... 0.0796 sec/batch
Epoch: 1/20... Training Step: 2585... Training loss: 1.8897... 0.0590 sec/batch
Epoch: 1/20... Training Step: 2586... Training loss: 2.0139... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2587... Training loss: 1.8380... 0.1084 sec/batch
Epoch: 1/20... Training Step: 2588... Training loss: 2.0243... 0.0645 sec/batch
Epoch: 1/20... Training Step: 2589... Training loss: 1.8610... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2590... Training loss: 1.9099... 0.0608 sec/batch
Epoch: 1/20... Training Step: 2591... Training loss: 1.8456... 0.0646 sec/batch
Epoch: 1/20... Training Step: 2592... Training loss: 1.8821... 0.0642 sec/batch
Epoch: 1/20... Training Step: 2593... Training loss: 1.9241... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2594... Training loss: 1.8482... 0.0646 sec/batch
Epoch: 1/20... Training Step: 2595... Training loss: 1.7751... 0.0923 sec/batch
Epoch: 1/20... Training Step: 2596... Training loss: 1.7922... 0.0803 sec/batch
Epoch: 1/20... Training Step: 2597... Training loss: 1.6625... 0.0601 sec/batch
Epoch: 1/20... Training Step: 2598... Training loss: 1.7759... 0.0699 sec/batch
Epoch: 1/20... Training Step: 2599... Training loss: 1.7662... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2600... Training loss: 1.7520... 0.0710 sec/batch
Epoch: 1/20... Training Step: 2601... Training loss: 1.7238... 0.0858 sec/batch
Epoch: 1/20... Training Step: 2602... Training loss: 1.7216... 0.0639 sec/batch
Epoch: 1/20... Training Step: 2603... Training loss: 1.8887... 0.0713 sec/batch
Epoch: 1/20... Training Step: 2604... Training loss: 1.8281... 0.0658 sec/batch
Epoch: 1/20... Training Step: 2605... Training loss: 1.7689... 0.1067 sec/batch
Epoch: 1/20... Training Step: 2606... Training loss: 1.7555... 0.0673 sec/batch
Epoch: 1/20... Training Step: 2607... Training loss: 1.8777... 0.0707 sec/batch
Epoch: 1/20... Training Step: 2608... Training loss: 1.9067... 0.0654 sec/batch
Epoch: 1/20... Training Step: 2609... Training loss: 1.6799... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2610... Training loss: 1.9131... 0.0532 sec/batch
Epoch: 1/20... Training Step: 2611... Training loss: 1.9230... 0.0709 sec/batch
Epoch: 1/20... Training Step: 2612... Training loss: 1.6908... 0.0629 sec/batch
Epoch: 1/20... Training Step: 2613... Training loss: 1.8513... 0.0622 sec/batch
Epoch: 1/20... Training Step: 2614... Training loss: 1.7922... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2615... Training loss: 1.5742... 0.0633 sec/batch
Epoch: 1/20... Training Step: 2616... Training loss: 1.7330... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2617... Training loss: 1.8623... 0.0676 sec/batch
Epoch: 1/20... Training Step: 2618... Training loss: 1.9186... 0.0601 sec/batch
Epoch: 1/20... Training Step: 2619... Training loss: 1.6991... 0.0863 sec/batch
Epoch: 1/20... Training Step: 2620... Training loss: 1.6694... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2621... Training loss: 1.9365... 0.0564 sec/batch
Epoch: 1/20... Training Step: 2622... Training loss: 1.6453... 0.0697 sec/batch
Epoch: 1/20... Training Step: 2623... Training loss: 1.7785... 0.0774 sec/batch
Epoch: 1/20... Training Step: 2624... Training loss: 1.8205... 0.0733 sec/batch
Epoch: 1/20... Training Step: 2625... Training loss: 1.8489... 0.0522 sec/batch
Epoch: 1/20... Training Step: 2626... Training loss: 1.7984... 0.0605 sec/batch
Epoch: 1/20... Training Step: 2627... Training loss: 1.8011... 0.0705 sec/batch
Epoch: 1/20... Training Step: 2628... Training loss: 1.8360... 0.0794 sec/batch
Epoch: 1/20... Training Step: 2629... Training loss: 1.6843... 0.0581 sec/batch
Epoch: 1/20... Training Step: 2630... Training loss: 2.0731... 0.0510 sec/batch
Epoch: 1/20... Training Step: 2631... Training loss: 1.9036... 0.0648 sec/batch
Epoch: 1/20... Training Step: 2632... Training loss: 1.7893... 0.0788 sec/batch
Epoch: 1/20... Training Step: 2633... Training loss: 1.8380... 0.0627 sec/batch
Epoch: 1/20... Training Step: 2634... Training loss: 1.8554... 0.0604 sec/batch
Epoch: 1/20... Training Step: 2635... Training loss: 1.7232... 0.0840 sec/batch
Epoch: 1/20... Training Step: 2636... Training loss: 1.8977... 0.0770 sec/batch
Epoch: 1/20... Training Step: 2637... Training loss: 1.8640... 0.0550 sec/batch
Epoch: 1/20... Training Step: 2638... Training loss: 1.7356... 0.0641 sec/batch
Epoch: 1/20... Training Step: 2639... Training loss: 1.8175... 0.0632 sec/batch
Epoch: 1/20... Training Step: 2640... Training loss: 1.7694... 0.0601 sec/batch
Epoch: 1/20... Training Step: 2641... Training loss: 1.9226... 0.0571 sec/batch
Epoch: 1/20... Training Step: 2642... Training loss: 1.7753... 0.0610 sec/batch
Epoch: 1/20... Training Step: 2643... Training loss: 1.7370... 0.0544 sec/batch
Epoch: 1/20... Training Step: 2644... Training loss: 1.8220... 0.0535 sec/batch
Epoch: 1/20... Training Step: 2645... Training loss: 1.8904... 0.0512 sec/batch
Epoch: 1/20... Training Step: 2646... Training loss: 1.8395... 0.0574 sec/batch
Epoch: 1/20... Training Step: 2647... Training loss: 1.8381... 0.0828 sec/batch
Epoch: 1/20... Training Step: 2648... Training loss: 1.8550... 0.0592 sec/batch
Epoch: 1/20... Training Step: 2649... Training loss: 1.7899... 0.0870 sec/batch
Epoch: 1/20... Training Step: 2650... Training loss: 1.9679... 0.0692 sec/batch
Epoch: 1/20... Training Step: 2651... Training loss: 1.8611... 0.0553 sec/batch
Epoch: 1/20... Training Step: 2652... Training loss: 1.8700... 0.0532 sec/batch
Epoch: 1/20... Training Step: 2653... Training loss: 1.9037... 0.0729 sec/batch
Epoch: 1/20... Training Step: 2654... Training loss: 1.9013... 0.0708 sec/batch
Epoch: 1/20... Training Step: 2655... Training loss: 1.7798... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2656... Training loss: 1.8746... 0.0573 sec/batch
Epoch: 1/20... Training Step: 2657... Training loss: 1.9422... 0.0831 sec/batch
Epoch: 1/20... Training Step: 2658... Training loss: 1.7757... 0.0772 sec/batch
Epoch: 1/20... Training Step: 2659... Training loss: 1.7671... 0.0524 sec/batch
Epoch: 1/20... Training Step: 2660... Training loss: 1.9535... 0.0609 sec/batch
Epoch: 1/20... Training Step: 2661... Training loss: 1.7173... 0.0519 sec/batch
Epoch: 1/20... Training Step: 2662... Training loss: 1.8387... 0.0740 sec/batch
Epoch: 1/20... Training Step: 2663... Training loss: 1.7523... 0.0569 sec/batch
Epoch: 1/20... Training Step: 2664... Training loss: 1.7985... 0.0649 sec/batch
Epoch: 1/20... Training Step: 2665... Training loss: 2.0284... 0.0787 sec/batch
Epoch: 1/20... Training Step: 2666... Training loss: 1.8137... 0.0799 sec/batch
Epoch: 1/20... Training Step: 2667... Training loss: 1.7790... 0.0597 sec/batch
Epoch: 1/20... Training Step: 2668... Training loss: 1.6728... 0.0639 sec/batch
Epoch: 1/20... Training Step: 2669... Training loss: 1.8590... 0.0860 sec/batch
Epoch: 1/20... Training Step: 2670... Training loss: 1.7752... 0.0551 sec/batch
Epoch: 1/20... Training Step: 2671... Training loss: 2.0870... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2672... Training loss: 1.8197... 0.0861 sec/batch
Epoch: 1/20... Training Step: 2673... Training loss: 1.9585... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2674... Training loss: 1.9075... 0.0645 sec/batch
Epoch: 1/20... Training Step: 2675... Training loss: 1.7607... 0.0757 sec/batch
Epoch: 1/20... Training Step: 2676... Training loss: 1.7918... 0.0692 sec/batch
Epoch: 1/20... Training Step: 2677... Training loss: 1.7328... 0.0664 sec/batch
Epoch: 1/20... Training Step: 2678... Training loss: 1.8302... 0.0646 sec/batch
Epoch: 1/20... Training Step: 2679... Training loss: 1.9349... 0.0732 sec/batch
Epoch: 1/20... Training Step: 2680... Training loss: 1.6732... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2681... Training loss: 1.8714... 0.0994 sec/batch
Epoch: 1/20... Training Step: 2682... Training loss: 1.7974... 0.0735 sec/batch
Epoch: 1/20... Training Step: 2683... Training loss: 1.8938... 0.0811 sec/batch
Epoch: 1/20... Training Step: 2684... Training loss: 1.8194... 0.0797 sec/batch
Epoch: 1/20... Training Step: 2685... Training loss: 1.7520... 0.0707 sec/batch
Epoch: 1/20... Training Step: 2686... Training loss: 1.7202... 0.0677 sec/batch
Epoch: 1/20... Training Step: 2687... Training loss: 1.7129... 0.0847 sec/batch
Epoch: 1/20... Training Step: 2688... Training loss: 1.7458... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2689... Training loss: 1.7707... 0.0600 sec/batch
Epoch: 1/20... Training Step: 2690... Training loss: 1.9015... 0.0767 sec/batch
Epoch: 1/20... Training Step: 2691... Training loss: 1.9005... 0.0671 sec/batch
Epoch: 1/20... Training Step: 2692... Training loss: 1.7918... 0.0594 sec/batch
Epoch: 1/20... Training Step: 2693... Training loss: 1.6773... 0.0918 sec/batch
Epoch: 1/20... Training Step: 2694... Training loss: 1.8988... 0.0876 sec/batch
Epoch: 1/20... Training Step: 2695... Training loss: 2.0181... 0.0770 sec/batch
Epoch: 1/20... Training Step: 2696... Training loss: 1.8570... 0.0718 sec/batch
Epoch: 1/20... Training Step: 2697... Training loss: 1.7626... 0.0697 sec/batch
Epoch: 1/20... Training Step: 2698... Training loss: 1.8141... 0.0557 sec/batch
Epoch: 1/20... Training Step: 2699... Training loss: 1.9896... 0.0980 sec/batch
Epoch: 1/20... Training Step: 2700... Training loss: 1.8953... 0.0731 sec/batch
Epoch: 1/20... Training Step: 2701... Training loss: 1.9456... 0.0608 sec/batch
Epoch: 1/20... Training Step: 2702... Training loss: 1.8232... 0.0643 sec/batch
Epoch: 1/20... Training Step: 2703... Training loss: 1.9072... 0.0746 sec/batch
Epoch: 1/20... Training Step: 2704... Training loss: 1.9580... 0.0777 sec/batch
Epoch: 1/20... Training Step: 2705... Training loss: 2.0227... 0.0635 sec/batch
Epoch: 1/20... Training Step: 2706... Training loss: 1.9280... 0.0587 sec/batch
Epoch: 1/20... Training Step: 2707... Training loss: 1.7945... 0.0921 sec/batch
Epoch: 1/20... Training Step: 2708... Training loss: 1.8122... 0.0607 sec/batch
Epoch: 1/20... Training Step: 2709... Training loss: 1.8847... 0.0566 sec/batch
Epoch: 1/20... Training Step: 2710... Training loss: 1.7396... 0.0878 sec/batch
Epoch: 1/20... Training Step: 2711... Training loss: 1.6815... 0.0775 sec/batch
Epoch: 1/20... Training Step: 2712... Training loss: 1.8590... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2713... Training loss: 1.8219... 0.0782 sec/batch
Epoch: 1/20... Training Step: 2714... Training loss: 1.8087... 0.0597 sec/batch
Epoch: 1/20... Training Step: 2715... Training loss: 1.7144... 0.0576 sec/batch
Epoch: 1/20... Training Step: 2716... Training loss: 1.8411... 0.0674 sec/batch
Epoch: 1/20... Training Step: 2717... Training loss: 1.7146... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2718... Training loss: 1.7301... 0.0622 sec/batch
Epoch: 1/20... Training Step: 2719... Training loss: 1.7950... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2720... Training loss: 1.7163... 0.0834 sec/batch
Epoch: 1/20... Training Step: 2721... Training loss: 1.8365... 0.0720 sec/batch
Epoch: 1/20... Training Step: 2722... Training loss: 1.7647... 0.0578 sec/batch
Epoch: 1/20... Training Step: 2723... Training loss: 1.8381... 0.0549 sec/batch
Epoch: 1/20... Training Step: 2724... Training loss: 1.7592... 0.0593 sec/batch
Epoch: 1/20... Training Step: 2725... Training loss: 1.8304... 0.0832 sec/batch
Epoch: 1/20... Training Step: 2726... Training loss: 1.9797... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2727... Training loss: 1.8585... 0.0494 sec/batch
Epoch: 1/20... Training Step: 2728... Training loss: 1.9380... 0.0807 sec/batch
Epoch: 1/20... Training Step: 2729... Training loss: 1.8562... 0.0804 sec/batch
Epoch: 1/20... Training Step: 2730... Training loss: 1.7522... 0.0556 sec/batch
Epoch: 1/20... Training Step: 2731... Training loss: 1.8727... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2732... Training loss: 1.7870... 0.0640 sec/batch
Epoch: 1/20... Training Step: 2733... Training loss: 1.8264... 0.0713 sec/batch
Epoch: 1/20... Training Step: 2734... Training loss: 1.8004... 0.0518 sec/batch
Epoch: 1/20... Training Step: 2735... Training loss: 1.7593... 0.0621 sec/batch
Epoch: 1/20... Training Step: 2736... Training loss: 1.8741... 0.0899 sec/batch
Epoch: 1/20... Training Step: 2737... Training loss: 1.7046... 0.0686 sec/batch
Epoch: 1/20... Training Step: 2738... Training loss: 1.7606... 0.0669 sec/batch
Epoch: 1/20... Training Step: 2739... Training loss: 1.9973... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2740... Training loss: 1.8063... 0.0671 sec/batch
Epoch: 1/20... Training Step: 2741... Training loss: 1.8479... 0.0826 sec/batch
Epoch: 1/20... Training Step: 2742... Training loss: 1.9131... 0.0727 sec/batch
Epoch: 1/20... Training Step: 2743... Training loss: 2.0032... 0.0595 sec/batch
Epoch: 1/20... Training Step: 2744... Training loss: 1.8640... 0.0646 sec/batch
Epoch: 1/20... Training Step: 2745... Training loss: 1.8757... 0.0739 sec/batch
Epoch: 1/20... Training Step: 2746... Training loss: 1.9538... 0.0498 sec/batch
Epoch: 1/20... Training Step: 2747... Training loss: 1.7635... 0.0661 sec/batch
Epoch: 1/20... Training Step: 2748... Training loss: 1.9230... 0.0621 sec/batch
Epoch: 1/20... Training Step: 2749... Training loss: 1.9833... 0.0645 sec/batch
Epoch: 1/20... Training Step: 2750... Training loss: 1.7436... 0.0575 sec/batch
Epoch: 1/20... Training Step: 2751... Training loss: 1.7765... 0.0728 sec/batch
Epoch: 1/20... Training Step: 2752... Training loss: 1.8924... 0.0871 sec/batch
Epoch: 1/20... Training Step: 2753... Training loss: 1.7759... 0.0597 sec/batch
Epoch: 1/20... Training Step: 2754... Training loss: 1.8446... 0.0558 sec/batch
Epoch: 1/20... Training Step: 2755... Training loss: 1.9467... 0.0652 sec/batch
Epoch: 1/20... Training Step: 2756... Training loss: 1.8583... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2757... Training loss: 1.8500... 0.0673 sec/batch
Epoch: 1/20... Training Step: 2758... Training loss: 1.8421... 0.0621 sec/batch
Epoch: 1/20... Training Step: 2759... Training loss: 1.7961... 0.0760 sec/batch
Epoch: 1/20... Training Step: 2760... Training loss: 1.8558... 0.0621 sec/batch
Epoch: 1/20... Training Step: 2761... Training loss: 1.7235... 0.0826 sec/batch
Epoch: 1/20... Training Step: 2762... Training loss: 1.8040... 0.0661 sec/batch
Epoch: 1/20... Training Step: 2763... Training loss: 1.8784... 0.0649 sec/batch
Epoch: 1/20... Training Step: 2764... Training loss: 1.6838... 0.0915 sec/batch
Epoch: 1/20... Training Step: 2765... Training loss: 1.9949... 0.0497 sec/batch
Epoch: 1/20... Training Step: 2766... Training loss: 1.7466... 0.0588 sec/batch
Epoch: 1/20... Training Step: 2767... Training loss: 1.6969... 0.0754 sec/batch
Epoch: 1/20... Training Step: 2768... Training loss: 1.8293... 0.0632 sec/batch
Epoch: 1/20... Training Step: 2769... Training loss: 1.9493... 0.0546 sec/batch
Epoch: 1/20... Training Step: 2770... Training loss: 1.9097... 0.0625 sec/batch
Epoch: 1/20... Training Step: 2771... Training loss: 1.8529... 0.0718 sec/batch
Epoch: 1/20... Training Step: 2772... Training loss: 1.9466... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2773... Training loss: 1.7438... 0.0573 sec/batch
Epoch: 1/20... Training Step: 2774... Training loss: 1.8208... 0.0849 sec/batch
Epoch: 1/20... Training Step: 2775... Training loss: 1.8011... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2776... Training loss: 1.7283... 0.0578 sec/batch
Epoch: 1/20... Training Step: 2777... Training loss: 1.8104... 0.0542 sec/batch
Epoch: 1/20... Training Step: 2778... Training loss: 1.6951... 0.0613 sec/batch
Epoch: 1/20... Training Step: 2779... Training loss: 1.7532... 0.0727 sec/batch
Epoch: 1/20... Training Step: 2780... Training loss: 1.7054... 0.0580 sec/batch
Epoch: 1/20... Training Step: 2781... Training loss: 1.7088... 0.0600 sec/batch
Epoch: 1/20... Training Step: 2782... Training loss: 1.9736... 0.0626 sec/batch
Epoch: 1/20... Training Step: 2783... Training loss: 1.7725... 0.0766 sec/batch
Epoch: 1/20... Training Step: 2784... Training loss: 1.7170... 0.0598 sec/batch
Epoch: 1/20... Training Step: 2785... Training loss: 1.7199... 0.0539 sec/batch
Epoch: 1/20... Training Step: 2786... Training loss: 1.9024... 0.0663 sec/batch
Epoch: 1/20... Training Step: 2787... Training loss: 1.7405... 0.0859 sec/batch
Epoch: 1/20... Training Step: 2788... Training loss: 1.7967... 0.0672 sec/batch
Epoch: 1/20... Training Step: 2789... Training loss: 1.7134... 0.0523 sec/batch
Epoch: 1/20... Training Step: 2790... Training loss: 1.7416... 0.0765 sec/batch
Epoch: 1/20... Training Step: 2791... Training loss: 1.7543... 0.0868 sec/batch
Epoch: 1/20... Training Step: 2792... Training loss: 1.6746... 0.0720 sec/batch
Epoch: 1/20... Training Step: 2793... Training loss: 1.6738... 0.0723 sec/batch
Epoch: 1/20... Training Step: 2794... Training loss: 1.6681... 0.0599 sec/batch
Epoch: 1/20... Training Step: 2795... Training loss: 1.6810... 0.0698 sec/batch
Epoch: 1/20... Training Step: 2796... Training loss: 1.9016... 0.0967 sec/batch
Epoch: 1/20... Training Step: 2797... Training loss: 1.6299... 0.0736 sec/batch
Epoch: 1/20... Training Step: 2798... Training loss: 1.7545... 0.0643 sec/batch
Epoch: 1/20... Training Step: 2799... Training loss: 1.8681... 0.0684 sec/batch
Epoch: 1/20... Training Step: 2800... Training loss: 1.6554... 0.0829 sec/batch
Epoch: 1/20... Training Step: 2801... Training loss: 1.8845... 0.0693 sec/batch
Epoch: 1/20... Training Step: 2802... Training loss: 1.8991... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2803... Training loss: 1.8101... 0.0618 sec/batch
Epoch: 1/20... Training Step: 2804... Training loss: 1.8080... 0.0619 sec/batch
Epoch: 1/20... Training Step: 2805... Training loss: 1.8114... 0.0746 sec/batch
Epoch: 1/20... Training Step: 2806... Training loss: 1.8768... 0.0764 sec/batch
Epoch: 1/20... Training Step: 2807... Training loss: 1.8948... 0.0634 sec/batch
Epoch: 1/20... Training Step: 2808... Training loss: 1.8236... 0.0632 sec/batch
Epoch: 1/20... Training Step: 2809... Training loss: 1.8334... 0.0860 sec/batch
Epoch: 1/20... Training Step: 2810... Training loss: 1.8719... 0.0834 sec/batch
Epoch: 1/20... Training Step: 2811... Training loss: 1.8313... 0.0686 sec/batch
Epoch: 1/20... Training Step: 2812... Training loss: 1.7817... 0.0801 sec/batch
Epoch: 1/20... Training Step: 2813... Training loss: 1.7816... 0.0688 sec/batch
Epoch: 1/20... Training Step: 2814... Training loss: 1.7744... 0.0677 sec/batch
Epoch: 1/20... Training Step: 2815... Training loss: 1.8826... 0.0692 sec/batch
Epoch: 1/20... Training Step: 2816... Training loss: 1.9411... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2817... Training loss: 1.8500... 0.0584 sec/batch
Epoch: 1/20... Training Step: 2818... Training loss: 1.9279... 0.0869 sec/batch
Epoch: 1/20... Training Step: 2819... Training loss: 1.7868... 0.0777 sec/batch
Epoch: 1/20... Training Step: 2820... Training loss: 1.9028... 0.0630 sec/batch
Epoch: 1/20... Training Step: 2821... Training loss: 1.9360... 0.0723 sec/batch
Epoch: 1/20... Training Step: 2822... Training loss: 1.7799... 0.0807 sec/batch
Epoch: 1/20... Training Step: 2823... Training loss: 1.8867... 0.0707 sec/batch
Epoch: 1/20... Training Step: 2824... Training loss: 1.6639... 0.0600 sec/batch
Epoch: 1/20... Training Step: 2825... Training loss: 1.9959... 0.0882 sec/batch
Epoch: 1/20... Training Step: 2826... Training loss: 1.8608... 0.0712 sec/batch
Epoch: 1/20... Training Step: 2827... Training loss: 1.8938... 0.0840 sec/batch
Epoch: 1/20... Training Step: 2828... Training loss: 1.8680... 0.1095 sec/batch
Epoch: 1/20... Training Step: 2829... Training loss: 1.8294... 0.0756 sec/batch
Epoch: 1/20... Training Step: 2830... Training loss: 1.8524... 0.0826 sec/batch
Epoch: 1/20... Training Step: 2831... Training loss: 1.8338... 0.0850 sec/batch
Epoch: 1/20... Training Step: 2832... Training loss: 1.9419... 0.0663 sec/batch
Epoch: 1/20... Training Step: 2833... Training loss: 1.8431... 0.0631 sec/batch
Epoch: 1/20... Training Step: 2834... Training loss: 1.9747... 0.0670 sec/batch
Epoch: 1/20... Training Step: 2835... Training loss: 1.8692... 0.0759 sec/batch
Epoch: 1/20... Training Step: 2836... Training loss: 2.0364... 0.0699 sec/batch
Epoch: 1/20... Training Step: 2837... Training loss: 1.9092... 0.1326 sec/batch
Epoch: 1/20... Training Step: 2838... Training loss: 1.8305... 0.1073 sec/batch
Epoch: 1/20... Training Step: 2839... Training loss: 1.8593... 0.1093 sec/batch
Epoch: 1/20... Training Step: 2840... Training loss: 1.8386... 0.0952 sec/batch
Epoch: 1/20... Training Step: 2841... Training loss: 1.9752... 0.1068 sec/batch
Epoch: 1/20... Training Step: 2842... Training loss: 1.9179... 0.1270 sec/batch
Epoch: 1/20... Training Step: 2843... Training loss: 1.6323... 0.1262 sec/batch
Epoch: 1/20... Training Step: 2844... Training loss: 1.7425... 0.1228 sec/batch
Epoch: 1/20... Training Step: 2845... Training loss: 1.7861... 0.1437 sec/batch
Epoch: 1/20... Training Step: 2846... Training loss: 1.8116... 0.0898 sec/batch
Epoch: 1/20... Training Step: 2847... Training loss: 1.8721... 0.1003 sec/batch
Epoch: 1/20... Training Step: 2848... Training loss: 1.9049... 0.1323 sec/batch
Epoch: 1/20... Training Step: 2849... Training loss: 1.8075... 0.1380 sec/batch
Epoch: 1/20... Training Step: 2850... Training loss: 1.9546... 0.1306 sec/batch
Epoch: 1/20... Training Step: 2851... Training loss: 1.8362... 0.1297 sec/batch
Epoch: 1/20... Training Step: 2852... Training loss: 1.9711... 0.1014 sec/batch
Epoch: 1/20... Training Step: 2853... Training loss: 1.9770... 0.1227 sec/batch
Epoch: 1/20... Training Step: 2854... Training loss: 1.8266... 0.1058 sec/batch
Epoch: 1/20... Training Step: 2855... Training loss: 1.7768... 0.1337 sec/batch
Epoch: 1/20... Training Step: 2856... Training loss: 1.9446... 0.1319 sec/batch
Epoch: 1/20... Training Step: 2857... Training loss: 1.8349... 0.1336 sec/batch
Epoch: 1/20... Training Step: 2858... Training loss: 1.9408... 0.0928 sec/batch
Epoch: 1/20... Training Step: 2859... Training loss: 1.6348... 0.1392 sec/batch
Epoch: 1/20... Training Step: 2860... Training loss: 1.7450... 0.1113 sec/batch
Epoch: 1/20... Training Step: 2861... Training loss: 1.8337... 0.0971 sec/batch
Epoch: 1/20... Training Step: 2862... Training loss: 1.6821... 0.0988 sec/batch
Epoch: 1/20... Training Step: 2863... Training loss: 1.7954... 0.1027 sec/batch
Epoch: 1/20... Training Step: 2864... Training loss: 1.8278... 0.1115 sec/batch
Epoch: 1/20... Training Step: 2865... Training loss: 1.7313... 0.1000 sec/batch
Epoch: 1/20... Training Step: 2866... Training loss: 1.7448... 0.1323 sec/batch
Epoch: 1/20... Training Step: 2867... Training loss: 1.7473... 0.1291 sec/batch
Epoch: 1/20... Training Step: 2868... Training loss: 1.8308... 0.1004 sec/batch
Epoch: 1/20... Training Step: 2869... Training loss: 1.8749... 0.0919 sec/batch
Epoch: 1/20... Training Step: 2870... Training loss: 1.8317... 0.1003 sec/batch
Epoch: 1/20... Training Step: 2871... Training loss: 1.7280... 0.1132 sec/batch
Epoch: 1/20... Training Step: 2872... Training loss: 1.7228... 0.0719 sec/batch
Epoch: 1/20... Training Step: 2873... Training loss: 1.8598... 0.0822 sec/batch
Epoch: 1/20... Training Step: 2874... Training loss: 1.7942... 0.0839 sec/batch
Epoch: 1/20... Training Step: 2875... Training loss: 1.8625... 0.0710 sec/batch
Epoch: 1/20... Training Step: 2876... Training loss: 2.0198... 0.0709 sec/batch
Epoch: 1/20... Training Step: 2877... Training loss: 1.8351... 0.0730 sec/batch
Epoch: 1/20... Training Step: 2878... Training loss: 1.7592... 0.0803 sec/batch
Epoch: 1/20... Training Step: 2879... Training loss: 1.9721... 0.0716 sec/batch
Epoch: 1/20... Training Step: 2880... Training loss: 1.7298... 0.0794 sec/batch
Epoch: 1/20... Training Step: 2881... Training loss: 1.9230... 0.0832 sec/batch
Epoch: 1/20... Training Step: 2882... Training loss: 1.9182... 0.0895 sec/batch
Epoch: 1/20... Training Step: 2883... Training loss: 1.8799... 0.1187 sec/batch
Epoch: 1/20... Training Step: 2884... Training loss: 1.8093... 0.0586 sec/batch
Epoch: 1/20... Training Step: 2885... Training loss: 1.8300... 0.0891 sec/batch
Epoch: 1/20... Training Step: 2886... Training loss: 1.6768... 0.0952 sec/batch
Epoch: 1/20... Training Step: 2887... Training loss: 1.7123... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2888... Training loss: 1.8489... 0.0681 sec/batch
Epoch: 1/20... Training Step: 2889... Training loss: 1.8903... 0.0902 sec/batch
Epoch: 1/20... Training Step: 2890... Training loss: 1.7673... 0.0716 sec/batch
Epoch: 1/20... Training Step: 2891... Training loss: 1.9052... 0.0738 sec/batch
Epoch: 1/20... Training Step: 2892... Training loss: 1.6801... 0.0882 sec/batch
Epoch: 1/20... Training Step: 2893... Training loss: 1.9417... 0.0727 sec/batch
Epoch: 1/20... Training Step: 2894... Training loss: 1.8188... 0.0617 sec/batch
Epoch: 1/20... Training Step: 2895... Training loss: 1.7475... 0.0948 sec/batch
Epoch: 1/20... Training Step: 2896... Training loss: 1.7257... 0.0755 sec/batch
Epoch: 1/20... Training Step: 2897... Training loss: 1.8161... 0.0684 sec/batch
Epoch: 1/20... Training Step: 2898... Training loss: 1.8926... 0.0739 sec/batch
Epoch: 1/20... Training Step: 2899... Training loss: 1.7832... 0.0783 sec/batch
Epoch: 1/20... Training Step: 2900... Training loss: 1.6827... 0.0691 sec/batch
Epoch: 1/20... Training Step: 2901... Training loss: 1.6255... 0.0729 sec/batch
Epoch: 1/20... Training Step: 2902... Training loss: 1.8146... 0.0746 sec/batch
Epoch: 1/20... Training Step: 2903... Training loss: 1.6744... 0.0717 sec/batch
Epoch: 1/20... Training Step: 2904... Training loss: 1.8117... 0.0830 sec/batch
Epoch: 1/20... Training Step: 2905... Training loss: 1.8236... 0.0719 sec/batch
Epoch: 1/20... Training Step: 2906... Training loss: 1.8310... 0.0772 sec/batch
Epoch: 1/20... Training Step: 2907... Training loss: 1.7897... 0.0773 sec/batch
Epoch: 1/20... Training Step: 2908... Training loss: 1.7968... 0.0723 sec/batch
Epoch: 1/20... Training Step: 2909... Training loss: 1.6390... 0.0644 sec/batch
Epoch: 1/20... Training Step: 2910... Training loss: 1.8350... 0.0822 sec/batch
Epoch: 1/20... Training Step: 2911... Training loss: 1.6379... 0.0786 sec/batch
Epoch: 1/20... Training Step: 2912... Training loss: 1.9167... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2913... Training loss: 1.7520... 0.0730 sec/batch
Epoch: 1/20... Training Step: 2914... Training loss: 1.7561... 0.0721 sec/batch
Epoch: 1/20... Training Step: 2915... Training loss: 1.9656... 0.0918 sec/batch
Epoch: 1/20... Training Step: 2916... Training loss: 1.8122... 0.0900 sec/batch
Epoch: 1/20... Training Step: 2917... Training loss: 1.8387... 0.0645 sec/batch
Epoch: 1/20... Training Step: 2918... Training loss: 1.8027... 0.0684 sec/batch
Epoch: 1/20... Training Step: 2919... Training loss: 1.8784... 0.0832 sec/batch
Epoch: 1/20... Training Step: 2920... Training loss: 1.9276... 0.0853 sec/batch
Epoch: 1/20... Training Step: 2921... Training loss: 1.9413... 0.0784 sec/batch
Epoch: 1/20... Training Step: 2922... Training loss: 1.7454... 0.0772 sec/batch
Epoch: 1/20... Training Step: 2923... Training loss: 1.7574... 0.0682 sec/batch
Epoch: 1/20... Training Step: 2924... Training loss: 1.9459... 0.0835 sec/batch
Epoch: 1/20... Training Step: 2925... Training loss: 1.7524... 0.0751 sec/batch
Epoch: 1/20... Training Step: 2926... Training loss: 1.7057... 0.0647 sec/batch
Epoch: 1/20... Training Step: 2927... Training loss: 1.7800... 0.0703 sec/batch
Epoch: 1/20... Training Step: 2928... Training loss: 1.8081... 0.0794 sec/batch
Epoch: 1/20... Training Step: 2929... Training loss: 1.7048... 0.0803 sec/batch
Epoch: 1/20... Training Step: 2930... Training loss: 1.7342... 0.0665 sec/batch
Epoch: 1/20... Training Step: 2931... Training loss: 1.7472... 0.0847 sec/batch
Epoch: 1/20... Training Step: 2932... Training loss: 1.6920... 0.0588 sec/batch
Epoch: 1/20... Training Step: 2933... Training loss: 1.7633... 0.0518 sec/batch
Epoch: 1/20... Training Step: 2934... Training loss: 1.7737... 0.0857 sec/batch
Epoch: 1/20... Training Step: 2935... Training loss: 1.7968... 0.0929 sec/batch
Epoch: 1/20... Training Step: 2936... Training loss: 1.8322... 0.0592 sec/batch
Epoch: 1/20... Training Step: 2937... Training loss: 1.8498... 0.0622 sec/batch
Epoch: 1/20... Training Step: 2938... Training loss: 1.7698... 0.0628 sec/batch
Epoch: 1/20... Training Step: 2939... Training loss: 1.7583... 0.0894 sec/batch
Epoch: 1/20... Training Step: 2940... Training loss: 1.8400... 0.0591 sec/batch
Epoch: 1/20... Training Step: 2941... Training loss: 1.9128... 0.0905 sec/batch
Epoch: 1/20... Training Step: 2942... Training loss: 1.9857... 0.0691 sec/batch
Epoch: 1/20... Training Step: 2943... Training loss: 1.8799... 0.0689 sec/batch
Epoch: 1/20... Training Step: 2944... Training loss: 1.7499... 0.0736 sec/batch
Epoch: 1/20... Training Step: 2945... Training loss: 1.7429... 0.0675 sec/batch
Epoch: 1/20... Training Step: 2946... Training loss: 1.7534... 0.0570 sec/batch
Epoch: 1/20... Training Step: 2947... Training loss: 1.7815... 0.0792 sec/batch
Epoch: 1/20... Training Step: 2948... Training loss: 1.7972... 0.0755 sec/batch
Epoch: 1/20... Training Step: 2949... Training loss: 1.6563... 0.0650 sec/batch
Epoch: 1/20... Training Step: 2950... Training loss: 1.5771... 0.0634 sec/batch
Epoch: 1/20... Training Step: 2951... Training loss: 1.7051... 0.0925 sec/batch
Epoch: 1/20... Training Step: 2952... Training loss: 1.7770... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2953... Training loss: 1.7531... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2954... Training loss: 1.7248... 0.0671 sec/batch
Epoch: 1/20... Training Step: 2955... Training loss: 1.6997... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2956... Training loss: 1.7848... 0.0683 sec/batch
Epoch: 1/20... Training Step: 2957... Training loss: 1.8163... 0.0880 sec/batch
Epoch: 1/20... Training Step: 2958... Training loss: 1.7708... 0.0649 sec/batch
Epoch: 1/20... Training Step: 2959... Training loss: 1.7751... 0.0569 sec/batch
Epoch: 1/20... Training Step: 2960... Training loss: 1.8622... 0.0552 sec/batch
Epoch: 1/20... Training Step: 2961... Training loss: 1.7666... 0.0872 sec/batch
Epoch: 1/20... Training Step: 2962... Training loss: 1.8565... 0.1025 sec/batch
Epoch: 1/20... Training Step: 2963... Training loss: 1.7794... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2964... Training loss: 1.7275... 0.0699 sec/batch
Epoch: 1/20... Training Step: 2965... Training loss: 1.7953... 0.0667 sec/batch
Epoch: 1/20... Training Step: 2966... Training loss: 1.7212... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2967... Training loss: 1.7119... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2968... Training loss: 1.8874... 0.0807 sec/batch
Epoch: 1/20... Training Step: 2969... Training loss: 1.7890... 0.0781 sec/batch
Epoch: 1/20... Training Step: 2970... Training loss: 1.7369... 0.0603 sec/batch
Epoch: 1/20... Training Step: 2971... Training loss: 1.9668... 0.0820 sec/batch
Epoch: 1/20... Training Step: 2972... Training loss: 1.6901... 0.0691 sec/batch
Epoch: 1/20... Training Step: 2973... Training loss: 1.7443... 0.0802 sec/batch
Epoch: 1/20... Training Step: 2974... Training loss: 1.9177... 0.0693 sec/batch
Epoch: 1/20... Training Step: 2975... Training loss: 1.8431... 0.0623 sec/batch
Epoch: 1/20... Training Step: 2976... Training loss: 1.7756... 0.0666 sec/batch
Epoch: 1/20... Training Step: 2977... Training loss: 1.6679... 0.0906 sec/batch
Epoch: 1/20... Training Step: 2978... Training loss: 1.8457... 0.0701 sec/batch
Epoch: 1/20... Training Step: 2979... Training loss: 1.9223... 0.0659 sec/batch
Epoch: 1/20... Training Step: 2980... Training loss: 2.0281... 0.0561 sec/batch
Epoch: 1/20... Training Step: 2981... Training loss: 1.9648... 0.0720 sec/batch
Epoch: 1/20... Training Step: 2982... Training loss: 1.9141... 0.0829 sec/batch
Epoch: 1/20... Training Step: 2983... Training loss: 1.6447... 0.0662 sec/batch
Epoch: 1/20... Training Step: 2984... Training loss: 1.6804... 0.0669 sec/batch
Epoch: 1/20... Training Step: 2985... Training loss: 1.7717... 0.0821 sec/batch
Epoch: 1/20... Training Step: 2986... Training loss: 1.8517... 0.0637 sec/batch
Epoch: 1/20... Training Step: 2987... Training loss: 1.7434... 0.0564 sec/batch
Epoch: 1/20... Training Step: 2988... Training loss: 1.8874... 0.0907 sec/batch
Epoch: 1/20... Training Step: 2989... Training loss: 1.8869... 0.0760 sec/batch
Epoch: 1/20... Training Step: 2990... Training loss: 1.8570... 0.0568 sec/batch
Epoch: 1/20... Training Step: 2991... Training loss: 1.8060... 0.0509 sec/batch
Epoch: 1/20... Training Step: 2992... Training loss: 1.8270... 0.0644 sec/batch
Epoch: 1/20... Training Step: 2993... Training loss: 1.6347... 0.0923 sec/batch
Epoch: 1/20... Training Step: 2994... Training loss: 1.8304... 0.0585 sec/batch
Epoch: 1/20... Training Step: 2995... Training loss: 1.9531... 0.0661 sec/batch
Epoch: 1/20... Training Step: 2996... Training loss: 1.9619... 0.0616 sec/batch
Epoch: 1/20... Training Step: 2997... Training loss: 2.0497... 0.0642 sec/batch
Epoch: 1/20... Training Step: 2998... Training loss: 1.8140... 0.0564 sec/batch
Epoch: 1/20... Training Step: 2999... Training loss: 1.7522... 0.0500 sec/batch
Epoch: 1/20... Training Step: 3000... Training loss: 1.8082... 0.0676 sec/batch
Epoch: 1/20... Training Step: 3001... Training loss: 1.7900... 0.0526 sec/batch
Epoch: 1/20... Training Step: 3002... Training loss: 1.5939... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3003... Training loss: 1.7717... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3004... Training loss: 1.7870... 0.0675 sec/batch
Epoch: 1/20... Training Step: 3005... Training loss: 1.8755... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3006... Training loss: 1.7832... 0.0691 sec/batch
Epoch: 1/20... Training Step: 3007... Training loss: 1.7861... 0.0530 sec/batch
Epoch: 1/20... Training Step: 3008... Training loss: 1.8659... 0.0748 sec/batch
Epoch: 1/20... Training Step: 3009... Training loss: 1.8411... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3010... Training loss: 1.9335... 0.0517 sec/batch
Epoch: 1/20... Training Step: 3011... Training loss: 1.7153... 0.0598 sec/batch
Epoch: 1/20... Training Step: 3012... Training loss: 1.9558... 0.0605 sec/batch
Epoch: 1/20... Training Step: 3013... Training loss: 1.7844... 0.0863 sec/batch
Epoch: 1/20... Training Step: 3014... Training loss: 1.8567... 0.0616 sec/batch
Epoch: 1/20... Training Step: 3015... Training loss: 1.8918... 0.0692 sec/batch
Epoch: 1/20... Training Step: 3016... Training loss: 1.7973... 0.0817 sec/batch
Epoch: 1/20... Training Step: 3017... Training loss: 1.8900... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3018... Training loss: 1.9068... 0.0780 sec/batch
Epoch: 1/20... Training Step: 3019... Training loss: 1.7448... 0.0723 sec/batch
Epoch: 1/20... Training Step: 3020... Training loss: 1.9574... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3021... Training loss: 1.8633... 0.0740 sec/batch
Epoch: 1/20... Training Step: 3022... Training loss: 1.8715... 0.0919 sec/batch
Epoch: 1/20... Training Step: 3023... Training loss: 1.7579... 0.0707 sec/batch
Epoch: 1/20... Training Step: 3024... Training loss: 1.8148... 0.0679 sec/batch
Epoch: 1/20... Training Step: 3025... Training loss: 1.7941... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3026... Training loss: 1.7720... 0.0599 sec/batch
Epoch: 1/20... Training Step: 3027... Training loss: 1.8389... 0.0741 sec/batch
Epoch: 1/20... Training Step: 3028... Training loss: 1.8848... 0.0711 sec/batch
Epoch: 1/20... Training Step: 3029... Training loss: 1.9530... 0.0765 sec/batch
Epoch: 1/20... Training Step: 3030... Training loss: 1.6779... 0.0878 sec/batch
Epoch: 1/20... Training Step: 3031... Training loss: 1.9303... 0.1052 sec/batch
Epoch: 1/20... Training Step: 3032... Training loss: 1.9000... 0.0791 sec/batch
Epoch: 1/20... Training Step: 3033... Training loss: 1.8163... 0.0594 sec/batch
Epoch: 1/20... Training Step: 3034... Training loss: 1.7808... 0.0961 sec/batch
Epoch: 1/20... Training Step: 3035... Training loss: 1.6258... 0.0584 sec/batch
Epoch: 1/20... Training Step: 3036... Training loss: 1.9391... 0.0646 sec/batch
Epoch: 1/20... Training Step: 3037... Training loss: 1.8324... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3038... Training loss: 1.9801... 0.0666 sec/batch
Epoch: 1/20... Training Step: 3039... Training loss: 1.7063... 0.0747 sec/batch
Epoch: 1/20... Training Step: 3040... Training loss: 1.9187... 0.0616 sec/batch
Epoch: 1/20... Training Step: 3041... Training loss: 1.7412... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3042... Training loss: 1.8039... 0.0734 sec/batch
Epoch: 1/20... Training Step: 3043... Training loss: 1.8468... 0.0641 sec/batch
Epoch: 1/20... Training Step: 3044... Training loss: 1.8227... 0.0783 sec/batch
Epoch: 1/20... Training Step: 3045... Training loss: 1.8362... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3046... Training loss: 1.8243... 0.0701 sec/batch
Epoch: 1/20... Training Step: 3047... Training loss: 1.6590... 0.0721 sec/batch
Epoch: 1/20... Training Step: 3048... Training loss: 2.0268... 0.0617 sec/batch
Epoch: 1/20... Training Step: 3049... Training loss: 1.8701... 0.0856 sec/batch
Epoch: 1/20... Training Step: 3050... Training loss: 1.8809... 0.0800 sec/batch
Epoch: 1/20... Training Step: 3051... Training loss: 1.8753... 0.0678 sec/batch
Epoch: 1/20... Training Step: 3052... Training loss: 1.8834... 0.0594 sec/batch
Epoch: 1/20... Training Step: 3053... Training loss: 2.0683... 0.0800 sec/batch
Epoch: 1/20... Training Step: 3054... Training loss: 1.8250... 0.0738 sec/batch
Epoch: 1/20... Training Step: 3055... Training loss: 1.7893... 0.0613 sec/batch
Epoch: 1/20... Training Step: 3056... Training loss: 1.7730... 0.0938 sec/batch
Epoch: 1/20... Training Step: 3057... Training loss: 1.9244... 0.0697 sec/batch
Epoch: 1/20... Training Step: 3058... Training loss: 1.8275... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3059... Training loss: 1.9639... 0.0999 sec/batch
Epoch: 1/20... Training Step: 3060... Training loss: 1.7650... 0.0540 sec/batch
Epoch: 1/20... Training Step: 3061... Training loss: 1.8812... 0.0585 sec/batch
Epoch: 1/20... Training Step: 3062... Training loss: 1.8542... 0.1045 sec/batch
Epoch: 1/20... Training Step: 3063... Training loss: 1.9346... 0.0773 sec/batch
Epoch: 1/20... Training Step: 3064... Training loss: 1.7553... 0.0670 sec/batch
Epoch: 1/20... Training Step: 3065... Training loss: 1.8253... 0.0950 sec/batch
Epoch: 1/20... Training Step: 3066... Training loss: 1.7947... 0.0834 sec/batch
Epoch: 1/20... Training Step: 3067... Training loss: 1.9629... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3068... Training loss: 1.8782... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3069... Training loss: 1.8264... 0.0769 sec/batch
Epoch: 1/20... Training Step: 3070... Training loss: 1.8457... 0.0642 sec/batch
Epoch: 1/20... Training Step: 3071... Training loss: 1.7652... 0.1026 sec/batch
Epoch: 1/20... Training Step: 3072... Training loss: 1.7789... 0.0792 sec/batch
Epoch: 1/20... Training Step: 3073... Training loss: 1.8724... 0.0816 sec/batch
Epoch: 1/20... Training Step: 3074... Training loss: 1.6464... 0.0871 sec/batch
Epoch: 1/20... Training Step: 3075... Training loss: 1.8762... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3076... Training loss: 1.9025... 0.0611 sec/batch
Epoch: 1/20... Training Step: 3077... Training loss: 1.8948... 0.0695 sec/batch
Epoch: 1/20... Training Step: 3078... Training loss: 1.8118... 0.0842 sec/batch
Epoch: 1/20... Training Step: 3079... Training loss: 1.6961... 0.0645 sec/batch
Epoch: 1/20... Training Step: 3080... Training loss: 1.8299... 0.0655 sec/batch
Epoch: 1/20... Training Step: 3081... Training loss: 1.8522... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3082... Training loss: 1.6752... 0.0638 sec/batch
Epoch: 1/20... Training Step: 3083... Training loss: 1.8574... 0.0737 sec/batch
Epoch: 1/20... Training Step: 3084... Training loss: 1.7420... 0.0859 sec/batch
Epoch: 1/20... Training Step: 3085... Training loss: 1.8632... 0.0638 sec/batch
Epoch: 1/20... Training Step: 3086... Training loss: 1.6881... 0.0618 sec/batch
Epoch: 1/20... Training Step: 3087... Training loss: 1.7059... 0.1072 sec/batch
Epoch: 1/20... Training Step: 3088... Training loss: 1.8003... 0.0958 sec/batch
Epoch: 1/20... Training Step: 3089... Training loss: 1.7797... 0.1167 sec/batch
Epoch: 1/20... Training Step: 3090... Training loss: 1.8026... 0.1061 sec/batch
Epoch: 1/20... Training Step: 3091... Training loss: 1.6566... 0.1245 sec/batch
Epoch: 1/20... Training Step: 3092... Training loss: 1.7253... 0.1269 sec/batch
Epoch: 1/20... Training Step: 3093... Training loss: 1.8774... 0.1660 sec/batch
Epoch: 1/20... Training Step: 3094... Training loss: 1.9240... 0.1380 sec/batch
Epoch: 1/20... Training Step: 3095... Training loss: 1.8867... 0.0860 sec/batch
Epoch: 1/20... Training Step: 3096... Training loss: 1.7657... 0.0947 sec/batch
Epoch: 1/20... Training Step: 3097... Training loss: 1.9157... 0.0993 sec/batch
Epoch: 1/20... Training Step: 3098... Training loss: 1.7752... 0.1008 sec/batch
Epoch: 1/20... Training Step: 3099... Training loss: 1.7141... 0.0952 sec/batch
Epoch: 1/20... Training Step: 3100... Training loss: 1.7539... 0.1200 sec/batch
Epoch: 1/20... Training Step: 3101... Training loss: 1.9259... 0.1202 sec/batch
Epoch: 1/20... Training Step: 3102... Training loss: 1.7366... 0.0854 sec/batch
Epoch: 1/20... Training Step: 3103... Training loss: 1.6383... 0.1018 sec/batch
Epoch: 1/20... Training Step: 3104... Training loss: 1.7543... 0.1274 sec/batch
Epoch: 1/20... Training Step: 3105... Training loss: 1.7561... 0.1115 sec/batch
Epoch: 1/20... Training Step: 3106... Training loss: 1.7346... 0.0991 sec/batch
Epoch: 1/20... Training Step: 3107... Training loss: 1.9053... 0.1368 sec/batch
Epoch: 1/20... Training Step: 3108... Training loss: 1.7158... 0.1225 sec/batch
Epoch: 1/20... Training Step: 3109... Training loss: 1.6624... 0.1345 sec/batch
Epoch: 1/20... Training Step: 3110... Training loss: 1.7554... 0.1062 sec/batch
Epoch: 1/20... Training Step: 3111... Training loss: 1.7316... 0.1420 sec/batch
Epoch: 1/20... Training Step: 3112... Training loss: 1.7658... 0.1450 sec/batch
Epoch: 1/20... Training Step: 3113... Training loss: 1.7367... 0.1452 sec/batch
Epoch: 1/20... Training Step: 3114... Training loss: 1.7541... 0.1291 sec/batch
Epoch: 1/20... Training Step: 3115... Training loss: 1.7656... 0.1472 sec/batch
Epoch: 1/20... Training Step: 3116... Training loss: 1.6453... 0.0985 sec/batch
Epoch: 1/20... Training Step: 3117... Training loss: 1.6814... 0.1617 sec/batch
Epoch: 1/20... Training Step: 3118... Training loss: 1.7119... 0.0788 sec/batch
Epoch: 1/20... Training Step: 3119... Training loss: 1.7898... 0.1322 sec/batch
Epoch: 1/20... Training Step: 3120... Training loss: 1.7208... 0.0906 sec/batch
Epoch: 1/20... Training Step: 3121... Training loss: 1.7716... 0.0844 sec/batch
Epoch: 1/20... Training Step: 3122... Training loss: 1.7857... 0.0757 sec/batch
Epoch: 1/20... Training Step: 3123... Training loss: 1.8782... 0.0705 sec/batch
Epoch: 1/20... Training Step: 3124... Training loss: 1.7886... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3125... Training loss: 1.7540... 0.0776 sec/batch
Epoch: 1/20... Training Step: 3126... Training loss: 1.7893... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3127... Training loss: 1.7797... 0.1009 sec/batch
Epoch: 1/20... Training Step: 3128... Training loss: 1.7475... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3129... Training loss: 1.6294... 0.0792 sec/batch
Epoch: 1/20... Training Step: 3130... Training loss: 1.7530... 0.0878 sec/batch
Epoch: 1/20... Training Step: 3131... Training loss: 1.7503... 0.0646 sec/batch
Epoch: 1/20... Training Step: 3132... Training loss: 1.7684... 0.0574 sec/batch
Epoch: 1/20... Training Step: 3133... Training loss: 2.0029... 0.1038 sec/batch
Epoch: 1/20... Training Step: 3134... Training loss: 1.8520... 0.0795 sec/batch
Epoch: 1/20... Training Step: 3135... Training loss: 1.9253... 0.0892 sec/batch
Epoch: 1/20... Training Step: 3136... Training loss: 1.8190... 0.1019 sec/batch
Epoch: 1/20... Training Step: 3137... Training loss: 1.6653... 0.0680 sec/batch
Epoch: 1/20... Training Step: 3138... Training loss: 1.7360... 0.0829 sec/batch
Epoch: 1/20... Training Step: 3139... Training loss: 1.8280... 0.0953 sec/batch
Epoch: 1/20... Training Step: 3140... Training loss: 1.9155... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3141... Training loss: 1.8418... 0.0747 sec/batch
Epoch: 1/20... Training Step: 3142... Training loss: 1.6005... 0.0921 sec/batch
Epoch: 1/20... Training Step: 3143... Training loss: 1.7996... 0.0740 sec/batch
Epoch: 1/20... Training Step: 3144... Training loss: 1.6741... 0.0654 sec/batch
Epoch: 1/20... Training Step: 3145... Training loss: 1.8413... 0.0885 sec/batch
Epoch: 1/20... Training Step: 3146... Training loss: 1.8600... 0.0660 sec/batch
Epoch: 1/20... Training Step: 3147... Training loss: 1.7899... 0.0778 sec/batch
Epoch: 1/20... Training Step: 3148... Training loss: 1.8163... 0.0758 sec/batch
Epoch: 1/20... Training Step: 3149... Training loss: 1.8171... 0.0840 sec/batch
Epoch: 1/20... Training Step: 3150... Training loss: 1.7359... 0.0673 sec/batch
Epoch: 1/20... Training Step: 3151... Training loss: 1.7936... 0.0828 sec/batch
Epoch: 1/20... Training Step: 3152... Training loss: 1.8478... 0.0869 sec/batch
Epoch: 1/20... Training Step: 3153... Training loss: 1.8169... 0.0742 sec/batch
Epoch: 1/20... Training Step: 3154... Training loss: 1.7093... 0.0742 sec/batch
Epoch: 1/20... Training Step: 3155... Training loss: 1.8239... 0.0767 sec/batch
Epoch: 1/20... Training Step: 3156... Training loss: 1.8282... 0.0639 sec/batch
Epoch: 1/20... Training Step: 3157... Training loss: 1.6706... 0.0902 sec/batch
Epoch: 1/20... Training Step: 3158... Training loss: 1.7713... 0.0840 sec/batch
Epoch: 1/20... Training Step: 3159... Training loss: 1.7354... 0.0830 sec/batch
Epoch: 1/20... Training Step: 3160... Training loss: 1.7481... 0.0758 sec/batch
Epoch: 1/20... Training Step: 3161... Training loss: 1.8750... 0.0671 sec/batch
Epoch: 1/20... Training Step: 3162... Training loss: 1.7786... 0.0738 sec/batch
Epoch: 1/20... Training Step: 3163... Training loss: 1.9195... 0.0977 sec/batch
Epoch: 1/20... Training Step: 3164... Training loss: 1.7543... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3165... Training loss: 1.6905... 0.0633 sec/batch
Epoch: 1/20... Training Step: 3166... Training loss: 1.7259... 0.0813 sec/batch
Epoch: 1/20... Training Step: 3167... Training loss: 1.7131... 0.0851 sec/batch
Epoch: 1/20... Training Step: 3168... Training loss: 1.7357... 0.0655 sec/batch
Epoch: 1/20... Training Step: 3169... Training loss: 1.7197... 0.0933 sec/batch
Epoch: 1/20... Training Step: 3170... Training loss: 1.8475... 0.0709 sec/batch
Epoch: 1/20... Training Step: 3171... Training loss: 1.8277... 0.0680 sec/batch
Epoch: 1/20... Training Step: 3172... Training loss: 1.7150... 0.0927 sec/batch
Epoch: 1/20... Training Step: 3173... Training loss: 1.8351... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3174... Training loss: 1.6784... 0.0603 sec/batch
Epoch: 1/20... Training Step: 3175... Training loss: 1.6966... 0.0846 sec/batch
Epoch: 1/20... Training Step: 3176... Training loss: 1.8500... 0.0639 sec/batch
Epoch: 1/20... Training Step: 3177... Training loss: 1.8852... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3178... Training loss: 1.6878... 0.0864 sec/batch
Epoch: 1/20... Training Step: 3179... Training loss: 1.8098... 0.0780 sec/batch
Epoch: 1/20... Training Step: 3180... Training loss: 1.7326... 0.0608 sec/batch
Epoch: 1/20... Training Step: 3181... Training loss: 1.7870... 0.0713 sec/batch
Epoch: 1/20... Training Step: 3182... Training loss: 1.8102... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3183... Training loss: 1.7657... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3184... Training loss: 1.7856... 0.0848 sec/batch
Epoch: 1/20... Training Step: 3185... Training loss: 1.9668... 0.0904 sec/batch
Epoch: 1/20... Training Step: 3186... Training loss: 1.8111... 0.0649 sec/batch
Epoch: 1/20... Training Step: 3187... Training loss: 1.6694... 0.0533 sec/batch
Epoch: 1/20... Training Step: 3188... Training loss: 1.8718... 0.0679 sec/batch
Epoch: 1/20... Training Step: 3189... Training loss: 1.7211... 0.0679 sec/batch
Epoch: 1/20... Training Step: 3190... Training loss: 1.7549... 0.0628 sec/batch
Epoch: 1/20... Training Step: 3191... Training loss: 1.7377... 0.0858 sec/batch
Epoch: 1/20... Training Step: 3192... Training loss: 1.6853... 0.0734 sec/batch
Epoch: 1/20... Training Step: 3193... Training loss: 1.9001... 0.0575 sec/batch
Epoch: 1/20... Training Step: 3194... Training loss: 1.8119... 0.0711 sec/batch
Epoch: 1/20... Training Step: 3195... Training loss: 1.8271... 0.0697 sec/batch
Epoch: 1/20... Training Step: 3196... Training loss: 1.7255... 0.0853 sec/batch
Epoch: 1/20... Training Step: 3197... Training loss: 1.8075... 0.0649 sec/batch
Epoch: 1/20... Training Step: 3198... Training loss: 1.8534... 0.0751 sec/batch
Epoch: 1/20... Training Step: 3199... Training loss: 1.7452... 0.0701 sec/batch
Epoch: 1/20... Training Step: 3200... Training loss: 1.8560... 0.0658 sec/batch
Epoch: 1/20... Training Step: 3201... Training loss: 1.9277... 0.0632 sec/batch
Epoch: 1/20... Training Step: 3202... Training loss: 1.7624... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3203... Training loss: 1.7264... 0.0820 sec/batch
Epoch: 1/20... Training Step: 3204... Training loss: 1.9259... 0.0675 sec/batch
Epoch: 1/20... Training Step: 3205... Training loss: 1.7359... 0.0634 sec/batch
Epoch: 1/20... Training Step: 3206... Training loss: 1.8146... 0.0873 sec/batch
Epoch: 1/20... Training Step: 3207... Training loss: 1.7698... 0.0812 sec/batch
Epoch: 1/20... Training Step: 3208... Training loss: 1.9091... 0.0644 sec/batch
Epoch: 1/20... Training Step: 3209... Training loss: 1.7382... 0.0937 sec/batch
Epoch: 1/20... Training Step: 3210... Training loss: 1.7423... 0.0608 sec/batch
Epoch: 1/20... Training Step: 3211... Training loss: 1.9396... 0.0637 sec/batch
Epoch: 1/20... Training Step: 3212... Training loss: 1.8440... 0.0849 sec/batch
Epoch: 1/20... Training Step: 3213... Training loss: 1.9263... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3214... Training loss: 1.8127... 0.0543 sec/batch
Epoch: 1/20... Training Step: 3215... Training loss: 1.7889... 0.0846 sec/batch
Epoch: 1/20... Training Step: 3216... Training loss: 1.8237... 0.0703 sec/batch
Epoch: 1/20... Training Step: 3217... Training loss: 1.7138... 0.0658 sec/batch
Epoch: 1/20... Training Step: 3218... Training loss: 1.7134... 0.0727 sec/batch
Epoch: 1/20... Training Step: 3219... Training loss: 1.6981... 0.0815 sec/batch
Epoch: 1/20... Training Step: 3220... Training loss: 1.8029... 0.0520 sec/batch
Epoch: 1/20... Training Step: 3221... Training loss: 1.7409... 0.0888 sec/batch
Epoch: 1/20... Training Step: 3222... Training loss: 1.8326... 0.0796 sec/batch
Epoch: 1/20... Training Step: 3223... Training loss: 1.8416... 0.0737 sec/batch
Epoch: 1/20... Training Step: 3224... Training loss: 1.7447... 0.0692 sec/batch
Epoch: 1/20... Training Step: 3225... Training loss: 1.7620... 0.0729 sec/batch
Epoch: 1/20... Training Step: 3226... Training loss: 1.6716... 0.0586 sec/batch
Epoch: 1/20... Training Step: 3227... Training loss: 1.7737... 0.0629 sec/batch
Epoch: 1/20... Training Step: 3228... Training loss: 1.6967... 0.0709 sec/batch
Epoch: 1/20... Training Step: 3229... Training loss: 1.7468... 0.0728 sec/batch
Epoch: 1/20... Training Step: 3230... Training loss: 1.8524... 0.0749 sec/batch
Epoch: 1/20... Training Step: 3231... Training loss: 1.8421... 0.0680 sec/batch
Epoch: 1/20... Training Step: 3232... Training loss: 1.7734... 0.0539 sec/batch
Epoch: 1/20... Training Step: 3233... Training loss: 1.7155... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3234... Training loss: 1.6648... 0.0695 sec/batch
Epoch: 1/20... Training Step: 3235... Training loss: 1.7939... 0.0629 sec/batch
Epoch: 1/20... Training Step: 3236... Training loss: 1.8299... 0.0620 sec/batch
Epoch: 1/20... Training Step: 3237... Training loss: 1.6872... 0.0829 sec/batch
Epoch: 1/20... Training Step: 3238... Training loss: 1.8045... 0.0727 sec/batch
Epoch: 1/20... Training Step: 3239... Training loss: 1.7651... 0.0562 sec/batch
Epoch: 1/20... Training Step: 3240... Training loss: 1.7469... 0.0634 sec/batch
Epoch: 1/20... Training Step: 3241... Training loss: 1.8149... 0.0895 sec/batch
Epoch: 1/20... Training Step: 3242... Training loss: 1.7108... 0.0868 sec/batch
Epoch: 1/20... Training Step: 3243... Training loss: 1.8969... 0.0548 sec/batch
Epoch: 1/20... Training Step: 3244... Training loss: 1.8202... 0.0620 sec/batch
Epoch: 1/20... Training Step: 3245... Training loss: 1.8206... 0.0744 sec/batch
Epoch: 1/20... Training Step: 3246... Training loss: 1.9071... 0.0709 sec/batch
Epoch: 1/20... Training Step: 3247... Training loss: 1.8831... 0.0549 sec/batch
Epoch: 1/20... Training Step: 3248... Training loss: 1.8774... 0.0768 sec/batch
Epoch: 1/20... Training Step: 3249... Training loss: 1.9716... 0.0755 sec/batch
Epoch: 1/20... Training Step: 3250... Training loss: 1.7923... 0.0649 sec/batch
Epoch: 1/20... Training Step: 3251... Training loss: 1.8700... 0.0500 sec/batch
Epoch: 1/20... Training Step: 3252... Training loss: 1.9725... 0.0656 sec/batch
Epoch: 1/20... Training Step: 3253... Training loss: 2.0579... 0.0784 sec/batch
Epoch: 1/20... Training Step: 3254... Training loss: 1.7767... 0.0618 sec/batch
Epoch: 1/20... Training Step: 3255... Training loss: 1.7755... 0.0651 sec/batch
Epoch: 1/20... Training Step: 3256... Training loss: 1.7989... 0.0603 sec/batch
Epoch: 1/20... Training Step: 3257... Training loss: 1.6656... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3258... Training loss: 1.8478... 0.0680 sec/batch
Epoch: 1/20... Training Step: 3259... Training loss: 1.9203... 0.0874 sec/batch
Epoch: 1/20... Training Step: 3260... Training loss: 1.7433... 0.0708 sec/batch
Epoch: 1/20... Training Step: 3261... Training loss: 1.8643... 0.0557 sec/batch
Epoch: 1/20... Training Step: 3262... Training loss: 1.8045... 0.0701 sec/batch
Epoch: 1/20... Training Step: 3263... Training loss: 1.7646... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3264... Training loss: 1.6504... 0.0526 sec/batch
Epoch: 1/20... Training Step: 3265... Training loss: 1.7506... 0.0689 sec/batch
Epoch: 1/20... Training Step: 3266... Training loss: 1.8007... 0.0627 sec/batch
Epoch: 1/20... Training Step: 3267... Training loss: 1.7439... 0.0665 sec/batch
Epoch: 1/20... Training Step: 3268... Training loss: 1.7791... 0.0486 sec/batch
Epoch: 1/20... Training Step: 3269... Training loss: 1.6644... 0.0665 sec/batch
Epoch: 1/20... Training Step: 3270... Training loss: 1.8824... 0.0826 sec/batch
Epoch: 1/20... Training Step: 3271... Training loss: 1.7798... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3272... Training loss: 1.7003... 0.0585 sec/batch
Epoch: 1/20... Training Step: 3273... Training loss: 1.9135... 0.0959 sec/batch
Epoch: 1/20... Training Step: 3274... Training loss: 1.7779... 0.0688 sec/batch
Epoch: 1/20... Training Step: 3275... Training loss: 1.7622... 0.0606 sec/batch
Epoch: 1/20... Training Step: 3276... Training loss: 1.8410... 0.0648 sec/batch
Epoch: 1/20... Training Step: 3277... Training loss: 1.8345... 0.0722 sec/batch
Epoch: 1/20... Training Step: 3278... Training loss: 1.7440... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3279... Training loss: 1.7522... 0.0774 sec/batch
Epoch: 1/20... Training Step: 3280... Training loss: 1.6467... 0.0798 sec/batch
Epoch: 1/20... Training Step: 3281... Training loss: 1.7617... 0.0559 sec/batch
Epoch: 1/20... Training Step: 3282... Training loss: 1.7391... 0.0786 sec/batch
Epoch: 1/20... Training Step: 3283... Training loss: 1.8001... 0.0652 sec/batch
Epoch: 1/20... Training Step: 3284... Training loss: 1.7984... 0.0507 sec/batch
Epoch: 1/20... Training Step: 3285... Training loss: 1.6820... 0.0655 sec/batch
Epoch: 1/20... Training Step: 3286... Training loss: 1.7643... 0.0816 sec/batch
Epoch: 1/20... Training Step: 3287... Training loss: 1.7565... 0.0599 sec/batch
Epoch: 1/20... Training Step: 3288... Training loss: 1.7504... 0.0586 sec/batch
Epoch: 1/20... Training Step: 3289... Training loss: 1.8873... 0.0741 sec/batch
Epoch: 1/20... Training Step: 3290... Training loss: 1.9194... 0.0745 sec/batch
Epoch: 1/20... Training Step: 3291... Training loss: 1.9275... 0.0557 sec/batch
Epoch: 1/20... Training Step: 3292... Training loss: 1.8530... 0.0722 sec/batch
Epoch: 1/20... Training Step: 3293... Training loss: 1.7759... 0.0673 sec/batch
Epoch: 1/20... Training Step: 3294... Training loss: 1.8335... 0.0548 sec/batch
Epoch: 1/20... Training Step: 3295... Training loss: 1.8305... 0.0689 sec/batch
Epoch: 1/20... Training Step: 3296... Training loss: 1.8749... 0.0765 sec/batch
Epoch: 1/20... Training Step: 3297... Training loss: 1.8492... 0.0896 sec/batch
Epoch: 1/20... Training Step: 3298... Training loss: 1.8433... 0.0572 sec/batch
Epoch: 1/20... Training Step: 3299... Training loss: 1.8148... 0.0602 sec/batch
Epoch: 1/20... Training Step: 3300... Training loss: 1.9067... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3301... Training loss: 1.8856... 0.0732 sec/batch
Epoch: 1/20... Training Step: 3302... Training loss: 1.8987... 0.0589 sec/batch
Epoch: 1/20... Training Step: 3303... Training loss: 1.7952... 0.0900 sec/batch
Epoch: 1/20... Training Step: 3304... Training loss: 1.8892... 0.0789 sec/batch
Epoch: 1/20... Training Step: 3305... Training loss: 1.9591... 0.0684 sec/batch
Epoch: 1/20... Training Step: 3306... Training loss: 1.8589... 0.0700 sec/batch
Epoch: 1/20... Training Step: 3307... Training loss: 1.6850... 0.0636 sec/batch
Epoch: 1/20... Training Step: 3308... Training loss: 1.7790... 0.0529 sec/batch
Epoch: 1/20... Training Step: 3309... Training loss: 1.7631... 0.0593 sec/batch
Epoch: 1/20... Training Step: 3310... Training loss: 1.8513... 0.0911 sec/batch
Epoch: 1/20... Training Step: 3311... Training loss: 1.7796... 0.0640 sec/batch
Epoch: 1/20... Training Step: 3312... Training loss: 1.7618... 0.0793 sec/batch
Epoch: 1/20... Training Step: 3313... Training loss: 1.7395... 0.0864 sec/batch
Epoch: 1/20... Training Step: 3314... Training loss: 1.8844... 0.0588 sec/batch
Epoch: 1/20... Training Step: 3315... Training loss: 1.7995... 0.0624 sec/batch
Epoch: 1/20... Training Step: 3316... Training loss: 1.9360... 0.0855 sec/batch
Epoch: 1/20... Training Step: 3317... Training loss: 1.8407... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3318... Training loss: 1.8030... 0.0602 sec/batch
Epoch: 1/20... Training Step: 3319... Training loss: 1.8401... 0.0788 sec/batch
Epoch: 1/20... Training Step: 3320... Training loss: 1.8263... 0.0872 sec/batch
Epoch: 1/20... Training Step: 3321... Training loss: 1.8067... 0.0661 sec/batch
Epoch: 1/20... Training Step: 3322... Training loss: 1.7429... 0.0648 sec/batch
Epoch: 1/20... Training Step: 3323... Training loss: 1.7787... 0.0629 sec/batch
Epoch: 1/20... Training Step: 3324... Training loss: 1.6599... 0.0658 sec/batch
Epoch: 1/20... Training Step: 3325... Training loss: 1.8673... 0.0870 sec/batch
Epoch: 1/20... Training Step: 3326... Training loss: 1.9031... 0.0818 sec/batch
Epoch: 1/20... Training Step: 3327... Training loss: 1.8103... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3328... Training loss: 1.8216... 0.0602 sec/batch
Epoch: 1/20... Training Step: 3329... Training loss: 1.9177... 0.0644 sec/batch
Epoch: 1/20... Training Step: 3330... Training loss: 1.8752... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3331... Training loss: 1.6679... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3332... Training loss: 1.9535... 0.0926 sec/batch
Epoch: 1/20... Training Step: 3333... Training loss: 1.8210... 0.0749 sec/batch
Epoch: 1/20... Training Step: 3334... Training loss: 1.7947... 0.0705 sec/batch
Epoch: 1/20... Training Step: 3335... Training loss: 1.9249... 0.0750 sec/batch
Epoch: 1/20... Training Step: 3336... Training loss: 1.9112... 0.0670 sec/batch
Epoch: 1/20... Training Step: 3337... Training loss: 1.8567... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3338... Training loss: 1.8346... 0.0971 sec/batch
Epoch: 1/20... Training Step: 3339... Training loss: 1.7237... 0.0930 sec/batch
Epoch: 1/20... Training Step: 3340... Training loss: 1.7524... 0.0948 sec/batch
Epoch: 1/20... Training Step: 3341... Training loss: 1.8260... 0.0591 sec/batch
Epoch: 1/20... Training Step: 3342... Training loss: 1.8033... 0.0544 sec/batch
Epoch: 1/20... Training Step: 3343... Training loss: 1.8039... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3344... Training loss: 1.7999... 0.0613 sec/batch
Epoch: 1/20... Training Step: 3345... Training loss: 1.7551... 0.0725 sec/batch
Epoch: 1/20... Training Step: 3346... Training loss: 1.9087... 0.0646 sec/batch
Epoch: 1/20... Training Step: 3347... Training loss: 1.9675... 0.0597 sec/batch
Epoch: 1/20... Training Step: 3348... Training loss: 1.8229... 0.0917 sec/batch
Epoch: 1/20... Training Step: 3349... Training loss: 1.8953... 0.0651 sec/batch
Epoch: 1/20... Training Step: 3350... Training loss: 1.6822... 0.0567 sec/batch
Epoch: 1/20... Training Step: 3351... Training loss: 1.8632... 0.0569 sec/batch
Epoch: 1/20... Training Step: 3352... Training loss: 1.6400... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3353... Training loss: 1.6603... 0.0676 sec/batch
Epoch: 1/20... Training Step: 3354... Training loss: 1.7405... 0.0695 sec/batch
Epoch: 1/20... Training Step: 3355... Training loss: 1.8113... 0.0670 sec/batch
Epoch: 1/20... Training Step: 3356... Training loss: 1.7666... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3357... Training loss: 1.8167... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3358... Training loss: 1.9340... 0.0578 sec/batch
Epoch: 1/20... Training Step: 3359... Training loss: 1.8207... 0.0605 sec/batch
Epoch: 1/20... Training Step: 3360... Training loss: 1.7206... 0.0629 sec/batch
Epoch: 1/20... Training Step: 3361... Training loss: 1.7338... 0.0661 sec/batch
Epoch: 1/20... Training Step: 3362... Training loss: 1.6498... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3363... Training loss: 1.7582... 0.0791 sec/batch
Epoch: 1/20... Training Step: 3364... Training loss: 1.7171... 0.0701 sec/batch
Epoch: 1/20... Training Step: 3365... Training loss: 1.8135... 0.0603 sec/batch
Epoch: 1/20... Training Step: 3366... Training loss: 1.7693... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3367... Training loss: 1.8074... 0.0542 sec/batch
Epoch: 1/20... Training Step: 3368... Training loss: 1.8985... 0.0830 sec/batch
Epoch: 1/20... Training Step: 3369... Training loss: 1.8361... 0.0608 sec/batch
Epoch: 1/20... Training Step: 3370... Training loss: 1.7869... 0.0614 sec/batch
Epoch: 1/20... Training Step: 3371... Training loss: 1.8454... 0.0746 sec/batch
Epoch: 1/20... Training Step: 3372... Training loss: 1.9236... 0.0724 sec/batch
Epoch: 1/20... Training Step: 3373... Training loss: 1.8241... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3374... Training loss: 1.7987... 0.0744 sec/batch
Epoch: 1/20... Training Step: 3375... Training loss: 1.8243... 0.0708 sec/batch
Epoch: 1/20... Training Step: 3376... Training loss: 1.7788... 0.0580 sec/batch
Epoch: 1/20... Training Step: 3377... Training loss: 1.9742... 0.0833 sec/batch
Epoch: 1/20... Training Step: 3378... Training loss: 1.6390... 0.0639 sec/batch
Epoch: 1/20... Training Step: 3379... Training loss: 1.6094... 0.0552 sec/batch
Epoch: 1/20... Training Step: 3380... Training loss: 1.6860... 0.0605 sec/batch
Epoch: 1/20... Training Step: 3381... Training loss: 1.7000... 0.0608 sec/batch
Epoch: 1/20... Training Step: 3382... Training loss: 1.7347... 0.0596 sec/batch
Epoch: 1/20... Training Step: 3383... Training loss: 1.7200... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3384... Training loss: 1.7715... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3385... Training loss: 1.7954... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3386... Training loss: 1.6977... 0.0672 sec/batch
Epoch: 1/20... Training Step: 3387... Training loss: 1.8354... 0.0753 sec/batch
Epoch: 1/20... Training Step: 3388... Training loss: 1.8458... 0.0775 sec/batch
Epoch: 1/20... Training Step: 3389... Training loss: 1.8414... 0.0614 sec/batch
Epoch: 1/20... Training Step: 3390... Training loss: 1.7607... 0.0602 sec/batch
Epoch: 1/20... Training Step: 3391... Training loss: 1.8232... 0.0846 sec/batch
Epoch: 1/20... Training Step: 3392... Training loss: 1.7523... 0.0747 sec/batch
Epoch: 1/20... Training Step: 3393... Training loss: 1.7956... 0.0544 sec/batch
Epoch: 1/20... Training Step: 3394... Training loss: 1.7614... 0.0615 sec/batch
Epoch: 1/20... Training Step: 3395... Training loss: 1.6473... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3396... Training loss: 1.9472... 0.0736 sec/batch
Epoch: 1/20... Training Step: 3397... Training loss: 1.8129... 0.0625 sec/batch
Epoch: 1/20... Training Step: 3398... Training loss: 1.7757... 0.0569 sec/batch
Epoch: 1/20... Training Step: 3399... Training loss: 1.8465... 0.0611 sec/batch
Epoch: 1/20... Training Step: 3400... Training loss: 1.7742... 0.0714 sec/batch
Epoch: 1/20... Training Step: 3401... Training loss: 1.7432... 0.0592 sec/batch
Epoch: 1/20... Training Step: 3402... Training loss: 1.8084... 0.0759 sec/batch
Epoch: 1/20... Training Step: 3403... Training loss: 1.7231... 0.0912 sec/batch
Epoch: 1/20... Training Step: 3404... Training loss: 1.6827... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3405... Training loss: 1.8087... 0.0640 sec/batch
Epoch: 1/20... Training Step: 3406... Training loss: 1.7591... 0.0766 sec/batch
Epoch: 1/20... Training Step: 3407... Training loss: 1.5480... 0.0665 sec/batch
Epoch: 1/20... Training Step: 3408... Training loss: 1.7168... 0.0719 sec/batch
Epoch: 1/20... Training Step: 3409... Training loss: 1.8343... 0.0691 sec/batch
Epoch: 1/20... Training Step: 3410... Training loss: 1.8756... 0.0875 sec/batch
Epoch: 1/20... Training Step: 3411... Training loss: 1.7549... 0.0801 sec/batch
Epoch: 1/20... Training Step: 3412... Training loss: 1.6113... 0.0707 sec/batch
Epoch: 1/20... Training Step: 3413... Training loss: 1.7012... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3414... Training loss: 1.6886... 0.0702 sec/batch
Epoch: 1/20... Training Step: 3415... Training loss: 1.7928... 0.0714 sec/batch
Epoch: 1/20... Training Step: 3416... Training loss: 1.7064... 0.0663 sec/batch
Epoch: 1/20... Training Step: 3417... Training loss: 1.8741... 0.0982 sec/batch
Epoch: 1/20... Training Step: 3418... Training loss: 1.8802... 0.0815 sec/batch
Epoch: 1/20... Training Step: 3419... Training loss: 1.8469... 0.0637 sec/batch
Epoch: 1/20... Training Step: 3420... Training loss: 1.9422... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3421... Training loss: 1.8446... 0.0726 sec/batch
Epoch: 1/20... Training Step: 3422... Training loss: 1.8118... 0.0560 sec/batch
Epoch: 1/20... Training Step: 3423... Training loss: 1.8228... 0.0836 sec/batch
Epoch: 1/20... Training Step: 3424... Training loss: 1.7502... 0.0905 sec/batch
Epoch: 1/20... Training Step: 3425... Training loss: 1.9150... 0.0543 sec/batch
Epoch: 1/20... Training Step: 3426... Training loss: 1.7590... 0.0541 sec/batch
Epoch: 1/20... Training Step: 3427... Training loss: 1.7556... 0.0890 sec/batch
Epoch: 1/20... Training Step: 3428... Training loss: 1.9061... 0.0702 sec/batch
Epoch: 1/20... Training Step: 3429... Training loss: 1.7910... 0.0622 sec/batch
Epoch: 1/20... Training Step: 3430... Training loss: 1.6663... 0.0949 sec/batch
Epoch: 1/20... Training Step: 3431... Training loss: 1.8165... 0.0683 sec/batch
Epoch: 1/20... Training Step: 3432... Training loss: 1.8454... 0.0698 sec/batch
Epoch: 1/20... Training Step: 3433... Training loss: 1.8732... 0.0929 sec/batch
Epoch: 1/20... Training Step: 3434... Training loss: 1.8405... 0.0651 sec/batch
Epoch: 1/20... Training Step: 3435... Training loss: 1.8296... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3436... Training loss: 1.7931... 0.0887 sec/batch
Epoch: 1/20... Training Step: 3437... Training loss: 1.7953... 0.0819 sec/batch
Epoch: 1/20... Training Step: 3438... Training loss: 1.8169... 0.0602 sec/batch
Epoch: 1/20... Training Step: 3439... Training loss: 1.9043... 0.0884 sec/batch
Epoch: 1/20... Training Step: 3440... Training loss: 1.9322... 0.0772 sec/batch
Epoch: 1/20... Training Step: 3441... Training loss: 1.7574... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3442... Training loss: 1.6808... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3443... Training loss: 1.7528... 0.0777 sec/batch
Epoch: 1/20... Training Step: 3444... Training loss: 1.8177... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3445... Training loss: 1.8263... 0.0853 sec/batch
Epoch: 1/20... Training Step: 3446... Training loss: 1.8238... 0.0665 sec/batch
Epoch: 1/20... Training Step: 3447... Training loss: 1.8475... 0.0586 sec/batch
Epoch: 1/20... Training Step: 3448... Training loss: 1.6966... 0.0690 sec/batch
Epoch: 1/20... Training Step: 3449... Training loss: 1.8065... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3450... Training loss: 1.8122... 0.0594 sec/batch
Epoch: 1/20... Training Step: 3451... Training loss: 1.9112... 0.0691 sec/batch
Epoch: 1/20... Training Step: 3452... Training loss: 1.8306... 0.0738 sec/batch
Epoch: 1/20... Training Step: 3453... Training loss: 1.7636... 0.0648 sec/batch
Epoch: 1/20... Training Step: 3454... Training loss: 1.6434... 0.0638 sec/batch
Epoch: 1/20... Training Step: 3455... Training loss: 1.6553... 0.0551 sec/batch
Epoch: 1/20... Training Step: 3456... Training loss: 1.9062... 0.0582 sec/batch
Epoch: 1/20... Training Step: 3457... Training loss: 1.7309... 0.0764 sec/batch
Epoch: 1/20... Training Step: 3458... Training loss: 1.6385... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3459... Training loss: 1.7570... 0.0849 sec/batch
Epoch: 1/20... Training Step: 3460... Training loss: 1.7423... 0.0575 sec/batch
Epoch: 1/20... Training Step: 3461... Training loss: 1.7526... 0.0555 sec/batch
Epoch: 1/20... Training Step: 3462... Training loss: 1.7539... 0.0807 sec/batch
Epoch: 1/20... Training Step: 3463... Training loss: 1.9538... 0.0652 sec/batch
Epoch: 1/20... Training Step: 3464... Training loss: 1.7338... 0.0685 sec/batch
Epoch: 1/20... Training Step: 3465... Training loss: 1.6961... 0.0629 sec/batch
Epoch: 1/20... Training Step: 3466... Training loss: 1.5342... 0.0678 sec/batch
Epoch: 1/20... Training Step: 3467... Training loss: 1.6207... 0.0833 sec/batch
Epoch: 1/20... Training Step: 3468... Training loss: 1.7577... 0.0640 sec/batch
Epoch: 1/20... Training Step: 3469... Training loss: 1.9076... 0.0642 sec/batch
Epoch: 1/20... Training Step: 3470... Training loss: 1.8555... 0.0834 sec/batch
Epoch: 1/20... Training Step: 3471... Training loss: 1.7785... 0.0772 sec/batch
Epoch: 1/20... Training Step: 3472... Training loss: 1.7797... 0.0585 sec/batch
Epoch: 1/20... Training Step: 3473... Training loss: 1.8932... 0.0585 sec/batch
Epoch: 1/20... Training Step: 3474... Training loss: 1.7931... 0.0622 sec/batch
Epoch: 1/20... Training Step: 3475... Training loss: 1.8110... 0.0619 sec/batch
Epoch: 1/20... Training Step: 3476... Training loss: 1.6630... 0.0634 sec/batch
Epoch: 1/20... Training Step: 3477... Training loss: 1.8852... 0.0725 sec/batch
Epoch: 1/20... Training Step: 3478... Training loss: 1.7921... 0.0571 sec/batch
Epoch: 1/20... Training Step: 3479... Training loss: 1.8041... 0.0534 sec/batch
Epoch: 1/20... Training Step: 3480... Training loss: 1.6541... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3481... Training loss: 1.7898... 0.0758 sec/batch
Epoch: 1/20... Training Step: 3482... Training loss: 1.7213... 0.0567 sec/batch
Epoch: 1/20... Training Step: 3483... Training loss: 1.9032... 0.0585 sec/batch
Epoch: 1/20... Training Step: 3484... Training loss: 1.8885... 0.0727 sec/batch
Epoch: 1/20... Training Step: 3485... Training loss: 1.7715... 0.0721 sec/batch
Epoch: 1/20... Training Step: 3486... Training loss: 1.9840... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3487... Training loss: 1.7802... 0.0605 sec/batch
Epoch: 1/20... Training Step: 3488... Training loss: 1.9637... 0.0777 sec/batch
Epoch: 1/20... Training Step: 3489... Training loss: 1.7871... 0.0542 sec/batch
Epoch: 1/20... Training Step: 3490... Training loss: 1.9412... 0.0692 sec/batch
Epoch: 1/20... Training Step: 3491... Training loss: 1.9477... 0.0611 sec/batch
Epoch: 1/20... Training Step: 3492... Training loss: 1.7388... 0.0584 sec/batch
Epoch: 1/20... Training Step: 3493... Training loss: 1.7047... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3494... Training loss: 1.7901... 0.0738 sec/batch
Epoch: 1/20... Training Step: 3495... Training loss: 1.8323... 0.0609 sec/batch
Epoch: 1/20... Training Step: 3496... Training loss: 1.7776... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3497... Training loss: 1.7952... 0.0942 sec/batch
Epoch: 1/20... Training Step: 3498... Training loss: 1.7957... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3499... Training loss: 1.7541... 0.0698 sec/batch
Epoch: 1/20... Training Step: 3500... Training loss: 1.7639... 0.0782 sec/batch
Epoch: 1/20... Training Step: 3501... Training loss: 1.6664... 0.0846 sec/batch
Epoch: 1/20... Training Step: 3502... Training loss: 1.7673... 0.0664 sec/batch
Epoch: 1/20... Training Step: 3503... Training loss: 1.6735... 0.0883 sec/batch
Epoch: 1/20... Training Step: 3504... Training loss: 1.7317... 0.0760 sec/batch
Epoch: 1/20... Training Step: 3505... Training loss: 1.7488... 0.0685 sec/batch
Epoch: 1/20... Training Step: 3506... Training loss: 1.7435... 0.0729 sec/batch
Epoch: 1/20... Training Step: 3507... Training loss: 1.6988... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3508... Training loss: 1.7373... 0.0537 sec/batch
Epoch: 1/20... Training Step: 3509... Training loss: 1.7690... 0.0552 sec/batch
Epoch: 1/20... Training Step: 3510... Training loss: 1.8210... 0.0778 sec/batch
Epoch: 1/20... Training Step: 3511... Training loss: 1.8428... 0.0641 sec/batch
Epoch: 1/20... Training Step: 3512... Training loss: 1.7011... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3513... Training loss: 1.7994... 0.1053 sec/batch
Epoch: 1/20... Training Step: 3514... Training loss: 1.7034... 0.0769 sec/batch
Epoch: 1/20... Training Step: 3515... Training loss: 1.6603... 0.0854 sec/batch
Epoch: 1/20... Training Step: 3516... Training loss: 1.8397... 0.0951 sec/batch
Epoch: 1/20... Training Step: 3517... Training loss: 1.7570... 0.0640 sec/batch
Epoch: 1/20... Training Step: 3518... Training loss: 1.8426... 0.0759 sec/batch
Epoch: 1/20... Training Step: 3519... Training loss: 1.8886... 0.0780 sec/batch
Epoch: 1/20... Training Step: 3520... Training loss: 1.7220... 0.0722 sec/batch
Epoch: 1/20... Training Step: 3521... Training loss: 1.8239... 0.0778 sec/batch
Epoch: 1/20... Training Step: 3522... Training loss: 1.9045... 0.1091 sec/batch
Epoch: 1/20... Training Step: 3523... Training loss: 1.7576... 0.0743 sec/batch
Epoch: 1/20... Training Step: 3524... Training loss: 1.7553... 0.0831 sec/batch
Epoch: 1/20... Training Step: 3525... Training loss: 1.5236... 0.0662 sec/batch
Epoch: 1/20... Training Step: 3526... Training loss: 1.7526... 0.0645 sec/batch
Epoch: 1/20... Training Step: 3527... Training loss: 1.7481... 0.0636 sec/batch
Epoch: 1/20... Training Step: 3528... Training loss: 1.8158... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3529... Training loss: 1.7174... 0.0806 sec/batch
Epoch: 1/20... Training Step: 3530... Training loss: 1.6189... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3531... Training loss: 1.6281... 0.0609 sec/batch
Epoch: 1/20... Training Step: 3532... Training loss: 1.8025... 0.0782 sec/batch
Epoch: 1/20... Training Step: 3533... Training loss: 1.8088... 0.0885 sec/batch
Epoch: 1/20... Training Step: 3534... Training loss: 1.8164... 0.0666 sec/batch
Epoch: 1/20... Training Step: 3535... Training loss: 1.8027... 0.1018 sec/batch
Epoch: 1/20... Training Step: 3536... Training loss: 1.6863... 0.0739 sec/batch
Epoch: 1/20... Training Step: 3537... Training loss: 1.6680... 0.0708 sec/batch
Epoch: 1/20... Training Step: 3538... Training loss: 1.8001... 0.0884 sec/batch
Epoch: 1/20... Training Step: 3539... Training loss: 1.9049... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3540... Training loss: 1.7532... 0.0766 sec/batch
Epoch: 1/20... Training Step: 3541... Training loss: 1.8589... 0.0680 sec/batch
Epoch: 1/20... Training Step: 3542... Training loss: 1.7611... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3543... Training loss: 1.6297... 0.0721 sec/batch
Epoch: 1/20... Training Step: 3544... Training loss: 1.7633... 0.0934 sec/batch
Epoch: 1/20... Training Step: 3545... Training loss: 1.7837... 0.0789 sec/batch
Epoch: 1/20... Training Step: 3546... Training loss: 1.6311... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3547... Training loss: 1.7642... 0.0829 sec/batch
Epoch: 1/20... Training Step: 3548... Training loss: 1.6120... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3549... Training loss: 1.7170... 0.0634 sec/batch
Epoch: 1/20... Training Step: 3550... Training loss: 1.8266... 0.0967 sec/batch
Epoch: 1/20... Training Step: 3551... Training loss: 1.7822... 0.0739 sec/batch
Epoch: 1/20... Training Step: 3552... Training loss: 1.7908... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3553... Training loss: 1.8224... 0.0766 sec/batch
Epoch: 1/20... Training Step: 3554... Training loss: 1.8241... 0.0865 sec/batch
Epoch: 1/20... Training Step: 3555... Training loss: 1.6457... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3556... Training loss: 1.9273... 0.0838 sec/batch
Epoch: 1/20... Training Step: 3557... Training loss: 1.6476... 0.0918 sec/batch
Epoch: 1/20... Training Step: 3558... Training loss: 1.6215... 0.0854 sec/batch
Epoch: 1/20... Training Step: 3559... Training loss: 1.8549... 0.0886 sec/batch
Epoch: 1/20... Training Step: 3560... Training loss: 1.8219... 0.0652 sec/batch
Epoch: 1/20... Training Step: 3561... Training loss: 1.6869... 0.0776 sec/batch
Epoch: 1/20... Training Step: 3562... Training loss: 1.7534... 0.0917 sec/batch
Epoch: 1/20... Training Step: 3563... Training loss: 1.8246... 0.0732 sec/batch
Epoch: 1/20... Training Step: 3564... Training loss: 1.7645... 0.1022 sec/batch
Epoch: 1/20... Training Step: 3565... Training loss: 1.7936... 0.1406 sec/batch
Epoch: 1/20... Training Step: 3566... Training loss: 1.8223... 0.1474 sec/batch
Epoch: 1/20... Training Step: 3567... Training loss: 1.8067... 0.1362 sec/batch
Epoch: 1/20... Training Step: 3568... Training loss: 1.7793... 0.1275 sec/batch
Epoch: 1/20... Training Step: 3569... Training loss: 1.8096... 0.0985 sec/batch
Epoch: 1/20... Training Step: 3570... Training loss: 1.5600... 0.0994 sec/batch
Epoch: 1/20... Training Step: 3571... Training loss: 1.8318... 0.1301 sec/batch
Epoch: 1/20... Training Step: 3572... Training loss: 1.9306... 0.1043 sec/batch
Epoch: 1/20... Training Step: 3573... Training loss: 1.8055... 0.0969 sec/batch
Epoch: 1/20... Training Step: 3574... Training loss: 1.7678... 0.1210 sec/batch
Epoch: 1/20... Training Step: 3575... Training loss: 1.5940... 0.1116 sec/batch
Epoch: 1/20... Training Step: 3576... Training loss: 1.7114... 0.0944 sec/batch
Epoch: 1/20... Training Step: 3577... Training loss: 1.8123... 0.1176 sec/batch
Epoch: 1/20... Training Step: 3578... Training loss: 1.6185... 0.1424 sec/batch
Epoch: 1/20... Training Step: 3579... Training loss: 1.6640... 0.1097 sec/batch
Epoch: 1/20... Training Step: 3580... Training loss: 1.6953... 0.1038 sec/batch
Epoch: 1/20... Training Step: 3581... Training loss: 1.7551... 0.1151 sec/batch
Epoch: 1/20... Training Step: 3582... Training loss: 1.7502... 0.0896 sec/batch
Epoch: 1/20... Training Step: 3583... Training loss: 1.6563... 0.0961 sec/batch
Epoch: 1/20... Training Step: 3584... Training loss: 1.8196... 0.1013 sec/batch
Epoch: 1/20... Training Step: 3585... Training loss: 1.7411... 0.1032 sec/batch
Epoch: 1/20... Training Step: 3586... Training loss: 1.9102... 0.1098 sec/batch
Epoch: 1/20... Training Step: 3587... Training loss: 1.8784... 0.1425 sec/batch
Epoch: 1/20... Training Step: 3588... Training loss: 1.7800... 0.1467 sec/batch
Epoch: 1/20... Training Step: 3589... Training loss: 1.6772... 0.1308 sec/batch
Epoch: 1/20... Training Step: 3590... Training loss: 1.7578... 0.1336 sec/batch
Epoch: 1/20... Training Step: 3591... Training loss: 1.7398... 0.1250 sec/batch
Epoch: 1/20... Training Step: 3592... Training loss: 1.6305... 0.1308 sec/batch
Epoch: 1/20... Training Step: 3593... Training loss: 1.7529... 0.1284 sec/batch
Epoch: 1/20... Training Step: 3594... Training loss: 2.5630... 0.1185 sec/batch
Epoch: 1/20... Training Step: 3595... Training loss: 2.0040... 0.1418 sec/batch
Epoch: 1/20... Training Step: 3596... Training loss: 1.7951... 0.1104 sec/batch
Epoch: 1/20... Training Step: 3597... Training loss: 2.0068... 0.1210 sec/batch
Epoch: 1/20... Training Step: 3598... Training loss: 1.8672... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3599... Training loss: 1.8685... 0.0714 sec/batch
Epoch: 1/20... Training Step: 3600... Training loss: 1.7882... 0.1012 sec/batch
Epoch: 1/20... Training Step: 3601... Training loss: 1.8894... 0.0713 sec/batch
Epoch: 1/20... Training Step: 3602... Training loss: 1.8179... 0.0673 sec/batch
Epoch: 1/20... Training Step: 3603... Training loss: 1.9386... 0.0878 sec/batch
Epoch: 1/20... Training Step: 3604... Training loss: 1.9423... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3605... Training loss: 1.8742... 0.0745 sec/batch
Epoch: 1/20... Training Step: 3606... Training loss: 1.8030... 0.0781 sec/batch
Epoch: 1/20... Training Step: 3607... Training loss: 1.8319... 0.0599 sec/batch
Epoch: 1/20... Training Step: 3608... Training loss: 1.7934... 0.0950 sec/batch
Epoch: 1/20... Training Step: 3609... Training loss: 1.9130... 0.0859 sec/batch
Epoch: 1/20... Training Step: 3610... Training loss: 1.8522... 0.0763 sec/batch
Epoch: 1/20... Training Step: 3611... Training loss: 1.8882... 0.0803 sec/batch
Epoch: 1/20... Training Step: 3612... Training loss: 1.7880... 0.0706 sec/batch
Epoch: 1/20... Training Step: 3613... Training loss: 1.7429... 0.0694 sec/batch
Epoch: 1/20... Training Step: 3614... Training loss: 1.7829... 0.0772 sec/batch
Epoch: 1/20... Training Step: 3615... Training loss: 1.7986... 0.0820 sec/batch
Epoch: 1/20... Training Step: 3616... Training loss: 1.7813... 0.0788 sec/batch
Epoch: 1/20... Training Step: 3617... Training loss: 1.7932... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3618... Training loss: 1.8723... 0.0786 sec/batch
Epoch: 1/20... Training Step: 3619... Training loss: 1.8410... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3620... Training loss: 1.7521... 0.0854 sec/batch
Epoch: 1/20... Training Step: 3621... Training loss: 1.7726... 0.0763 sec/batch
Epoch: 1/20... Training Step: 3622... Training loss: 1.8930... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3623... Training loss: 1.8591... 0.0675 sec/batch
Epoch: 1/20... Training Step: 3624... Training loss: 1.9620... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3625... Training loss: 1.7429... 0.0742 sec/batch
Epoch: 1/20... Training Step: 3626... Training loss: 1.8284... 0.0745 sec/batch
Epoch: 1/20... Training Step: 3627... Training loss: 2.0532... 0.0882 sec/batch
Epoch: 1/20... Training Step: 3628... Training loss: 1.7522... 0.0729 sec/batch
Epoch: 1/20... Training Step: 3629... Training loss: 1.9196... 0.0663 sec/batch
Epoch: 1/20... Training Step: 3630... Training loss: 1.7682... 0.0692 sec/batch
Epoch: 1/20... Training Step: 3631... Training loss: 1.8708... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3632... Training loss: 1.8801... 0.0742 sec/batch
Epoch: 1/20... Training Step: 3633... Training loss: 1.8873... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3634... Training loss: 1.8817... 0.0679 sec/batch
Epoch: 1/20... Training Step: 3635... Training loss: 1.8069... 0.0752 sec/batch
Epoch: 1/20... Training Step: 3636... Training loss: 1.9118... 0.0924 sec/batch
Epoch: 1/20... Training Step: 3637... Training loss: 1.7662... 0.0698 sec/batch
Epoch: 1/20... Training Step: 3638... Training loss: 1.8694... 0.0718 sec/batch
Epoch: 1/20... Training Step: 3639... Training loss: 1.8670... 0.0814 sec/batch
Epoch: 1/20... Training Step: 3640... Training loss: 1.7286... 0.0620 sec/batch
Epoch: 1/20... Training Step: 3641... Training loss: 1.9074... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3642... Training loss: 1.9077... 0.0910 sec/batch
Epoch: 1/20... Training Step: 3643... Training loss: 2.0165... 0.0705 sec/batch
Epoch: 1/20... Training Step: 3644... Training loss: 1.9518... 0.0787 sec/batch
Epoch: 1/20... Training Step: 3645... Training loss: 1.8546... 0.0740 sec/batch
Epoch: 1/20... Training Step: 3646... Training loss: 1.6605... 0.0513 sec/batch
Epoch: 1/20... Training Step: 3647... Training loss: 1.8330... 0.0885 sec/batch
Epoch: 1/20... Training Step: 3648... Training loss: 1.9189... 0.0818 sec/batch
Epoch: 1/20... Training Step: 3649... Training loss: 1.7036... 0.0613 sec/batch
Epoch: 1/20... Training Step: 3650... Training loss: 1.8936... 0.0699 sec/batch
Epoch: 1/20... Training Step: 3651... Training loss: 1.9064... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3652... Training loss: 1.7618... 0.0632 sec/batch
Epoch: 1/20... Training Step: 3653... Training loss: 1.8584... 0.0687 sec/batch
Epoch: 1/20... Training Step: 3654... Training loss: 1.7715... 0.0741 sec/batch
Epoch: 1/20... Training Step: 3655... Training loss: 2.0877... 0.0584 sec/batch
Epoch: 1/20... Training Step: 3656... Training loss: 2.0216... 0.0676 sec/batch
Epoch: 1/20... Training Step: 3657... Training loss: 1.8035... 0.0666 sec/batch
Epoch: 1/20... Training Step: 3658... Training loss: 1.9346... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3659... Training loss: 1.8334... 0.0933 sec/batch
Epoch: 1/20... Training Step: 3660... Training loss: 1.8809... 0.0761 sec/batch
Epoch: 1/20... Training Step: 3661... Training loss: 1.9048... 0.0615 sec/batch
Epoch: 1/20... Training Step: 3662... Training loss: 1.8227... 0.0548 sec/batch
Epoch: 1/20... Training Step: 3663... Training loss: 1.8961... 0.0571 sec/batch
Epoch: 1/20... Training Step: 3664... Training loss: 1.9391... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3665... Training loss: 1.7327... 0.0793 sec/batch
Epoch: 1/20... Training Step: 3666... Training loss: 1.8797... 0.0553 sec/batch
Epoch: 1/20... Training Step: 3667... Training loss: 1.7492... 0.0810 sec/batch
Epoch: 1/20... Training Step: 3668... Training loss: 1.9457... 0.0841 sec/batch
Epoch: 1/20... Training Step: 3669... Training loss: 1.9283... 0.0584 sec/batch
Epoch: 1/20... Training Step: 3670... Training loss: 1.8474... 0.0515 sec/batch
Epoch: 1/20... Training Step: 3671... Training loss: 1.7497... 0.0657 sec/batch
Epoch: 1/20... Training Step: 3672... Training loss: 1.8259... 0.0696 sec/batch
Epoch: 1/20... Training Step: 3673... Training loss: 1.8272... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3674... Training loss: 1.8159... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3675... Training loss: 1.7296... 0.0695 sec/batch
Epoch: 1/20... Training Step: 3676... Training loss: 1.9573... 0.0689 sec/batch
Epoch: 1/20... Training Step: 3677... Training loss: 1.8552... 0.0598 sec/batch
Epoch: 1/20... Training Step: 3678... Training loss: 1.8406... 0.0526 sec/batch
Epoch: 1/20... Training Step: 3679... Training loss: 1.9302... 0.0703 sec/batch
Epoch: 1/20... Training Step: 3680... Training loss: 1.9235... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3681... Training loss: 1.8509... 0.0685 sec/batch
Epoch: 1/20... Training Step: 3682... Training loss: 1.8549... 0.0580 sec/batch
Epoch: 1/20... Training Step: 3683... Training loss: 1.7302... 0.0622 sec/batch
Epoch: 1/20... Training Step: 3684... Training loss: 1.7011... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3685... Training loss: 1.8208... 0.0556 sec/batch
Epoch: 1/20... Training Step: 3686... Training loss: 1.8830... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3687... Training loss: 1.8839... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3688... Training loss: 1.7697... 0.0667 sec/batch
Epoch: 1/20... Training Step: 3689... Training loss: 1.8214... 0.0538 sec/batch
Epoch: 1/20... Training Step: 3690... Training loss: 1.7613... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3691... Training loss: 1.9359... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3692... Training loss: 1.8445... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3693... Training loss: 1.9670... 0.0535 sec/batch
Epoch: 1/20... Training Step: 3694... Training loss: 1.8432... 0.0625 sec/batch
Epoch: 1/20... Training Step: 3695... Training loss: 1.8841... 0.0739 sec/batch
Epoch: 1/20... Training Step: 3696... Training loss: 1.8950... 0.0707 sec/batch
Epoch: 1/20... Training Step: 3697... Training loss: 1.7923... 0.0506 sec/batch
Epoch: 1/20... Training Step: 3698... Training loss: 1.8201... 0.0785 sec/batch
Epoch: 1/20... Training Step: 3699... Training loss: 1.8195... 0.0650 sec/batch
Epoch: 1/20... Training Step: 3700... Training loss: 1.7394... 0.0644 sec/batch
Epoch: 1/20... Training Step: 3701... Training loss: 1.9634... 0.0627 sec/batch
Epoch: 1/20... Training Step: 3702... Training loss: 1.8690... 0.0697 sec/batch
Epoch: 1/20... Training Step: 3703... Training loss: 1.7379... 0.0803 sec/batch
Epoch: 1/20... Training Step: 3704... Training loss: 1.9797... 0.0584 sec/batch
Epoch: 1/20... Training Step: 3705... Training loss: 1.7213... 0.0558 sec/batch
Epoch: 1/20... Training Step: 3706... Training loss: 1.6883... 0.0910 sec/batch
Epoch: 1/20... Training Step: 3707... Training loss: 1.7251... 0.0711 sec/batch
Epoch: 1/20... Training Step: 3708... Training loss: 1.8414... 0.0606 sec/batch
Epoch: 1/20... Training Step: 3709... Training loss: 1.7969... 0.0614 sec/batch
Epoch: 1/20... Training Step: 3710... Training loss: 1.7761... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3711... Training loss: 1.8279... 0.0631 sec/batch
Epoch: 1/20... Training Step: 3712... Training loss: 2.0235... 0.0508 sec/batch
Epoch: 1/20... Training Step: 3713... Training loss: 1.8362... 0.0523 sec/batch
Epoch: 1/20... Training Step: 3714... Training loss: 1.9537... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3715... Training loss: 2.0968... 0.0681 sec/batch
Epoch: 1/20... Training Step: 3716... Training loss: 1.7888... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3717... Training loss: 1.8176... 0.0600 sec/batch
Epoch: 1/20... Training Step: 3718... Training loss: 1.8110... 0.0603 sec/batch
Epoch: 1/20... Training Step: 3719... Training loss: 1.8213... 0.0632 sec/batch
Epoch: 1/20... Training Step: 3720... Training loss: 1.7319... 0.0550 sec/batch
Epoch: 1/20... Training Step: 3721... Training loss: 1.7564... 0.0522 sec/batch
Epoch: 1/20... Training Step: 3722... Training loss: 1.6845... 0.0610 sec/batch
Epoch: 1/20... Training Step: 3723... Training loss: 1.9546... 0.0715 sec/batch
Epoch: 1/20... Training Step: 3724... Training loss: 1.9120... 0.0648 sec/batch
Epoch: 1/20... Training Step: 3725... Training loss: 1.9032... 0.0535 sec/batch
Epoch: 1/20... Training Step: 3726... Training loss: 1.8900... 0.0624 sec/batch
Epoch: 1/20... Training Step: 3727... Training loss: 1.9688... 0.0626 sec/batch
Epoch: 1/20... Training Step: 3728... Training loss: 1.9777... 0.0554 sec/batch
Epoch: 1/20... Training Step: 3729... Training loss: 2.0529... 0.0611 sec/batch
Epoch: 1/20... Training Step: 3730... Training loss: 1.8946... 0.0544 sec/batch
Epoch: 1/20... Training Step: 3731... Training loss: 1.7953... 0.0688 sec/batch
Epoch: 1/20... Training Step: 3732... Training loss: 1.8960... 0.0617 sec/batch
Epoch: 1/20... Training Step: 3733... Training loss: 1.7743... 0.0600 sec/batch
Epoch: 1/20... Training Step: 3734... Training loss: 1.6978... 0.0606 sec/batch
Epoch: 1/20... Training Step: 3735... Training loss: 1.8484... 0.0630 sec/batch
Epoch: 1/20... Training Step: 3736... Training loss: 1.9021... 0.0624 sec/batch
Epoch: 1/20... Training Step: 3737... Training loss: 1.7926... 0.0506 sec/batch
Epoch: 1/20... Training Step: 3738... Training loss: 2.0099... 0.0515 sec/batch
Epoch: 1/20... Training Step: 3739... Training loss: 2.0414... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3740... Training loss: 1.8174... 0.0728 sec/batch
Epoch: 1/20... Training Step: 3741... Training loss: 1.9105... 0.0550 sec/batch
Epoch: 1/20... Training Step: 3742... Training loss: 1.9923... 0.0610 sec/batch
Epoch: 1/20... Training Step: 3743... Training loss: 1.9113... 0.0607 sec/batch
Epoch: 1/20... Training Step: 3744... Training loss: 1.9597... 0.0693 sec/batch
Epoch: 1/20... Training Step: 3745... Training loss: 1.7888... 0.0764 sec/batch
Epoch: 1/20... Training Step: 3746... Training loss: 2.0286... 0.0663 sec/batch
Epoch: 1/20... Training Step: 3747... Training loss: 1.8477... 0.0669 sec/batch
Epoch: 1/20... Training Step: 3748... Training loss: 2.0674... 0.0728 sec/batch
Epoch: 1/20... Training Step: 3749... Training loss: 1.7969... 0.0759 sec/batch
Epoch: 1/20... Training Step: 3750... Training loss: 1.8764... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3751... Training loss: 1.8008... 0.0587 sec/batch
Epoch: 1/20... Training Step: 3752... Training loss: 2.0408... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3753... Training loss: 1.7813... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3754... Training loss: 1.8931... 0.0587 sec/batch
Epoch: 1/20... Training Step: 3755... Training loss: 1.8649... 0.0699 sec/batch
Epoch: 1/20... Training Step: 3756... Training loss: 1.9127... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3757... Training loss: 1.8473... 0.0503 sec/batch
Epoch: 1/20... Training Step: 3758... Training loss: 2.0442... 0.0727 sec/batch
Epoch: 1/20... Training Step: 3759... Training loss: 1.8586... 0.0605 sec/batch
Epoch: 1/20... Training Step: 3760... Training loss: 1.7833... 0.0621 sec/batch
Epoch: 1/20... Training Step: 3761... Training loss: 1.7962... 0.0550 sec/batch
Epoch: 1/20... Training Step: 3762... Training loss: 1.7433... 0.0693 sec/batch
Epoch: 1/20... Training Step: 3763... Training loss: 1.7623... 0.0751 sec/batch
Epoch: 1/20... Training Step: 3764... Training loss: 1.8390... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3765... Training loss: 1.7947... 0.0529 sec/batch
Epoch: 1/20... Training Step: 3766... Training loss: 1.8695... 0.0648 sec/batch
Epoch: 1/20... Training Step: 3767... Training loss: 1.8459... 0.0533 sec/batch
Epoch: 1/20... Training Step: 3768... Training loss: 1.9249... 0.0647 sec/batch
Epoch: 1/20... Training Step: 3769... Training loss: 1.9542... 0.0588 sec/batch
Epoch: 1/20... Training Step: 3770... Training loss: 1.7790... 0.0676 sec/batch
Epoch: 1/20... Training Step: 3771... Training loss: 1.7846... 0.0647 sec/batch
Epoch: 1/20... Training Step: 3772... Training loss: 1.7418... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3773... Training loss: 1.8851... 0.0569 sec/batch
Epoch: 1/20... Training Step: 3774... Training loss: 1.8309... 0.0703 sec/batch
Epoch: 1/20... Training Step: 3775... Training loss: 1.9051... 0.0742 sec/batch
Epoch: 1/20... Training Step: 3776... Training loss: 1.8417... 0.0660 sec/batch
Epoch: 1/20... Training Step: 3777... Training loss: 1.8029... 0.0651 sec/batch
Epoch: 1/20... Training Step: 3778... Training loss: 1.8152... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3779... Training loss: 1.8479... 0.0726 sec/batch
Epoch: 1/20... Training Step: 3780... Training loss: 2.0154... 0.0657 sec/batch
Epoch: 1/20... Training Step: 3781... Training loss: 1.7807... 0.0693 sec/batch
Epoch: 1/20... Training Step: 3782... Training loss: 1.6863... 0.0636 sec/batch
Epoch: 1/20... Training Step: 3783... Training loss: 1.7721... 0.0591 sec/batch
Epoch: 1/20... Training Step: 3784... Training loss: 1.7609... 0.1007 sec/batch
Epoch: 1/20... Training Step: 3785... Training loss: 1.7400... 0.0820 sec/batch
Epoch: 1/20... Training Step: 3786... Training loss: 1.7264... 0.0611 sec/batch
Epoch: 1/20... Training Step: 3787... Training loss: 1.8485... 0.0606 sec/batch
Epoch: 1/20... Training Step: 3788... Training loss: 1.7757... 0.0675 sec/batch
Epoch: 1/20... Training Step: 3789... Training loss: 1.6907... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3790... Training loss: 1.7385... 0.0568 sec/batch
Epoch: 1/20... Training Step: 3791... Training loss: 1.6702... 0.0627 sec/batch
Epoch: 1/20... Training Step: 3792... Training loss: 1.8305... 0.0750 sec/batch
Epoch: 1/20... Training Step: 3793... Training loss: 1.8352... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3794... Training loss: 2.0064... 0.0615 sec/batch
Epoch: 1/20... Training Step: 3795... Training loss: 1.9566... 0.0647 sec/batch
Epoch: 1/20... Training Step: 3796... Training loss: 1.9834... 0.0688 sec/batch
Epoch: 1/20... Training Step: 3797... Training loss: 2.0453... 0.0536 sec/batch
Epoch: 1/20... Training Step: 3798... Training loss: 2.0216... 0.0699 sec/batch
Epoch: 1/20... Training Step: 3799... Training loss: 1.9186... 0.0765 sec/batch
Epoch: 1/20... Training Step: 3800... Training loss: 1.7780... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3801... Training loss: 1.9471... 0.0651 sec/batch
Epoch: 1/20... Training Step: 3802... Training loss: 1.6978... 0.0780 sec/batch
Epoch: 1/20... Training Step: 3803... Training loss: 1.7885... 0.0554 sec/batch
Epoch: 1/20... Training Step: 3804... Training loss: 1.7225... 0.0616 sec/batch
Epoch: 1/20... Training Step: 3805... Training loss: 1.8585... 0.0598 sec/batch
Epoch: 1/20... Training Step: 3806... Training loss: 1.6978... 0.0919 sec/batch
Epoch: 1/20... Training Step: 3807... Training loss: 1.9115... 0.0497 sec/batch
Epoch: 1/20... Training Step: 3808... Training loss: 2.3265... 0.0503 sec/batch
Epoch: 1/20... Training Step: 3809... Training loss: 1.8770... 0.0741 sec/batch
Epoch: 1/20... Training Step: 3810... Training loss: 1.8882... 0.0703 sec/batch
Epoch: 1/20... Training Step: 3811... Training loss: 1.6397... 0.0726 sec/batch
Epoch: 1/20... Training Step: 3812... Training loss: 1.8988... 0.0557 sec/batch
Epoch: 1/20... Training Step: 3813... Training loss: 1.6576... 0.0675 sec/batch
Epoch: 1/20... Training Step: 3814... Training loss: 1.7245... 0.0622 sec/batch
Epoch: 1/20... Training Step: 3815... Training loss: 2.0095... 0.0725 sec/batch
Epoch: 1/20... Training Step: 3816... Training loss: 2.3919... 0.0616 sec/batch
Epoch: 1/20... Training Step: 3817... Training loss: 2.4775... 0.0899 sec/batch
Epoch: 1/20... Training Step: 3818... Training loss: 2.3457... 0.0724 sec/batch
Epoch: 1/20... Training Step: 3819... Training loss: 2.3585... 0.0617 sec/batch
Epoch: 1/20... Training Step: 3820... Training loss: 2.3632... 0.0606 sec/batch
Epoch: 1/20... Training Step: 3821... Training loss: 2.1159... 0.0713 sec/batch
Epoch: 1/20... Training Step: 3822... Training loss: 2.1006... 0.0640 sec/batch
Epoch: 1/20... Training Step: 3823... Training loss: 2.1884... 0.0603 sec/batch
Epoch: 1/20... Training Step: 3824... Training loss: 2.0008... 0.0496 sec/batch
Epoch: 1/20... Training Step: 3825... Training loss: 1.9714... 0.0673 sec/batch
Epoch: 1/20... Training Step: 3826... Training loss: 1.8131... 0.0722 sec/batch
Epoch: 1/20... Training Step: 3827... Training loss: 1.7250... 0.0792 sec/batch
Epoch: 1/20... Training Step: 3828... Training loss: 1.8356... 0.0588 sec/batch
Epoch: 1/20... Training Step: 3829... Training loss: 1.7890... 0.0850 sec/batch
Epoch: 1/20... Training Step: 3830... Training loss: 1.9500... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3831... Training loss: 1.7584... 0.0538 sec/batch
Epoch: 1/20... Training Step: 3832... Training loss: 1.6892... 0.0560 sec/batch
Epoch: 1/20... Training Step: 3833... Training loss: 1.7826... 0.0614 sec/batch
Epoch: 1/20... Training Step: 3834... Training loss: 1.6501... 0.0535 sec/batch
Epoch: 1/20... Training Step: 3835... Training loss: 1.9138... 0.0661 sec/batch
Epoch: 1/20... Training Step: 3836... Training loss: 1.7722... 0.0643 sec/batch
Epoch: 1/20... Training Step: 3837... Training loss: 1.8927... 0.0587 sec/batch
Epoch: 1/20... Training Step: 3838... Training loss: 1.9136... 0.0627 sec/batch
Epoch: 1/20... Training Step: 3839... Training loss: 1.9316... 0.0771 sec/batch
Epoch: 1/20... Training Step: 3840... Training loss: 1.9593... 0.0638 sec/batch
Epoch: 1/20... Training Step: 3841... Training loss: 1.8713... 0.0604 sec/batch
Epoch: 1/20... Training Step: 3842... Training loss: 1.9804... 0.0513 sec/batch
Epoch: 1/20... Training Step: 3843... Training loss: 2.0932... 0.0545 sec/batch
Epoch: 1/20... Training Step: 3844... Training loss: 2.0895... 0.0575 sec/batch
Epoch: 1/20... Training Step: 3845... Training loss: 1.9196... 0.0685 sec/batch
Epoch: 1/20... Training Step: 3846... Training loss: 1.8948... 0.0515 sec/batch
Epoch: 1/20... Training Step: 3847... Training loss: 1.8356... 0.0572 sec/batch
Epoch: 1/20... Training Step: 3848... Training loss: 1.9448... 0.0578 sec/batch
Epoch: 1/20... Training Step: 3849... Training loss: 1.7447... 0.0776 sec/batch
Epoch: 1/20... Training Step: 3850... Training loss: 1.7274... 0.0580 sec/batch
Epoch: 1/20... Training Step: 3851... Training loss: 1.9376... 0.0670 sec/batch
Epoch: 1/20... Training Step: 3852... Training loss: 1.9651... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3853... Training loss: 1.9886... 0.0589 sec/batch
Epoch: 1/20... Training Step: 3854... Training loss: 1.6966... 0.0540 sec/batch
Epoch: 1/20... Training Step: 3855... Training loss: 1.9112... 0.0666 sec/batch
Epoch: 1/20... Training Step: 3856... Training loss: 1.7071... 0.0618 sec/batch
Epoch: 1/20... Training Step: 3857... Training loss: 1.9656... 0.0657 sec/batch
Epoch: 1/20... Training Step: 3858... Training loss: 1.7676... 0.0669 sec/batch
Epoch: 1/20... Training Step: 3859... Training loss: 1.9230... 0.0570 sec/batch
Epoch: 1/20... Training Step: 3860... Training loss: 1.8759... 0.0644 sec/batch
Epoch: 1/20... Training Step: 3861... Training loss: 1.9711... 0.0767 sec/batch
Epoch: 1/20... Training Step: 3862... Training loss: 1.7072... 0.0493 sec/batch
Epoch: 1/20... Training Step: 3863... Training loss: 1.8114... 0.0744 sec/batch
Epoch: 1/20... Training Step: 3864... Training loss: 1.8477... 0.0708 sec/batch
Epoch: 1/20... Training Step: 3865... Training loss: 1.8886... 0.0600 sec/batch
Epoch: 1/20... Training Step: 3866... Training loss: 1.8525... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3867... Training loss: 1.8604... 0.0679 sec/batch
Epoch: 1/20... Training Step: 3868... Training loss: 1.7930... 0.0706 sec/batch
Epoch: 1/20... Training Step: 3869... Training loss: 1.8524... 0.0592 sec/batch
Epoch: 1/20... Training Step: 3870... Training loss: 1.9227... 0.0504 sec/batch
Epoch: 1/20... Training Step: 3871... Training loss: 1.8432... 0.0694 sec/batch
Epoch: 1/20... Training Step: 3872... Training loss: 1.8754... 0.0638 sec/batch
Epoch: 1/20... Training Step: 3873... Training loss: 1.8419... 0.0555 sec/batch
Epoch: 1/20... Training Step: 3874... Training loss: 1.7693... 0.0632 sec/batch
Epoch: 1/20... Training Step: 3875... Training loss: 1.7043... 0.0698 sec/batch
Epoch: 1/20... Training Step: 3876... Training loss: 1.8046... 0.0659 sec/batch
Epoch: 1/20... Training Step: 3877... Training loss: 1.8899... 0.0652 sec/batch
Epoch: 1/20... Training Step: 3878... Training loss: 1.7654... 0.0534 sec/batch
Epoch: 1/20... Training Step: 3879... Training loss: 1.7559... 0.0668 sec/batch
Epoch: 1/20... Training Step: 3880... Training loss: 1.8250... 0.0725 sec/batch
Epoch: 1/20... Training Step: 3881... Training loss: 1.8715... 0.0677 sec/batch
Epoch: 1/20... Training Step: 3882... Training loss: 1.8962... 0.0598 sec/batch
Epoch: 1/20... Training Step: 3883... Training loss: 1.8153... 0.0707 sec/batch
Epoch: 1/20... Training Step: 3884... Training loss: 1.8070... 0.0642 sec/batch
Epoch: 1/20... Training Step: 3885... Training loss: 1.8422... 0.0686 sec/batch
Epoch: 1/20... Training Step: 3886... Training loss: 1.7893... 0.0832 sec/batch
Epoch: 1/20... Training Step: 3887... Training loss: 1.9077... 0.0723 sec/batch
Epoch: 1/20... Training Step: 3888... Training loss: 2.0209... 0.0581 sec/batch
Epoch: 1/20... Training Step: 3889... Training loss: 1.7927... 0.0678 sec/batch
Epoch: 1/20... Training Step: 3890... Training loss: 1.8872... 0.0660 sec/batch
Epoch: 1/20... Training Step: 3891... Training loss: 1.9245... 0.0658 sec/batch
Epoch: 1/20... Training Step: 3892... Training loss: 1.7975... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3893... Training loss: 1.7800... 0.0674 sec/batch
Epoch: 1/20... Training Step: 3894... Training loss: 1.8762... 0.0710 sec/batch
Epoch: 1/20... Training Step: 3895... Training loss: 1.8238... 0.0806 sec/batch
Epoch: 1/20... Training Step: 3896... Training loss: 2.1215... 0.0859 sec/batch
Epoch: 1/20... Training Step: 3897... Training loss: 1.9337... 0.0653 sec/batch
Epoch: 1/20... Training Step: 3898... Training loss: 1.9137... 0.0743 sec/batch
Epoch: 1/20... Training Step: 3899... Training loss: 1.8827... 0.0656 sec/batch
Epoch: 1/20... Training Step: 3900... Training loss: 1.9048... 0.0607 sec/batch
Epoch: 1/20... Training Step: 3901... Training loss: 1.8316... 0.0663 sec/batch
Epoch: 1/20... Training Step: 3902... Training loss: 2.0715... 0.0673 sec/batch
Epoch: 1/20... Training Step: 3903... Training loss: 1.9869... 0.0678 sec/batch
Epoch: 1/20... Training Step: 3904... Training loss: 1.9871... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3905... Training loss: 2.1139... 0.0682 sec/batch
Epoch: 1/20... Training Step: 3906... Training loss: 2.0387... 0.0741 sec/batch
Epoch: 1/20... Training Step: 3907... Training loss: 1.9318... 0.0658 sec/batch
Epoch: 1/20... Training Step: 3908... Training loss: 1.8539... 0.0779 sec/batch
Epoch: 1/20... Training Step: 3909... Training loss: 2.0551... 0.0716 sec/batch
Epoch: 1/20... Training Step: 3910... Training loss: 2.0234... 0.0828 sec/batch
Epoch: 1/20... Training Step: 3911... Training loss: 2.0200... 0.0698 sec/batch
Epoch: 1/20... Training Step: 3912... Training loss: 1.8895... 0.0829 sec/batch
Epoch: 1/20... Training Step: 3913... Training loss: 1.8658... 0.0621 sec/batch
Epoch: 1/20... Training Step: 3914... Training loss: 1.8775... 0.0684 sec/batch
Epoch: 1/20... Training Step: 3915... Training loss: 1.8271... 0.0721 sec/batch
Epoch: 1/20... Training Step: 3916... Training loss: 1.8069... 0.0626 sec/batch
Epoch: 1/20... Training Step: 3917... Training loss: 1.8519... 0.0847 sec/batch
Epoch: 1/20... Training Step: 3918... Training loss: 1.8182... 0.0703 sec/batch
Epoch: 1/20... Training Step: 3919... Training loss: 1.8438... 0.0766 sec/batch
Epoch: 1/20... Training Step: 3920... Training loss: 2.1099... 0.0783 sec/batch
Epoch: 1/20... Training Step: 3921... Training loss: 1.7545... 0.0587 sec/batch
Epoch: 1/20... Training Step: 3922... Training loss: 1.9374... 0.0786 sec/batch
Epoch: 1/20... Training Step: 3923... Training loss: 1.7993... 0.0725 sec/batch
Epoch: 1/20... Training Step: 3924... Training loss: 1.8407... 0.0551 sec/batch
Epoch: 1/20... Training Step: 3925... Training loss: 1.9588... 0.0743 sec/batch
Epoch: 1/20... Training Step: 3926... Training loss: 1.8703... 0.0685 sec/batch
Epoch: 1/20... Training Step: 3927... Training loss: 1.7958... 0.0655 sec/batch
Epoch: 1/20... Training Step: 3928... Training loss: 1.8019... 0.0672 sec/batch
Epoch: 1/20... Training Step: 3929... Training loss: 1.7853... 0.0665 sec/batch
Epoch: 1/20... Training Step: 3930... Training loss: 1.8786... 0.0728 sec/batch
Epoch: 1/20... Training Step: 3931... Training loss: 1.8193... 0.0712 sec/batch
Epoch: 1/20... Training Step: 3932... Training loss: 2.0980... 0.0613 sec/batch
Epoch: 1/20... Training Step: 3933... Training loss: 2.0398... 0.0669 sec/batch
Epoch: 1/20... Training Step: 3934... Training loss: 1.7855... 0.0902 sec/batch
Epoch: 1/20... Training Step: 3935... Training loss: 1.7794... 0.0911 sec/batch
Epoch: 1/20... Training Step: 3936... Training loss: 1.7549... 0.0628 sec/batch
Epoch: 1/20... Training Step: 3937... Training loss: 1.8776... 0.0601 sec/batch
Epoch: 1/20... Training Step: 3938... Training loss: 1.8125... 0.0595 sec/batch
Epoch: 1/20... Training Step: 3939... Training loss: 2.0639... 0.0829 sec/batch
Epoch: 1/20... Training Step: 3940... Training loss: 1.8459... 0.0649 sec/batch
Epoch: 1/20... Training Step: 3941... Training loss: 1.8823... 0.0720 sec/batch
Epoch: 1/20... Training Step: 3942... Training loss: 1.9198... 0.0770 sec/batch
Epoch: 1/20... Training Step: 3943... Training loss: 1.7935... 0.0598 sec/batch
Epoch: 1/20... Training Step: 3944... Training loss: 1.7597... 0.0872 sec/batch
Epoch: 1/20... Training Step: 3945... Training loss: 1.7583... 0.0706 sec/batch
Epoch: 1/20... Training Step: 3946... Training loss: 2.0563... 0.0619 sec/batch
Epoch: 1/20... Training Step: 3947... Training loss: 1.8787... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3948... Training loss: 1.9205... 0.0697 sec/batch
Epoch: 1/20... Training Step: 3949... Training loss: 1.7943... 0.0567 sec/batch
Epoch: 1/20... Training Step: 3950... Training loss: 1.7327... 0.0730 sec/batch
Epoch: 1/20... Training Step: 3951... Training loss: 1.7684... 0.0746 sec/batch
Epoch: 1/20... Training Step: 3952... Training loss: 1.7081... 0.0654 sec/batch
Epoch: 1/20... Training Step: 3953... Training loss: 1.8050... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3954... Training loss: 1.8696... 0.0566 sec/batch
Epoch: 1/20... Training Step: 3955... Training loss: 1.7430... 0.0660 sec/batch
Epoch: 1/20... Training Step: 3956... Training loss: 1.9692... 0.0612 sec/batch
Epoch: 1/20... Training Step: 3957... Training loss: 1.8934... 0.0687 sec/batch
Epoch: 1/20... Training Step: 3958... Training loss: 1.8618... 0.0709 sec/batch
Epoch: 1/20... Training Step: 3959... Training loss: 1.6946... 0.0635 sec/batch
Epoch: 1/20... Training Step: 3960... Training loss: 1.8535... 0.0520 sec/batch
Epoch: 1/20... Training Step: 3961... Training loss: 1.8305... 0.0545 sec/batch
Epoch: 1/20... Training Step: 3962... Training loss: 1.8373... 0.0695 sec/batch
Epoch: 1/20... Training Step: 3963... Training loss: 1.9552... 0.0617 sec/batch
Epoch: 1/20... Training Step: 3964... Training loss: 1.7494... 0.0531 sec/batch
Epoch: 1/20... Training Step: 3965... Training loss: 1.7720... 0.0544 sec/batch
Epoch: 1/20... Training Step: 3966... Training loss: 1.8316... 0.0823 sec/batch
Epoch: 1/20... Training Step: 3967... Training loss: 1.8058... 0.0649 sec/batch
Epoch: 1/20... Training Step: 3968... Training loss: 1.8216... 0.0575 sec/batch
Epoch: 1/20... Training Step: 3969... Training loss: 1.8745... 0.0717 sec/batch
Epoch: 1/20... Training Step: 3970... Training loss: 1.9307... 0.0678 sec/batch
Epoch: 2/20... Training Step: 3971... Training loss: 1.7703... 0.0579 sec/batch
Epoch: 2/20... Training Step: 3972... Training loss: 1.7227... 0.0616 sec/batch
Epoch: 2/20... Training Step: 3973... Training loss: 1.7643... 0.0702 sec/batch
Epoch: 2/20... Training Step: 3974... Training loss: 1.7002... 0.0581 sec/batch
Epoch: 2/20... Training Step: 3975... Training loss: 1.8550... 0.0583 sec/batch
Epoch: 2/20... Training Step: 3976... Training loss: 1.7840... 0.0763 sec/batch
Epoch: 2/20... Training Step: 3977... Training loss: 1.6586... 0.0689 sec/batch
Epoch: 2/20... Training Step: 3978... Training loss: 1.7865... 0.0657 sec/batch
Epoch: 2/20... Training Step: 3979... Training loss: 1.8773... 0.0527 sec/batch
Epoch: 2/20... Training Step: 3980... Training loss: 1.8437... 0.0696 sec/batch
Epoch: 2/20... Training Step: 3981... Training loss: 1.7175... 0.0529 sec/batch
Epoch: 2/20... Training Step: 3982... Training loss: 1.6238... 0.0620 sec/batch
Epoch: 2/20... Training Step: 3983... Training loss: 1.7224... 0.0592 sec/batch
Epoch: 2/20... Training Step: 3984... Training loss: 1.7392... 0.0550 sec/batch
Epoch: 2/20... Training Step: 3985... Training loss: 1.5954... 0.0726 sec/batch
Epoch: 2/20... Training Step: 3986... Training loss: 1.7428... 0.0591 sec/batch
Epoch: 2/20... Training Step: 3987... Training loss: 1.6899... 0.0525 sec/batch
Epoch: 2/20... Training Step: 3988... Training loss: 1.6673... 0.0534 sec/batch
Epoch: 2/20... Training Step: 3989... Training loss: 1.7780... 0.0694 sec/batch
Epoch: 2/20... Training Step: 3990... Training loss: 1.9104... 0.0582 sec/batch
Epoch: 2/20... Training Step: 3991... Training loss: 1.8385... 0.0744 sec/batch
Epoch: 2/20... Training Step: 3992... Training loss: 1.7084... 0.0673 sec/batch
Epoch: 2/20... Training Step: 3993... Training loss: 1.7622... 0.0812 sec/batch
Epoch: 2/20... Training Step: 3994... Training loss: 1.7153... 0.0718 sec/batch
Epoch: 2/20... Training Step: 3995... Training loss: 1.7187... 0.0671 sec/batch
Epoch: 2/20... Training Step: 3996... Training loss: 1.8527... 0.0669 sec/batch
Epoch: 2/20... Training Step: 3997... Training loss: 1.9298... 0.0572 sec/batch
Epoch: 2/20... Training Step: 3998... Training loss: 1.9760... 0.0555 sec/batch
Epoch: 2/20... Training Step: 3999... Training loss: 1.9989... 0.0665 sec/batch
Epoch: 2/20... Training Step: 4000... Training loss: 1.7814... 0.0653 sec/batch
Epoch: 2/20... Training Step: 4001... Training loss: 1.8833... 0.0551 sec/batch
Epoch: 2/20... Training Step: 4002... Training loss: 1.9509... 0.0585 sec/batch
Epoch: 2/20... Training Step: 4003... Training loss: 1.7983... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4004... Training loss: 1.8059... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4005... Training loss: 1.8121... 0.0684 sec/batch
Epoch: 2/20... Training Step: 4006... Training loss: 1.6503... 0.0747 sec/batch
Epoch: 2/20... Training Step: 4007... Training loss: 1.7859... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4008... Training loss: 1.8551... 0.0576 sec/batch
Epoch: 2/20... Training Step: 4009... Training loss: 1.8038... 0.0784 sec/batch
Epoch: 2/20... Training Step: 4010... Training loss: 1.8715... 0.0847 sec/batch
Epoch: 2/20... Training Step: 4011... Training loss: 1.8327... 0.0787 sec/batch
Epoch: 2/20... Training Step: 4012... Training loss: 1.7853... 0.0683 sec/batch
Epoch: 2/20... Training Step: 4013... Training loss: 1.8070... 0.0768 sec/batch
Epoch: 2/20... Training Step: 4014... Training loss: 1.7206... 0.0610 sec/batch
Epoch: 2/20... Training Step: 4015... Training loss: 1.7781... 0.0636 sec/batch
Epoch: 2/20... Training Step: 4016... Training loss: 1.9131... 0.0871 sec/batch
Epoch: 2/20... Training Step: 4017... Training loss: 1.7969... 0.0644 sec/batch
Epoch: 2/20... Training Step: 4018... Training loss: 1.7957... 0.0804 sec/batch
Epoch: 2/20... Training Step: 4019... Training loss: 1.8899... 0.0589 sec/batch
Epoch: 2/20... Training Step: 4020... Training loss: 1.9668... 0.0661 sec/batch
Epoch: 2/20... Training Step: 4021... Training loss: 1.7463... 0.0848 sec/batch
Epoch: 2/20... Training Step: 4022... Training loss: 1.8140... 0.0747 sec/batch
Epoch: 2/20... Training Step: 4023... Training loss: 1.9064... 0.0601 sec/batch
Epoch: 2/20... Training Step: 4024... Training loss: 1.8953... 0.0667 sec/batch
Epoch: 2/20... Training Step: 4025... Training loss: 1.8766... 0.0631 sec/batch
Epoch: 2/20... Training Step: 4026... Training loss: 1.7764... 0.0797 sec/batch
Epoch: 2/20... Training Step: 4027... Training loss: 1.7900... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4028... Training loss: 1.7123... 0.0690 sec/batch
Epoch: 2/20... Training Step: 4029... Training loss: 1.8194... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4030... Training loss: 1.7354... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4031... Training loss: 2.0053... 0.0730 sec/batch
Epoch: 2/20... Training Step: 4032... Training loss: 1.9714... 0.0607 sec/batch
Epoch: 2/20... Training Step: 4033... Training loss: 1.8840... 0.0848 sec/batch
Epoch: 2/20... Training Step: 4034... Training loss: 1.6194... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4035... Training loss: 1.9213... 0.0689 sec/batch
Epoch: 2/20... Training Step: 4036... Training loss: 1.8925... 0.0744 sec/batch
Epoch: 2/20... Training Step: 4037... Training loss: 1.7862... 0.0868 sec/batch
Epoch: 2/20... Training Step: 4038... Training loss: 1.9752... 0.0635 sec/batch
Epoch: 2/20... Training Step: 4039... Training loss: 1.9799... 0.0659 sec/batch
Epoch: 2/20... Training Step: 4040... Training loss: 1.7606... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4041... Training loss: 1.8402... 0.0633 sec/batch
Epoch: 2/20... Training Step: 4042... Training loss: 1.7705... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4043... Training loss: 1.7439... 0.0603 sec/batch
Epoch: 2/20... Training Step: 4044... Training loss: 1.7998... 0.0652 sec/batch
Epoch: 2/20... Training Step: 4045... Training loss: 1.7071... 0.0557 sec/batch
Epoch: 2/20... Training Step: 4046... Training loss: 1.7900... 0.0662 sec/batch
Epoch: 2/20... Training Step: 4047... Training loss: 1.6905... 0.0677 sec/batch
Epoch: 2/20... Training Step: 4048... Training loss: 1.6920... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4049... Training loss: 1.7686... 0.0588 sec/batch
Epoch: 2/20... Training Step: 4050... Training loss: 1.7686... 0.0678 sec/batch
Epoch: 2/20... Training Step: 4051... Training loss: 1.7301... 0.0668 sec/batch
Epoch: 2/20... Training Step: 4052... Training loss: 1.6977... 0.0687 sec/batch
Epoch: 2/20... Training Step: 4053... Training loss: 1.7424... 0.0601 sec/batch
Epoch: 2/20... Training Step: 4054... Training loss: 1.7617... 0.0849 sec/batch
Epoch: 2/20... Training Step: 4055... Training loss: 1.7005... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4056... Training loss: 1.7881... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4057... Training loss: 1.8210... 0.0563 sec/batch
Epoch: 2/20... Training Step: 4058... Training loss: 1.7128... 0.0831 sec/batch
Epoch: 2/20... Training Step: 4059... Training loss: 1.7505... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4060... Training loss: 1.8190... 0.0691 sec/batch
Epoch: 2/20... Training Step: 4061... Training loss: 1.7257... 0.0551 sec/batch
Epoch: 2/20... Training Step: 4062... Training loss: 1.8110... 0.0761 sec/batch
Epoch: 2/20... Training Step: 4063... Training loss: 1.8704... 0.0709 sec/batch
Epoch: 2/20... Training Step: 4064... Training loss: 1.8406... 0.0594 sec/batch
Epoch: 2/20... Training Step: 4065... Training loss: 1.7468... 0.0824 sec/batch
Epoch: 2/20... Training Step: 4066... Training loss: 1.7545... 0.0971 sec/batch
Epoch: 2/20... Training Step: 4067... Training loss: 1.8470... 0.0852 sec/batch
Epoch: 2/20... Training Step: 4068... Training loss: 1.8546... 0.0995 sec/batch
Epoch: 2/20... Training Step: 4069... Training loss: 1.7664... 0.0908 sec/batch
Epoch: 2/20... Training Step: 4070... Training loss: 1.7569... 0.1068 sec/batch
Epoch: 2/20... Training Step: 4071... Training loss: 1.9040... 0.0998 sec/batch
Epoch: 2/20... Training Step: 4072... Training loss: 1.8032... 0.0999 sec/batch
Epoch: 2/20... Training Step: 4073... Training loss: 1.8911... 0.0981 sec/batch
Epoch: 2/20... Training Step: 4074... Training loss: 1.6600... 0.1129 sec/batch
Epoch: 2/20... Training Step: 4075... Training loss: 1.7507... 0.0963 sec/batch
Epoch: 2/20... Training Step: 4076... Training loss: 1.7505... 0.1133 sec/batch
Epoch: 2/20... Training Step: 4077... Training loss: 1.7116... 0.1317 sec/batch
Epoch: 2/20... Training Step: 4078... Training loss: 1.6927... 0.1110 sec/batch
Epoch: 2/20... Training Step: 4079... Training loss: 1.6494... 0.1311 sec/batch
Epoch: 2/20... Training Step: 4080... Training loss: 1.8262... 0.0938 sec/batch
Epoch: 2/20... Training Step: 4081... Training loss: 1.7027... 0.0836 sec/batch
Epoch: 2/20... Training Step: 4082... Training loss: 1.7876... 0.1215 sec/batch
Epoch: 2/20... Training Step: 4083... Training loss: 1.7079... 0.1028 sec/batch
Epoch: 2/20... Training Step: 4084... Training loss: 1.6954... 0.1133 sec/batch
Epoch: 2/20... Training Step: 4085... Training loss: 1.7912... 0.1303 sec/batch
Epoch: 2/20... Training Step: 4086... Training loss: 1.8334... 0.1136 sec/batch
Epoch: 2/20... Training Step: 4087... Training loss: 1.7655... 0.1214 sec/batch
Epoch: 2/20... Training Step: 4088... Training loss: 1.6361... 0.1085 sec/batch
Epoch: 2/20... Training Step: 4089... Training loss: 1.7481... 0.1096 sec/batch
Epoch: 2/20... Training Step: 4090... Training loss: 1.6900... 0.1326 sec/batch
Epoch: 2/20... Training Step: 4091... Training loss: 1.6539... 0.0980 sec/batch
Epoch: 2/20... Training Step: 4092... Training loss: 1.6313... 0.1146 sec/batch
Epoch: 2/20... Training Step: 4093... Training loss: 1.7845... 0.1078 sec/batch
Epoch: 2/20... Training Step: 4094... Training loss: 1.7386... 0.0993 sec/batch
Epoch: 2/20... Training Step: 4095... Training loss: 1.7626... 0.1210 sec/batch
Epoch: 2/20... Training Step: 4096... Training loss: 1.7638... 0.0963 sec/batch
Epoch: 2/20... Training Step: 4097... Training loss: 1.7490... 0.0948 sec/batch
Epoch: 2/20... Training Step: 4098... Training loss: 1.6676... 0.1036 sec/batch
Epoch: 2/20... Training Step: 4099... Training loss: 1.8386... 0.0998 sec/batch
Epoch: 2/20... Training Step: 4100... Training loss: 1.7442... 0.1170 sec/batch
Epoch: 2/20... Training Step: 4101... Training loss: 1.8988... 0.0994 sec/batch
Epoch: 2/20... Training Step: 4102... Training loss: 1.7906... 0.0982 sec/batch
Epoch: 2/20... Training Step: 4103... Training loss: 1.8153... 0.0921 sec/batch
Epoch: 2/20... Training Step: 4104... Training loss: 1.7888... 0.0735 sec/batch
Epoch: 2/20... Training Step: 4105... Training loss: 1.7882... 0.1050 sec/batch
Epoch: 2/20... Training Step: 4106... Training loss: 1.6599... 0.0825 sec/batch
Epoch: 2/20... Training Step: 4107... Training loss: 1.8817... 0.0677 sec/batch
Epoch: 2/20... Training Step: 4108... Training loss: 1.6667... 0.0780 sec/batch
Epoch: 2/20... Training Step: 4109... Training loss: 1.7852... 0.0733 sec/batch
Epoch: 2/20... Training Step: 4110... Training loss: 1.7644... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4111... Training loss: 1.7623... 0.0822 sec/batch
Epoch: 2/20... Training Step: 4112... Training loss: 1.6911... 0.0961 sec/batch
Epoch: 2/20... Training Step: 4113... Training loss: 1.7948... 0.0813 sec/batch
Epoch: 2/20... Training Step: 4114... Training loss: 1.8437... 0.0622 sec/batch
Epoch: 2/20... Training Step: 4115... Training loss: 1.6544... 0.0836 sec/batch
Epoch: 2/20... Training Step: 4116... Training loss: 1.8477... 0.0738 sec/batch
Epoch: 2/20... Training Step: 4117... Training loss: 1.8043... 0.0761 sec/batch
Epoch: 2/20... Training Step: 4118... Training loss: 1.6700... 0.0968 sec/batch
Epoch: 2/20... Training Step: 4119... Training loss: 1.7282... 0.0750 sec/batch
Epoch: 2/20... Training Step: 4120... Training loss: 1.6468... 0.0798 sec/batch
Epoch: 2/20... Training Step: 4121... Training loss: 1.8093... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4122... Training loss: 1.7441... 0.0638 sec/batch
Epoch: 2/20... Training Step: 4123... Training loss: 1.7685... 0.0701 sec/batch
Epoch: 2/20... Training Step: 4124... Training loss: 1.7453... 0.0857 sec/batch
Epoch: 2/20... Training Step: 4125... Training loss: 1.9320... 0.0793 sec/batch
Epoch: 2/20... Training Step: 4126... Training loss: 1.7048... 0.0922 sec/batch
Epoch: 2/20... Training Step: 4127... Training loss: 1.6170... 0.0850 sec/batch
Epoch: 2/20... Training Step: 4128... Training loss: 1.8648... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4129... Training loss: 2.0290... 0.0661 sec/batch
Epoch: 2/20... Training Step: 4130... Training loss: 1.9955... 0.0830 sec/batch
Epoch: 2/20... Training Step: 4131... Training loss: 1.8194... 0.0855 sec/batch
Epoch: 2/20... Training Step: 4132... Training loss: 1.7423... 0.0887 sec/batch
Epoch: 2/20... Training Step: 4133... Training loss: 1.8310... 0.0710 sec/batch
Epoch: 2/20... Training Step: 4134... Training loss: 1.7914... 0.0775 sec/batch
Epoch: 2/20... Training Step: 4135... Training loss: 1.8375... 0.0842 sec/batch
Epoch: 2/20... Training Step: 4136... Training loss: 1.7754... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4137... Training loss: 1.7300... 0.0711 sec/batch
Epoch: 2/20... Training Step: 4138... Training loss: 1.9586... 0.0886 sec/batch
Epoch: 2/20... Training Step: 4139... Training loss: 1.8009... 0.0860 sec/batch
Epoch: 2/20... Training Step: 4140... Training loss: 1.8726... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4141... Training loss: 1.8127... 0.0827 sec/batch
Epoch: 2/20... Training Step: 4142... Training loss: 1.8425... 0.0750 sec/batch
Epoch: 2/20... Training Step: 4143... Training loss: 1.8120... 0.0614 sec/batch
Epoch: 2/20... Training Step: 4144... Training loss: 1.8064... 0.0695 sec/batch
Epoch: 2/20... Training Step: 4145... Training loss: 1.8257... 0.0736 sec/batch
Epoch: 2/20... Training Step: 4146... Training loss: 1.8107... 0.0658 sec/batch
Epoch: 2/20... Training Step: 4147... Training loss: 1.7262... 0.0845 sec/batch
Epoch: 2/20... Training Step: 4148... Training loss: 1.7490... 0.0919 sec/batch
Epoch: 2/20... Training Step: 4149... Training loss: 1.8018... 0.0800 sec/batch
Epoch: 2/20... Training Step: 4150... Training loss: 1.7097... 0.0908 sec/batch
Epoch: 2/20... Training Step: 4151... Training loss: 1.7521... 0.0841 sec/batch
Epoch: 2/20... Training Step: 4152... Training loss: 1.8535... 0.0611 sec/batch
Epoch: 2/20... Training Step: 4153... Training loss: 1.7600... 0.0874 sec/batch
Epoch: 2/20... Training Step: 4154... Training loss: 1.7117... 0.0741 sec/batch
Epoch: 2/20... Training Step: 4155... Training loss: 1.7216... 0.0588 sec/batch
Epoch: 2/20... Training Step: 4156... Training loss: 1.7406... 0.0714 sec/batch
Epoch: 2/20... Training Step: 4157... Training loss: 1.6309... 0.0938 sec/batch
Epoch: 2/20... Training Step: 4158... Training loss: 1.7534... 0.0705 sec/batch
Epoch: 2/20... Training Step: 4159... Training loss: 1.7581... 0.0617 sec/batch
Epoch: 2/20... Training Step: 4160... Training loss: 1.8075... 0.0636 sec/batch
Epoch: 2/20... Training Step: 4161... Training loss: 1.6996... 0.0624 sec/batch
Epoch: 2/20... Training Step: 4162... Training loss: 1.7178... 0.0627 sec/batch
Epoch: 2/20... Training Step: 4163... Training loss: 1.7546... 0.0854 sec/batch
Epoch: 2/20... Training Step: 4164... Training loss: 1.8861... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4165... Training loss: 1.7999... 0.0600 sec/batch
Epoch: 2/20... Training Step: 4166... Training loss: 1.6951... 0.0881 sec/batch
Epoch: 2/20... Training Step: 4167... Training loss: 1.6880... 0.0728 sec/batch
Epoch: 2/20... Training Step: 4168... Training loss: 1.8588... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4169... Training loss: 1.7680... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4170... Training loss: 1.8375... 0.0837 sec/batch
Epoch: 2/20... Training Step: 4171... Training loss: 1.7119... 0.0576 sec/batch
Epoch: 2/20... Training Step: 4172... Training loss: 1.8660... 0.0795 sec/batch
Epoch: 2/20... Training Step: 4173... Training loss: 1.7540... 0.0712 sec/batch
Epoch: 2/20... Training Step: 4174... Training loss: 1.8723... 0.0635 sec/batch
Epoch: 2/20... Training Step: 4175... Training loss: 1.8194... 0.0869 sec/batch
Epoch: 2/20... Training Step: 4176... Training loss: 1.8661... 0.0697 sec/batch
Epoch: 2/20... Training Step: 4177... Training loss: 1.7901... 0.0683 sec/batch
Epoch: 2/20... Training Step: 4178... Training loss: 1.7731... 0.0820 sec/batch
Epoch: 2/20... Training Step: 4179... Training loss: 1.8561... 0.0734 sec/batch
Epoch: 2/20... Training Step: 4180... Training loss: 1.7703... 0.0559 sec/batch
Epoch: 2/20... Training Step: 4181... Training loss: 1.6815... 0.0839 sec/batch
Epoch: 2/20... Training Step: 4182... Training loss: 1.7592... 0.0739 sec/batch
Epoch: 2/20... Training Step: 4183... Training loss: 1.7864... 0.0564 sec/batch
Epoch: 2/20... Training Step: 4184... Training loss: 1.8684... 0.0795 sec/batch
Epoch: 2/20... Training Step: 4185... Training loss: 1.7690... 0.0834 sec/batch
Epoch: 2/20... Training Step: 4186... Training loss: 1.7483... 0.0681 sec/batch
Epoch: 2/20... Training Step: 4187... Training loss: 1.8979... 0.0824 sec/batch
Epoch: 2/20... Training Step: 4188... Training loss: 1.8438... 0.0873 sec/batch
Epoch: 2/20... Training Step: 4189... Training loss: 1.7688... 0.0734 sec/batch
Epoch: 2/20... Training Step: 4190... Training loss: 1.8019... 0.0773 sec/batch
Epoch: 2/20... Training Step: 4191... Training loss: 1.8064... 0.0677 sec/batch
Epoch: 2/20... Training Step: 4192... Training loss: 1.7776... 0.0685 sec/batch
Epoch: 2/20... Training Step: 4193... Training loss: 1.8314... 0.0700 sec/batch
Epoch: 2/20... Training Step: 4194... Training loss: 1.8535... 0.0611 sec/batch
Epoch: 2/20... Training Step: 4195... Training loss: 1.8173... 0.0731 sec/batch
Epoch: 2/20... Training Step: 4196... Training loss: 1.8891... 0.0657 sec/batch
Epoch: 2/20... Training Step: 4197... Training loss: 1.9605... 0.0681 sec/batch
Epoch: 2/20... Training Step: 4198... Training loss: 1.9109... 0.0705 sec/batch
Epoch: 2/20... Training Step: 4199... Training loss: 1.7347... 0.0586 sec/batch
Epoch: 2/20... Training Step: 4200... Training loss: 1.7103... 0.0693 sec/batch
Epoch: 2/20... Training Step: 4201... Training loss: 1.7835... 0.0760 sec/batch
Epoch: 2/20... Training Step: 4202... Training loss: 1.7914... 0.0610 sec/batch
Epoch: 2/20... Training Step: 4203... Training loss: 1.7130... 0.0687 sec/batch
Epoch: 2/20... Training Step: 4204... Training loss: 1.8410... 0.0609 sec/batch
Epoch: 2/20... Training Step: 4205... Training loss: 1.8052... 0.0747 sec/batch
Epoch: 2/20... Training Step: 4206... Training loss: 1.9708... 0.0653 sec/batch
Epoch: 2/20... Training Step: 4207... Training loss: 1.8389... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4208... Training loss: 1.8818... 0.0619 sec/batch
Epoch: 2/20... Training Step: 4209... Training loss: 1.7913... 0.0831 sec/batch
Epoch: 2/20... Training Step: 4210... Training loss: 1.8092... 0.0596 sec/batch
Epoch: 2/20... Training Step: 4211... Training loss: 1.8337... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4212... Training loss: 1.8724... 0.0551 sec/batch
Epoch: 2/20... Training Step: 4213... Training loss: 1.8299... 0.0502 sec/batch
Epoch: 2/20... Training Step: 4214... Training loss: 1.8770... 0.0700 sec/batch
Epoch: 2/20... Training Step: 4215... Training loss: 1.9197... 0.0680 sec/batch
Epoch: 2/20... Training Step: 4216... Training loss: 1.9059... 0.0680 sec/batch
Epoch: 2/20... Training Step: 4217... Training loss: 1.8113... 0.0829 sec/batch
Epoch: 2/20... Training Step: 4218... Training loss: 1.8497... 0.0743 sec/batch
Epoch: 2/20... Training Step: 4219... Training loss: 1.6973... 0.0669 sec/batch
Epoch: 2/20... Training Step: 4220... Training loss: 1.7747... 0.0801 sec/batch
Epoch: 2/20... Training Step: 4221... Training loss: 1.7195... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4222... Training loss: 1.8932... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4223... Training loss: 1.7961... 0.0559 sec/batch
Epoch: 2/20... Training Step: 4224... Training loss: 1.8311... 0.0766 sec/batch
Epoch: 2/20... Training Step: 4225... Training loss: 1.9025... 0.0692 sec/batch
Epoch: 2/20... Training Step: 4226... Training loss: 1.8029... 0.0655 sec/batch
Epoch: 2/20... Training Step: 4227... Training loss: 1.7870... 0.0693 sec/batch
Epoch: 2/20... Training Step: 4228... Training loss: 1.7763... 0.0613 sec/batch
Epoch: 2/20... Training Step: 4229... Training loss: 1.7195... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4230... Training loss: 1.6823... 0.0599 sec/batch
Epoch: 2/20... Training Step: 4231... Training loss: 1.7250... 0.0604 sec/batch
Epoch: 2/20... Training Step: 4232... Training loss: 1.7006... 0.0545 sec/batch
Epoch: 2/20... Training Step: 4233... Training loss: 1.7860... 0.0585 sec/batch
Epoch: 2/20... Training Step: 4234... Training loss: 1.7542... 0.0833 sec/batch
Epoch: 2/20... Training Step: 4235... Training loss: 1.6511... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4236... Training loss: 1.7262... 0.0523 sec/batch
Epoch: 2/20... Training Step: 4237... Training loss: 1.6665... 0.0553 sec/batch
Epoch: 2/20... Training Step: 4238... Training loss: 1.7397... 0.0793 sec/batch
Epoch: 2/20... Training Step: 4239... Training loss: 1.8294... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4240... Training loss: 1.8666... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4241... Training loss: 1.7870... 0.0720 sec/batch
Epoch: 2/20... Training Step: 4242... Training loss: 1.7510... 0.0559 sec/batch
Epoch: 2/20... Training Step: 4243... Training loss: 1.9602... 0.0787 sec/batch
Epoch: 2/20... Training Step: 4244... Training loss: 1.8240... 0.0689 sec/batch
Epoch: 2/20... Training Step: 4245... Training loss: 1.8631... 0.0651 sec/batch
Epoch: 2/20... Training Step: 4246... Training loss: 1.7712... 0.0554 sec/batch
Epoch: 2/20... Training Step: 4247... Training loss: 1.7325... 0.0691 sec/batch
Epoch: 2/20... Training Step: 4248... Training loss: 1.8340... 0.0572 sec/batch
Epoch: 2/20... Training Step: 4249... Training loss: 1.7578... 0.0592 sec/batch
Epoch: 2/20... Training Step: 4250... Training loss: 1.8401... 0.0531 sec/batch
Epoch: 2/20... Training Step: 4251... Training loss: 1.7355... 0.0544 sec/batch
Epoch: 2/20... Training Step: 4252... Training loss: 1.7303... 0.0612 sec/batch
Epoch: 2/20... Training Step: 4253... Training loss: 1.6749... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4254... Training loss: 1.8317... 0.0531 sec/batch
Epoch: 2/20... Training Step: 4255... Training loss: 1.6526... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4256... Training loss: 1.7259... 0.0791 sec/batch
Epoch: 2/20... Training Step: 4257... Training loss: 1.6270... 0.0796 sec/batch
Epoch: 2/20... Training Step: 4258... Training loss: 1.7820... 0.0616 sec/batch
Epoch: 2/20... Training Step: 4259... Training loss: 1.6859... 0.0631 sec/batch
Epoch: 2/20... Training Step: 4260... Training loss: 1.7551... 0.0703 sec/batch
Epoch: 2/20... Training Step: 4261... Training loss: 1.7590... 0.0596 sec/batch
Epoch: 2/20... Training Step: 4262... Training loss: 1.7111... 0.0609 sec/batch
Epoch: 2/20... Training Step: 4263... Training loss: 1.7383... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4264... Training loss: 1.6774... 0.0608 sec/batch
Epoch: 2/20... Training Step: 4265... Training loss: 1.8487... 0.0527 sec/batch
Epoch: 2/20... Training Step: 4266... Training loss: 1.7678... 0.0735 sec/batch
Epoch: 2/20... Training Step: 4267... Training loss: 1.6372... 0.0708 sec/batch
Epoch: 2/20... Training Step: 4268... Training loss: 1.7802... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4269... Training loss: 1.6615... 0.0606 sec/batch
Epoch: 2/20... Training Step: 4270... Training loss: 1.7016... 0.0879 sec/batch
Epoch: 2/20... Training Step: 4271... Training loss: 1.8160... 0.0799 sec/batch
Epoch: 2/20... Training Step: 4272... Training loss: 1.7633... 0.0667 sec/batch
Epoch: 2/20... Training Step: 4273... Training loss: 1.7584... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4274... Training loss: 1.8101... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4275... Training loss: 1.8370... 0.0687 sec/batch
Epoch: 2/20... Training Step: 4276... Training loss: 1.6357... 0.0673 sec/batch
Epoch: 2/20... Training Step: 4277... Training loss: 1.9100... 0.0688 sec/batch
Epoch: 2/20... Training Step: 4278... Training loss: 1.7992... 0.0803 sec/batch
Epoch: 2/20... Training Step: 4279... Training loss: 1.7775... 0.0760 sec/batch
Epoch: 2/20... Training Step: 4280... Training loss: 1.9053... 0.0778 sec/batch
Epoch: 2/20... Training Step: 4281... Training loss: 1.8196... 0.0803 sec/batch
Epoch: 2/20... Training Step: 4282... Training loss: 1.7589... 0.0634 sec/batch
Epoch: 2/20... Training Step: 4283... Training loss: 1.7354... 0.0811 sec/batch
Epoch: 2/20... Training Step: 4284... Training loss: 1.8312... 0.0753 sec/batch
Epoch: 2/20... Training Step: 4285... Training loss: 1.8741... 0.0624 sec/batch
Epoch: 2/20... Training Step: 4286... Training loss: 1.8679... 0.0806 sec/batch
Epoch: 2/20... Training Step: 4287... Training loss: 1.7438... 0.0790 sec/batch
Epoch: 2/20... Training Step: 4288... Training loss: 1.7297... 0.0751 sec/batch
Epoch: 2/20... Training Step: 4289... Training loss: 1.7905... 0.0675 sec/batch
Epoch: 2/20... Training Step: 4290... Training loss: 1.8550... 0.0811 sec/batch
Epoch: 2/20... Training Step: 4291... Training loss: 1.7968... 0.0734 sec/batch
Epoch: 2/20... Training Step: 4292... Training loss: 1.5930... 0.0642 sec/batch
Epoch: 2/20... Training Step: 4293... Training loss: 1.8768... 0.0748 sec/batch
Epoch: 2/20... Training Step: 4294... Training loss: 1.7623... 0.0664 sec/batch
Epoch: 2/20... Training Step: 4295... Training loss: 1.7551... 0.0877 sec/batch
Epoch: 2/20... Training Step: 4296... Training loss: 1.8571... 0.0820 sec/batch
Epoch: 2/20... Training Step: 4297... Training loss: 1.8831... 0.0743 sec/batch
Epoch: 2/20... Training Step: 4298... Training loss: 1.8155... 0.0766 sec/batch
Epoch: 2/20... Training Step: 4299... Training loss: 1.8794... 0.0672 sec/batch
Epoch: 2/20... Training Step: 4300... Training loss: 1.7589... 0.0672 sec/batch
Epoch: 2/20... Training Step: 4301... Training loss: 1.8496... 0.0657 sec/batch
Epoch: 2/20... Training Step: 4302... Training loss: 1.8692... 0.0684 sec/batch
Epoch: 2/20... Training Step: 4303... Training loss: 1.8069... 0.0584 sec/batch
Epoch: 2/20... Training Step: 4304... Training loss: 1.8926... 0.0728 sec/batch
Epoch: 2/20... Training Step: 4305... Training loss: 1.9197... 0.0625 sec/batch
Epoch: 2/20... Training Step: 4306... Training loss: 1.8807... 0.0766 sec/batch
Epoch: 2/20... Training Step: 4307... Training loss: 1.6955... 0.0601 sec/batch
Epoch: 2/20... Training Step: 4308... Training loss: 1.6613... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4309... Training loss: 1.7817... 0.0935 sec/batch
Epoch: 2/20... Training Step: 4310... Training loss: 1.8411... 0.0593 sec/batch
Epoch: 2/20... Training Step: 4311... Training loss: 2.1006... 0.0588 sec/batch
Epoch: 2/20... Training Step: 4312... Training loss: 2.1087... 0.0698 sec/batch
Epoch: 2/20... Training Step: 4313... Training loss: 1.9568... 0.0774 sec/batch
Epoch: 2/20... Training Step: 4314... Training loss: 1.7903... 0.0726 sec/batch
Epoch: 2/20... Training Step: 4315... Training loss: 1.7958... 0.0751 sec/batch
Epoch: 2/20... Training Step: 4316... Training loss: 1.9311... 0.0624 sec/batch
Epoch: 2/20... Training Step: 4317... Training loss: 1.9035... 0.0810 sec/batch
Epoch: 2/20... Training Step: 4318... Training loss: 1.8451... 0.0552 sec/batch
Epoch: 2/20... Training Step: 4319... Training loss: 1.8105... 0.0682 sec/batch
Epoch: 2/20... Training Step: 4320... Training loss: 1.8151... 0.0755 sec/batch
Epoch: 2/20... Training Step: 4321... Training loss: 1.7729... 0.0636 sec/batch
Epoch: 2/20... Training Step: 4322... Training loss: 1.6542... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4323... Training loss: 1.8820... 0.0977 sec/batch
Epoch: 2/20... Training Step: 4324... Training loss: 1.8363... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4325... Training loss: 1.8453... 0.1001 sec/batch
Epoch: 2/20... Training Step: 4326... Training loss: 1.7904... 0.1087 sec/batch
Epoch: 2/20... Training Step: 4327... Training loss: 1.8689... 0.1186 sec/batch
Epoch: 2/20... Training Step: 4328... Training loss: 1.8113... 0.1306 sec/batch
Epoch: 2/20... Training Step: 4329... Training loss: 1.7093... 0.1362 sec/batch
Epoch: 2/20... Training Step: 4330... Training loss: 1.8335... 0.1180 sec/batch
Epoch: 2/20... Training Step: 4331... Training loss: 1.8280... 0.1158 sec/batch
Epoch: 2/20... Training Step: 4332... Training loss: 1.7813... 0.1365 sec/batch
Epoch: 2/20... Training Step: 4333... Training loss: 1.8865... 0.1286 sec/batch
Epoch: 2/20... Training Step: 4334... Training loss: 1.6494... 0.1037 sec/batch
Epoch: 2/20... Training Step: 4335... Training loss: 1.8417... 0.1272 sec/batch
Epoch: 2/20... Training Step: 4336... Training loss: 1.7169... 0.1245 sec/batch
Epoch: 2/20... Training Step: 4337... Training loss: 1.8923... 0.1083 sec/batch
Epoch: 2/20... Training Step: 4338... Training loss: 1.7976... 0.1447 sec/batch
Epoch: 2/20... Training Step: 4339... Training loss: 1.8494... 0.1245 sec/batch
Epoch: 2/20... Training Step: 4340... Training loss: 1.8927... 0.1346 sec/batch
Epoch: 2/20... Training Step: 4341... Training loss: 1.7665... 0.1146 sec/batch
Epoch: 2/20... Training Step: 4342... Training loss: 1.7568... 0.1039 sec/batch
Epoch: 2/20... Training Step: 4343... Training loss: 1.7710... 0.1031 sec/batch
Epoch: 2/20... Training Step: 4344... Training loss: 1.7270... 0.1231 sec/batch
Epoch: 2/20... Training Step: 4345... Training loss: 1.8005... 0.1259 sec/batch
Epoch: 2/20... Training Step: 4346... Training loss: 1.7862... 0.1001 sec/batch
Epoch: 2/20... Training Step: 4347... Training loss: 1.8835... 0.0905 sec/batch
Epoch: 2/20... Training Step: 4348... Training loss: 1.8012... 0.1223 sec/batch
Epoch: 2/20... Training Step: 4349... Training loss: 1.7600... 0.1036 sec/batch
Epoch: 2/20... Training Step: 4350... Training loss: 1.8074... 0.0942 sec/batch
Epoch: 2/20... Training Step: 4351... Training loss: 1.8105... 0.0918 sec/batch
Epoch: 2/20... Training Step: 4352... Training loss: 1.8147... 0.1303 sec/batch
Epoch: 2/20... Training Step: 4353... Training loss: 1.9038... 0.0943 sec/batch
Epoch: 2/20... Training Step: 4354... Training loss: 1.9649... 0.1074 sec/batch
Epoch: 2/20... Training Step: 4355... Training loss: 1.8740... 0.1452 sec/batch
Epoch: 2/20... Training Step: 4356... Training loss: 1.7476... 0.1009 sec/batch
Epoch: 2/20... Training Step: 4357... Training loss: 1.7834... 0.1028 sec/batch
Epoch: 2/20... Training Step: 4358... Training loss: 1.8560... 0.1337 sec/batch
Epoch: 2/20... Training Step: 4359... Training loss: 1.7398... 0.0874 sec/batch
Epoch: 2/20... Training Step: 4360... Training loss: 1.7935... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4361... Training loss: 1.7231... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4362... Training loss: 1.8614... 0.0737 sec/batch
Epoch: 2/20... Training Step: 4363... Training loss: 1.8049... 0.0977 sec/batch
Epoch: 2/20... Training Step: 4364... Training loss: 1.7292... 0.0820 sec/batch
Epoch: 2/20... Training Step: 4365... Training loss: 1.8010... 0.0865 sec/batch
Epoch: 2/20... Training Step: 4366... Training loss: 1.7815... 0.0732 sec/batch
Epoch: 2/20... Training Step: 4367... Training loss: 1.7830... 0.0769 sec/batch
Epoch: 2/20... Training Step: 4368... Training loss: 1.7811... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4369... Training loss: 1.7882... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4370... Training loss: 1.7748... 0.0908 sec/batch
Epoch: 2/20... Training Step: 4371... Training loss: 1.6786... 0.0692 sec/batch
Epoch: 2/20... Training Step: 4372... Training loss: 1.7977... 0.0863 sec/batch
Epoch: 2/20... Training Step: 4373... Training loss: 1.7519... 0.0799 sec/batch
Epoch: 2/20... Training Step: 4374... Training loss: 1.6687... 0.0802 sec/batch
Epoch: 2/20... Training Step: 4375... Training loss: 1.8901... 0.0746 sec/batch
Epoch: 2/20... Training Step: 4376... Training loss: 1.8127... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4377... Training loss: 1.7864... 0.0633 sec/batch
Epoch: 2/20... Training Step: 4378... Training loss: 1.8395... 0.0798 sec/batch
Epoch: 2/20... Training Step: 4379... Training loss: 1.7132... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4380... Training loss: 1.7777... 0.0655 sec/batch
Epoch: 2/20... Training Step: 4381... Training loss: 1.7824... 0.0751 sec/batch
Epoch: 2/20... Training Step: 4382... Training loss: 1.7381... 0.0770 sec/batch
Epoch: 2/20... Training Step: 4383... Training loss: 1.9674... 0.0941 sec/batch
Epoch: 2/20... Training Step: 4384... Training loss: 1.7667... 0.0922 sec/batch
Epoch: 2/20... Training Step: 4385... Training loss: 1.8393... 0.0787 sec/batch
Epoch: 2/20... Training Step: 4386... Training loss: 1.8025... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4387... Training loss: 1.8844... 0.0759 sec/batch
Epoch: 2/20... Training Step: 4388... Training loss: 1.6516... 0.0873 sec/batch
Epoch: 2/20... Training Step: 4389... Training loss: 1.7151... 0.0797 sec/batch
Epoch: 2/20... Training Step: 4390... Training loss: 1.7949... 0.0777 sec/batch
Epoch: 2/20... Training Step: 4391... Training loss: 1.7493... 0.0626 sec/batch
Epoch: 2/20... Training Step: 4392... Training loss: 1.7886... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4393... Training loss: 1.7379... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4394... Training loss: 1.7281... 0.0745 sec/batch
Epoch: 2/20... Training Step: 4395... Training loss: 1.7333... 0.0674 sec/batch
Epoch: 2/20... Training Step: 4396... Training loss: 1.8203... 0.0804 sec/batch
Epoch: 2/20... Training Step: 4397... Training loss: 1.8316... 0.0804 sec/batch
Epoch: 2/20... Training Step: 4398... Training loss: 1.7463... 0.0842 sec/batch
Epoch: 2/20... Training Step: 4399... Training loss: 1.8428... 0.0932 sec/batch
Epoch: 2/20... Training Step: 4400... Training loss: 1.7744... 0.1019 sec/batch
Epoch: 2/20... Training Step: 4401... Training loss: 1.7005... 0.0716 sec/batch
Epoch: 2/20... Training Step: 4402... Training loss: 1.7947... 0.0697 sec/batch
Epoch: 2/20... Training Step: 4403... Training loss: 1.6698... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4404... Training loss: 1.7311... 0.0651 sec/batch
Epoch: 2/20... Training Step: 4405... Training loss: 1.6660... 0.0908 sec/batch
Epoch: 2/20... Training Step: 4406... Training loss: 1.7650... 0.0853 sec/batch
Epoch: 2/20... Training Step: 4407... Training loss: 1.8052... 0.0692 sec/batch
Epoch: 2/20... Training Step: 4408... Training loss: 1.7267... 0.0776 sec/batch
Epoch: 2/20... Training Step: 4409... Training loss: 1.6739... 0.0793 sec/batch
Epoch: 2/20... Training Step: 4410... Training loss: 1.7862... 0.0647 sec/batch
Epoch: 2/20... Training Step: 4411... Training loss: 1.9328... 0.0675 sec/batch
Epoch: 2/20... Training Step: 4412... Training loss: 1.7134... 0.0838 sec/batch
Epoch: 2/20... Training Step: 4413... Training loss: 1.8014... 0.0608 sec/batch
Epoch: 2/20... Training Step: 4414... Training loss: 1.7021... 0.1004 sec/batch
Epoch: 2/20... Training Step: 4415... Training loss: 1.8145... 0.0673 sec/batch
Epoch: 2/20... Training Step: 4416... Training loss: 1.8362... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4417... Training loss: 2.0000... 0.0601 sec/batch
Epoch: 2/20... Training Step: 4418... Training loss: 1.6475... 0.0736 sec/batch
Epoch: 2/20... Training Step: 4419... Training loss: 1.6235... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4420... Training loss: 1.8783... 0.0648 sec/batch
Epoch: 2/20... Training Step: 4421... Training loss: 1.7662... 0.0631 sec/batch
Epoch: 2/20... Training Step: 4422... Training loss: 1.7529... 0.0691 sec/batch
Epoch: 2/20... Training Step: 4423... Training loss: 1.6463... 0.0683 sec/batch
Epoch: 2/20... Training Step: 4424... Training loss: 1.8229... 0.0656 sec/batch
Epoch: 2/20... Training Step: 4425... Training loss: 1.7442... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4426... Training loss: 1.6668... 0.0576 sec/batch
Epoch: 2/20... Training Step: 4427... Training loss: 1.7599... 0.0528 sec/batch
Epoch: 2/20... Training Step: 4428... Training loss: 1.8116... 0.0735 sec/batch
Epoch: 2/20... Training Step: 4429... Training loss: 1.9109... 0.0804 sec/batch
Epoch: 2/20... Training Step: 4430... Training loss: 1.8973... 0.0633 sec/batch
Epoch: 2/20... Training Step: 4431... Training loss: 1.6782... 0.0775 sec/batch
Epoch: 2/20... Training Step: 4432... Training loss: 1.8111... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4433... Training loss: 1.7790... 0.0637 sec/batch
Epoch: 2/20... Training Step: 4434... Training loss: 1.6996... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4435... Training loss: 1.7958... 0.0836 sec/batch
Epoch: 2/20... Training Step: 4436... Training loss: 1.7609... 0.0627 sec/batch
Epoch: 2/20... Training Step: 4437... Training loss: 1.7629... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4438... Training loss: 1.7992... 0.0853 sec/batch
Epoch: 2/20... Training Step: 4439... Training loss: 1.6718... 0.0619 sec/batch
Epoch: 2/20... Training Step: 4440... Training loss: 1.8200... 0.0732 sec/batch
Epoch: 2/20... Training Step: 4441... Training loss: 1.7339... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4442... Training loss: 1.7205... 0.0877 sec/batch
Epoch: 2/20... Training Step: 4443... Training loss: 1.8681... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4444... Training loss: 1.8123... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4445... Training loss: 1.7873... 0.0626 sec/batch
Epoch: 2/20... Training Step: 4446... Training loss: 1.6095... 0.0663 sec/batch
Epoch: 2/20... Training Step: 4447... Training loss: 1.7059... 0.0767 sec/batch
Epoch: 2/20... Training Step: 4448... Training loss: 1.7062... 0.0637 sec/batch
Epoch: 2/20... Training Step: 4449... Training loss: 1.7185... 0.0749 sec/batch
Epoch: 2/20... Training Step: 4450... Training loss: 1.8630... 0.0736 sec/batch
Epoch: 2/20... Training Step: 4451... Training loss: 1.7405... 0.0654 sec/batch
Epoch: 2/20... Training Step: 4452... Training loss: 1.7415... 0.0647 sec/batch
Epoch: 2/20... Training Step: 4453... Training loss: 1.7987... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4454... Training loss: 1.7613... 0.0667 sec/batch
Epoch: 2/20... Training Step: 4455... Training loss: 1.8154... 0.0664 sec/batch
Epoch: 2/20... Training Step: 4456... Training loss: 1.9216... 0.0782 sec/batch
Epoch: 2/20... Training Step: 4457... Training loss: 1.9524... 0.0685 sec/batch
Epoch: 2/20... Training Step: 4458... Training loss: 1.8237... 0.0515 sec/batch
Epoch: 2/20... Training Step: 4459... Training loss: 1.8429... 0.0706 sec/batch
Epoch: 2/20... Training Step: 4460... Training loss: 1.8331... 0.0732 sec/batch
Epoch: 2/20... Training Step: 4461... Training loss: 1.6896... 0.0593 sec/batch
Epoch: 2/20... Training Step: 4462... Training loss: 1.7570... 0.0569 sec/batch
Epoch: 2/20... Training Step: 4463... Training loss: 1.9184... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4464... Training loss: 1.8673... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4465... Training loss: 1.6639... 0.0690 sec/batch
Epoch: 2/20... Training Step: 4466... Training loss: 1.6781... 0.0571 sec/batch
Epoch: 2/20... Training Step: 4467... Training loss: 1.6166... 0.0725 sec/batch
Epoch: 2/20... Training Step: 4468... Training loss: 1.7876... 0.0711 sec/batch
Epoch: 2/20... Training Step: 4469... Training loss: 1.6898... 0.0743 sec/batch
Epoch: 2/20... Training Step: 4470... Training loss: 1.8853... 0.0638 sec/batch
Epoch: 2/20... Training Step: 4471... Training loss: 1.6636... 0.0764 sec/batch
Epoch: 2/20... Training Step: 4472... Training loss: 1.8776... 0.0802 sec/batch
Epoch: 2/20... Training Step: 4473... Training loss: 1.8395... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4474... Training loss: 1.7571... 0.0901 sec/batch
Epoch: 2/20... Training Step: 4475... Training loss: 1.7371... 0.0971 sec/batch
Epoch: 2/20... Training Step: 4476... Training loss: 1.7748... 0.0831 sec/batch
Epoch: 2/20... Training Step: 4477... Training loss: 1.8839... 0.0808 sec/batch
Epoch: 2/20... Training Step: 4478... Training loss: 1.8922... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4479... Training loss: 1.8449... 0.0598 sec/batch
Epoch: 2/20... Training Step: 4480... Training loss: 1.7993... 0.0805 sec/batch
Epoch: 2/20... Training Step: 4481... Training loss: 1.7927... 0.0764 sec/batch
Epoch: 2/20... Training Step: 4482... Training loss: 1.9753... 0.0803 sec/batch
Epoch: 2/20... Training Step: 4483... Training loss: 1.7341... 0.1085 sec/batch
Epoch: 2/20... Training Step: 4484... Training loss: 1.6934... 0.0651 sec/batch
Epoch: 2/20... Training Step: 4485... Training loss: 1.8939... 0.0596 sec/batch
Epoch: 2/20... Training Step: 4486... Training loss: 1.8454... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4487... Training loss: 1.8439... 0.0842 sec/batch
Epoch: 2/20... Training Step: 4488... Training loss: 1.7729... 0.0796 sec/batch
Epoch: 2/20... Training Step: 4489... Training loss: 1.7974... 0.0709 sec/batch
Epoch: 2/20... Training Step: 4490... Training loss: 1.6964... 0.0836 sec/batch
Epoch: 2/20... Training Step: 4491... Training loss: 1.7105... 0.0789 sec/batch
Epoch: 2/20... Training Step: 4492... Training loss: 1.8492... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4493... Training loss: 1.7881... 0.0894 sec/batch
Epoch: 2/20... Training Step: 4494... Training loss: 1.6808... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4495... Training loss: 1.7556... 0.0866 sec/batch
Epoch: 2/20... Training Step: 4496... Training loss: 1.7176... 0.0712 sec/batch
Epoch: 2/20... Training Step: 4497... Training loss: 1.7194... 0.0774 sec/batch
Epoch: 2/20... Training Step: 4498... Training loss: 1.8197... 0.0741 sec/batch
Epoch: 2/20... Training Step: 4499... Training loss: 1.8243... 0.0596 sec/batch
Epoch: 2/20... Training Step: 4500... Training loss: 1.7802... 0.0625 sec/batch
Epoch: 2/20... Training Step: 4501... Training loss: 1.8313... 0.0750 sec/batch
Epoch: 2/20... Training Step: 4502... Training loss: 1.8213... 0.0752 sec/batch
Epoch: 2/20... Training Step: 4503... Training loss: 1.7188... 0.0767 sec/batch
Epoch: 2/20... Training Step: 4504... Training loss: 1.7036... 0.0651 sec/batch
Epoch: 2/20... Training Step: 4505... Training loss: 1.7892... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4506... Training loss: 1.7489... 0.0943 sec/batch
Epoch: 2/20... Training Step: 4507... Training loss: 1.5680... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4508... Training loss: 1.5999... 0.0710 sec/batch
Epoch: 2/20... Training Step: 4509... Training loss: 1.6792... 0.0827 sec/batch
Epoch: 2/20... Training Step: 4510... Training loss: 1.6303... 0.0644 sec/batch
Epoch: 2/20... Training Step: 4511... Training loss: 1.6765... 0.0710 sec/batch
Epoch: 2/20... Training Step: 4512... Training loss: 1.7894... 0.0678 sec/batch
Epoch: 2/20... Training Step: 4513... Training loss: 1.7992... 0.0811 sec/batch
Epoch: 2/20... Training Step: 4514... Training loss: 1.6935... 0.1039 sec/batch
Epoch: 2/20... Training Step: 4515... Training loss: 1.7626... 0.0791 sec/batch
Epoch: 2/20... Training Step: 4516... Training loss: 1.6844... 0.0744 sec/batch
Epoch: 2/20... Training Step: 4517... Training loss: 1.7285... 0.0929 sec/batch
Epoch: 2/20... Training Step: 4518... Training loss: 1.6708... 0.0776 sec/batch
Epoch: 2/20... Training Step: 4519... Training loss: 1.6803... 0.0669 sec/batch
Epoch: 2/20... Training Step: 4520... Training loss: 1.6209... 0.0739 sec/batch
Epoch: 2/20... Training Step: 4521... Training loss: 1.6792... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4522... Training loss: 1.8242... 0.0590 sec/batch
Epoch: 2/20... Training Step: 4523... Training loss: 1.6825... 0.0658 sec/batch
Epoch: 2/20... Training Step: 4524... Training loss: 1.7420... 0.0791 sec/batch
Epoch: 2/20... Training Step: 4525... Training loss: 1.7898... 0.0858 sec/batch
Epoch: 2/20... Training Step: 4526... Training loss: 1.7740... 0.0692 sec/batch
Epoch: 2/20... Training Step: 4527... Training loss: 1.7222... 0.0713 sec/batch
Epoch: 2/20... Training Step: 4528... Training loss: 1.6587... 0.0577 sec/batch
Epoch: 2/20... Training Step: 4529... Training loss: 1.7858... 0.0866 sec/batch
Epoch: 2/20... Training Step: 4530... Training loss: 1.7810... 0.0746 sec/batch
Epoch: 2/20... Training Step: 4531... Training loss: 1.7466... 0.0610 sec/batch
Epoch: 2/20... Training Step: 4532... Training loss: 1.7871... 0.0557 sec/batch
Epoch: 2/20... Training Step: 4533... Training loss: 1.6689... 0.0704 sec/batch
Epoch: 2/20... Training Step: 4534... Training loss: 1.7032... 0.0809 sec/batch
Epoch: 2/20... Training Step: 4535... Training loss: 1.6960... 0.0665 sec/batch
Epoch: 2/20... Training Step: 4536... Training loss: 1.7217... 0.0585 sec/batch
Epoch: 2/20... Training Step: 4537... Training loss: 1.7086... 0.0767 sec/batch
Epoch: 2/20... Training Step: 4538... Training loss: 1.8460... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4539... Training loss: 1.7887... 0.0622 sec/batch
Epoch: 2/20... Training Step: 4540... Training loss: 1.8181... 0.0730 sec/batch
Epoch: 2/20... Training Step: 4541... Training loss: 1.7693... 0.0717 sec/batch
Epoch: 2/20... Training Step: 4542... Training loss: 1.7494... 0.0853 sec/batch
Epoch: 2/20... Training Step: 4543... Training loss: 1.6713... 0.0853 sec/batch
Epoch: 2/20... Training Step: 4544... Training loss: 1.7128... 0.0820 sec/batch
Epoch: 2/20... Training Step: 4545... Training loss: 1.6395... 0.0784 sec/batch
Epoch: 2/20... Training Step: 4546... Training loss: 1.7095... 0.0778 sec/batch
Epoch: 2/20... Training Step: 4547... Training loss: 1.7964... 0.0693 sec/batch
Epoch: 2/20... Training Step: 4548... Training loss: 1.8239... 0.0612 sec/batch
Epoch: 2/20... Training Step: 4549... Training loss: 1.7524... 0.0688 sec/batch
Epoch: 2/20... Training Step: 4550... Training loss: 1.9195... 0.0765 sec/batch
Epoch: 2/20... Training Step: 4551... Training loss: 1.8596... 0.0806 sec/batch
Epoch: 2/20... Training Step: 4552... Training loss: 1.6508... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4553... Training loss: 1.7571... 0.0714 sec/batch
Epoch: 2/20... Training Step: 4554... Training loss: 1.8010... 0.0634 sec/batch
Epoch: 2/20... Training Step: 4555... Training loss: 1.5916... 0.0974 sec/batch
Epoch: 2/20... Training Step: 4556... Training loss: 1.7844... 0.0678 sec/batch
Epoch: 2/20... Training Step: 4557... Training loss: 1.5855... 0.0578 sec/batch
Epoch: 2/20... Training Step: 4558... Training loss: 1.7840... 0.0537 sec/batch
Epoch: 2/20... Training Step: 4559... Training loss: 2.0527... 0.0622 sec/batch
Epoch: 2/20... Training Step: 4560... Training loss: 1.8803... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4561... Training loss: 1.8640... 0.0800 sec/batch
Epoch: 2/20... Training Step: 4562... Training loss: 1.6798... 0.0604 sec/batch
Epoch: 2/20... Training Step: 4563... Training loss: 1.7207... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4564... Training loss: 1.8835... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4565... Training loss: 1.9587... 0.0616 sec/batch
Epoch: 2/20... Training Step: 4566... Training loss: 1.8528... 0.0678 sec/batch
Epoch: 2/20... Training Step: 4567... Training loss: 1.6572... 0.0673 sec/batch
Epoch: 2/20... Training Step: 4568... Training loss: 1.8324... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4569... Training loss: 1.7680... 0.0782 sec/batch
Epoch: 2/20... Training Step: 4570... Training loss: 1.6606... 0.0775 sec/batch
Epoch: 2/20... Training Step: 4571... Training loss: 1.9204... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4572... Training loss: 1.6581... 0.0793 sec/batch
Epoch: 2/20... Training Step: 4573... Training loss: 1.7717... 0.0783 sec/batch
Epoch: 2/20... Training Step: 4574... Training loss: 1.6665... 0.0577 sec/batch
Epoch: 2/20... Training Step: 4575... Training loss: 1.7413... 0.0797 sec/batch
Epoch: 2/20... Training Step: 4576... Training loss: 1.7569... 0.0794 sec/batch
Epoch: 2/20... Training Step: 4577... Training loss: 1.6549... 0.0567 sec/batch
Epoch: 2/20... Training Step: 4578... Training loss: 1.8188... 0.0762 sec/batch
Epoch: 2/20... Training Step: 4579... Training loss: 1.8681... 0.0674 sec/batch
Epoch: 2/20... Training Step: 4580... Training loss: 1.7921... 0.0549 sec/batch
Epoch: 2/20... Training Step: 4581... Training loss: 1.8291... 0.0693 sec/batch
Epoch: 2/20... Training Step: 4582... Training loss: 1.6984... 0.0734 sec/batch
Epoch: 2/20... Training Step: 4583... Training loss: 1.7588... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4584... Training loss: 1.8757... 0.0578 sec/batch
Epoch: 2/20... Training Step: 4585... Training loss: 1.7226... 0.0645 sec/batch
Epoch: 2/20... Training Step: 4586... Training loss: 1.7550... 0.0756 sec/batch
Epoch: 2/20... Training Step: 4587... Training loss: 1.7701... 0.0586 sec/batch
Epoch: 2/20... Training Step: 4588... Training loss: 1.6484... 0.0539 sec/batch
Epoch: 2/20... Training Step: 4589... Training loss: 1.6419... 0.0594 sec/batch
Epoch: 2/20... Training Step: 4590... Training loss: 1.6130... 0.0732 sec/batch
Epoch: 2/20... Training Step: 4591... Training loss: 1.6587... 0.0682 sec/batch
Epoch: 2/20... Training Step: 4592... Training loss: 1.7392... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4593... Training loss: 1.7851... 0.0738 sec/batch
Epoch: 2/20... Training Step: 4594... Training loss: 1.8056... 0.0697 sec/batch
Epoch: 2/20... Training Step: 4595... Training loss: 1.6606... 0.0659 sec/batch
Epoch: 2/20... Training Step: 4596... Training loss: 1.7621... 0.0716 sec/batch
Epoch: 2/20... Training Step: 4597... Training loss: 1.6920... 0.0676 sec/batch
Epoch: 2/20... Training Step: 4598... Training loss: 1.6485... 0.0586 sec/batch
Epoch: 2/20... Training Step: 4599... Training loss: 1.7868... 0.0721 sec/batch
Epoch: 2/20... Training Step: 4600... Training loss: 1.8570... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4601... Training loss: 1.8369... 0.0573 sec/batch
Epoch: 2/20... Training Step: 4602... Training loss: 1.5960... 0.0663 sec/batch
Epoch: 2/20... Training Step: 4603... Training loss: 1.7077... 0.0690 sec/batch
Epoch: 2/20... Training Step: 4604... Training loss: 1.8123... 0.0732 sec/batch
Epoch: 2/20... Training Step: 4605... Training loss: 1.9387... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4606... Training loss: 1.8313... 0.0591 sec/batch
Epoch: 2/20... Training Step: 4607... Training loss: 1.7637... 0.0800 sec/batch
Epoch: 2/20... Training Step: 4608... Training loss: 1.7538... 0.0560 sec/batch
Epoch: 2/20... Training Step: 4609... Training loss: 1.7863... 0.0711 sec/batch
Epoch: 2/20... Training Step: 4610... Training loss: 1.6743... 0.0568 sec/batch
Epoch: 2/20... Training Step: 4611... Training loss: 1.7541... 0.0796 sec/batch
Epoch: 2/20... Training Step: 4612... Training loss: 1.7151... 0.0539 sec/batch
Epoch: 2/20... Training Step: 4613... Training loss: 1.7498... 0.0796 sec/batch
Epoch: 2/20... Training Step: 4614... Training loss: 1.6098... 0.0727 sec/batch
Epoch: 2/20... Training Step: 4615... Training loss: 1.7405... 0.0709 sec/batch
Epoch: 2/20... Training Step: 4616... Training loss: 1.9092... 0.0656 sec/batch
Epoch: 2/20... Training Step: 4617... Training loss: 1.7437... 0.0626 sec/batch
Epoch: 2/20... Training Step: 4618... Training loss: 1.8061... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4619... Training loss: 1.7948... 0.0623 sec/batch
Epoch: 2/20... Training Step: 4620... Training loss: 1.6465... 0.0745 sec/batch
Epoch: 2/20... Training Step: 4621... Training loss: 1.7553... 0.0538 sec/batch
Epoch: 2/20... Training Step: 4622... Training loss: 1.9318... 0.0598 sec/batch
Epoch: 2/20... Training Step: 4623... Training loss: 1.8699... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4624... Training loss: 1.7063... 0.0765 sec/batch
Epoch: 2/20... Training Step: 4625... Training loss: 1.8057... 0.0744 sec/batch
Epoch: 2/20... Training Step: 4626... Training loss: 1.8157... 0.0565 sec/batch
Epoch: 2/20... Training Step: 4627... Training loss: 1.6783... 0.0684 sec/batch
Epoch: 2/20... Training Step: 4628... Training loss: 1.8642... 0.0558 sec/batch
Epoch: 2/20... Training Step: 4629... Training loss: 1.7770... 0.0697 sec/batch
Epoch: 2/20... Training Step: 4630... Training loss: 1.8014... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4631... Training loss: 1.7623... 0.0576 sec/batch
Epoch: 2/20... Training Step: 4632... Training loss: 1.7755... 0.0580 sec/batch
Epoch: 2/20... Training Step: 4633... Training loss: 1.8075... 0.0673 sec/batch
Epoch: 2/20... Training Step: 4634... Training loss: 1.8817... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4635... Training loss: 1.8435... 0.0798 sec/batch
Epoch: 2/20... Training Step: 4636... Training loss: 1.8569... 0.0926 sec/batch
Epoch: 2/20... Training Step: 4637... Training loss: 1.8728... 0.0882 sec/batch
Epoch: 2/20... Training Step: 4638... Training loss: 1.6382... 0.0759 sec/batch
Epoch: 2/20... Training Step: 4639... Training loss: 1.6883... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4640... Training loss: 1.7618... 0.0607 sec/batch
Epoch: 2/20... Training Step: 4641... Training loss: 1.7723... 0.0655 sec/batch
Epoch: 2/20... Training Step: 4642... Training loss: 1.7709... 0.0782 sec/batch
Epoch: 2/20... Training Step: 4643... Training loss: 1.8692... 0.0525 sec/batch
Epoch: 2/20... Training Step: 4644... Training loss: 1.8715... 0.0784 sec/batch
Epoch: 2/20... Training Step: 4645... Training loss: 1.7503... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4646... Training loss: 1.7519... 0.0590 sec/batch
Epoch: 2/20... Training Step: 4647... Training loss: 1.9704... 0.0611 sec/batch
Epoch: 2/20... Training Step: 4648... Training loss: 1.8680... 0.0760 sec/batch
Epoch: 2/20... Training Step: 4649... Training loss: 1.7659... 0.0648 sec/batch
Epoch: 2/20... Training Step: 4650... Training loss: 1.8728... 0.0735 sec/batch
Epoch: 2/20... Training Step: 4651... Training loss: 1.7619... 0.0545 sec/batch
Epoch: 2/20... Training Step: 4652... Training loss: 1.6999... 0.0702 sec/batch
Epoch: 2/20... Training Step: 4653... Training loss: 1.7717... 0.0762 sec/batch
Epoch: 2/20... Training Step: 4654... Training loss: 1.6951... 0.0824 sec/batch
Epoch: 2/20... Training Step: 4655... Training loss: 1.6898... 0.0621 sec/batch
Epoch: 2/20... Training Step: 4656... Training loss: 1.7388... 0.0597 sec/batch
Epoch: 2/20... Training Step: 4657... Training loss: 1.8144... 0.0733 sec/batch
Epoch: 2/20... Training Step: 4658... Training loss: 1.6105... 0.0680 sec/batch
Epoch: 2/20... Training Step: 4659... Training loss: 1.7150... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4660... Training loss: 1.7286... 0.0950 sec/batch
Epoch: 2/20... Training Step: 4661... Training loss: 1.8536... 0.0754 sec/batch
Epoch: 2/20... Training Step: 4662... Training loss: 1.8531... 0.0784 sec/batch
Epoch: 2/20... Training Step: 4663... Training loss: 1.7817... 0.0734 sec/batch
Epoch: 2/20... Training Step: 4664... Training loss: 1.8247... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4665... Training loss: 1.7273... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4666... Training loss: 1.8606... 0.0858 sec/batch
Epoch: 2/20... Training Step: 4667... Training loss: 1.8366... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4668... Training loss: 1.8148... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4669... Training loss: 1.8668... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4670... Training loss: 1.6983... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4671... Training loss: 1.8523... 0.0692 sec/batch
Epoch: 2/20... Training Step: 4672... Training loss: 1.7584... 0.0660 sec/batch
Epoch: 2/20... Training Step: 4673... Training loss: 1.7880... 0.0648 sec/batch
Epoch: 2/20... Training Step: 4674... Training loss: 1.7578... 0.0542 sec/batch
Epoch: 2/20... Training Step: 4675... Training loss: 1.7333... 0.0800 sec/batch
Epoch: 2/20... Training Step: 4676... Training loss: 1.7903... 0.0863 sec/batch
Epoch: 2/20... Training Step: 4677... Training loss: 1.8228... 0.0631 sec/batch
Epoch: 2/20... Training Step: 4678... Training loss: 1.7600... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4679... Training loss: 1.8196... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4680... Training loss: 1.8229... 0.0584 sec/batch
Epoch: 2/20... Training Step: 4681... Training loss: 1.7176... 0.0663 sec/batch
Epoch: 2/20... Training Step: 4682... Training loss: 1.7832... 0.0807 sec/batch
Epoch: 2/20... Training Step: 4683... Training loss: 1.8490... 0.0689 sec/batch
Epoch: 2/20... Training Step: 4684... Training loss: 1.6877... 0.0542 sec/batch
Epoch: 2/20... Training Step: 4685... Training loss: 1.7295... 0.0696 sec/batch
Epoch: 2/20... Training Step: 4686... Training loss: 1.6633... 0.0694 sec/batch
Epoch: 2/20... Training Step: 4687... Training loss: 1.8325... 0.0767 sec/batch
Epoch: 2/20... Training Step: 4688... Training loss: 1.6515... 0.0733 sec/batch
Epoch: 2/20... Training Step: 4689... Training loss: 1.7204... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4690... Training loss: 1.8495... 0.0557 sec/batch
Epoch: 2/20... Training Step: 4691... Training loss: 1.7577... 0.0685 sec/batch
Epoch: 2/20... Training Step: 4692... Training loss: 1.6086... 0.0724 sec/batch
Epoch: 2/20... Training Step: 4693... Training loss: 1.7728... 0.0676 sec/batch
Epoch: 2/20... Training Step: 4694... Training loss: 1.8156... 0.0600 sec/batch
Epoch: 2/20... Training Step: 4695... Training loss: 1.6998... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4696... Training loss: 1.6980... 0.0804 sec/batch
Epoch: 2/20... Training Step: 4697... Training loss: 1.6871... 0.0713 sec/batch
Epoch: 2/20... Training Step: 4698... Training loss: 1.7014... 0.0644 sec/batch
Epoch: 2/20... Training Step: 4699... Training loss: 1.7046... 0.0619 sec/batch
Epoch: 2/20... Training Step: 4700... Training loss: 1.6876... 0.0659 sec/batch
Epoch: 2/20... Training Step: 4701... Training loss: 1.7515... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4702... Training loss: 1.7056... 0.0769 sec/batch
Epoch: 2/20... Training Step: 4703... Training loss: 1.7383... 0.0658 sec/batch
Epoch: 2/20... Training Step: 4704... Training loss: 1.6786... 0.0681 sec/batch
Epoch: 2/20... Training Step: 4705... Training loss: 1.6228... 0.0679 sec/batch
Epoch: 2/20... Training Step: 4706... Training loss: 1.8872... 0.0528 sec/batch
Epoch: 2/20... Training Step: 4707... Training loss: 1.7587... 0.0584 sec/batch
Epoch: 2/20... Training Step: 4708... Training loss: 1.8022... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4709... Training loss: 1.7732... 0.0681 sec/batch
Epoch: 2/20... Training Step: 4710... Training loss: 1.7155... 0.0602 sec/batch
Epoch: 2/20... Training Step: 4711... Training loss: 1.7481... 0.0644 sec/batch
Epoch: 2/20... Training Step: 4712... Training loss: 1.5945... 0.0642 sec/batch
Epoch: 2/20... Training Step: 4713... Training loss: 1.6050... 0.0663 sec/batch
Epoch: 2/20... Training Step: 4714... Training loss: 1.7157... 0.0528 sec/batch
Epoch: 2/20... Training Step: 4715... Training loss: 1.7739... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4716... Training loss: 1.8185... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4717... Training loss: 1.7922... 0.0605 sec/batch
Epoch: 2/20... Training Step: 4718... Training loss: 1.7361... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4719... Training loss: 1.7330... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4720... Training loss: 1.7978... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4721... Training loss: 1.7055... 0.0540 sec/batch
Epoch: 2/20... Training Step: 4722... Training loss: 2.0226... 0.0572 sec/batch
Epoch: 2/20... Training Step: 4723... Training loss: 1.8428... 0.0511 sec/batch
Epoch: 2/20... Training Step: 4724... Training loss: 1.6663... 0.0701 sec/batch
Epoch: 2/20... Training Step: 4725... Training loss: 1.6803... 0.0582 sec/batch
Epoch: 2/20... Training Step: 4726... Training loss: 1.6672... 0.0612 sec/batch
Epoch: 2/20... Training Step: 4727... Training loss: 1.7215... 0.0726 sec/batch
Epoch: 2/20... Training Step: 4728... Training loss: 1.6964... 0.0691 sec/batch
Epoch: 2/20... Training Step: 4729... Training loss: 1.7404... 0.0499 sec/batch
Epoch: 2/20... Training Step: 4730... Training loss: 1.6866... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4731... Training loss: 1.6882... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4732... Training loss: 1.9000... 0.0651 sec/batch
Epoch: 2/20... Training Step: 4733... Training loss: 1.7932... 0.0599 sec/batch
Epoch: 2/20... Training Step: 4734... Training loss: 1.7454... 0.0595 sec/batch
Epoch: 2/20... Training Step: 4735... Training loss: 1.7376... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4736... Training loss: 1.7923... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4737... Training loss: 1.7213... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4738... Training loss: 1.7557... 0.0727 sec/batch
Epoch: 2/20... Training Step: 4739... Training loss: 1.8910... 0.0577 sec/batch
Epoch: 2/20... Training Step: 4740... Training loss: 1.7408... 0.0488 sec/batch
Epoch: 2/20... Training Step: 4741... Training loss: 1.7697... 0.0560 sec/batch
Epoch: 2/20... Training Step: 4742... Training loss: 1.6696... 0.0606 sec/batch
Epoch: 2/20... Training Step: 4743... Training loss: 1.8264... 0.0549 sec/batch
Epoch: 2/20... Training Step: 4744... Training loss: 1.9466... 0.0672 sec/batch
Epoch: 2/20... Training Step: 4745... Training loss: 1.8144... 0.0597 sec/batch
Epoch: 2/20... Training Step: 4746... Training loss: 1.7788... 0.0728 sec/batch
Epoch: 2/20... Training Step: 4747... Training loss: 1.8632... 0.0797 sec/batch
Epoch: 2/20... Training Step: 4748... Training loss: 1.9270... 0.0530 sec/batch
Epoch: 2/20... Training Step: 4749... Training loss: 1.7447... 0.0708 sec/batch
Epoch: 2/20... Training Step: 4750... Training loss: 1.8303... 0.0756 sec/batch
Epoch: 2/20... Training Step: 4751... Training loss: 1.7162... 0.0754 sec/batch
Epoch: 2/20... Training Step: 4752... Training loss: 1.8212... 0.0897 sec/batch
Epoch: 2/20... Training Step: 4753... Training loss: 1.7680... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4754... Training loss: 1.6818... 0.0723 sec/batch
Epoch: 2/20... Training Step: 4755... Training loss: 1.7729... 0.0632 sec/batch
Epoch: 2/20... Training Step: 4756... Training loss: 1.8105... 0.0658 sec/batch
Epoch: 2/20... Training Step: 4757... Training loss: 1.7979... 0.0751 sec/batch
Epoch: 2/20... Training Step: 4758... Training loss: 1.5702... 0.0602 sec/batch
Epoch: 2/20... Training Step: 4759... Training loss: 1.8185... 0.0764 sec/batch
Epoch: 2/20... Training Step: 4760... Training loss: 1.7645... 0.0746 sec/batch
Epoch: 2/20... Training Step: 4761... Training loss: 1.8051... 0.0690 sec/batch
Epoch: 2/20... Training Step: 4762... Training loss: 1.6397... 0.0770 sec/batch
Epoch: 2/20... Training Step: 4763... Training loss: 1.7892... 0.0612 sec/batch
Epoch: 2/20... Training Step: 4764... Training loss: 1.7150... 0.0571 sec/batch
Epoch: 2/20... Training Step: 4765... Training loss: 1.6712... 0.0785 sec/batch
Epoch: 2/20... Training Step: 4766... Training loss: 1.6678... 0.0721 sec/batch
Epoch: 2/20... Training Step: 4767... Training loss: 1.8295... 0.0637 sec/batch
Epoch: 2/20... Training Step: 4768... Training loss: 1.6196... 0.0747 sec/batch
Epoch: 2/20... Training Step: 4769... Training loss: 1.7409... 0.0764 sec/batch
Epoch: 2/20... Training Step: 4770... Training loss: 1.7536... 0.0623 sec/batch
Epoch: 2/20... Training Step: 4771... Training loss: 1.7523... 0.0626 sec/batch
Epoch: 2/20... Training Step: 4772... Training loss: 1.6605... 0.0619 sec/batch
Epoch: 2/20... Training Step: 4773... Training loss: 1.7602... 0.0698 sec/batch
Epoch: 2/20... Training Step: 4774... Training loss: 1.9061... 0.0724 sec/batch
Epoch: 2/20... Training Step: 4775... Training loss: 1.9084... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4776... Training loss: 1.8215... 0.0551 sec/batch
Epoch: 2/20... Training Step: 4777... Training loss: 1.7727... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4778... Training loss: 1.6854... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4779... Training loss: 1.8254... 0.0760 sec/batch
Epoch: 2/20... Training Step: 4780... Training loss: 1.7124... 0.0838 sec/batch
Epoch: 2/20... Training Step: 4781... Training loss: 1.7114... 0.0795 sec/batch
Epoch: 2/20... Training Step: 4782... Training loss: 1.7409... 0.0676 sec/batch
Epoch: 2/20... Training Step: 4783... Training loss: 1.7231... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4784... Training loss: 1.7447... 0.0748 sec/batch
Epoch: 2/20... Training Step: 4785... Training loss: 1.6869... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4786... Training loss: 1.6479... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4787... Training loss: 1.6615... 0.0639 sec/batch
Epoch: 2/20... Training Step: 4788... Training loss: 1.6416... 0.0722 sec/batch
Epoch: 2/20... Training Step: 4789... Training loss: 1.6672... 0.0685 sec/batch
Epoch: 2/20... Training Step: 4790... Training loss: 1.7848... 0.0817 sec/batch
Epoch: 2/20... Training Step: 4791... Training loss: 1.7666... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4792... Training loss: 1.8775... 0.0572 sec/batch
Epoch: 2/20... Training Step: 4793... Training loss: 1.7205... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4794... Training loss: 1.6740... 0.0682 sec/batch
Epoch: 2/20... Training Step: 4795... Training loss: 1.6448... 0.0695 sec/batch
Epoch: 2/20... Training Step: 4796... Training loss: 1.6981... 0.0710 sec/batch
Epoch: 2/20... Training Step: 4797... Training loss: 1.7672... 0.0683 sec/batch
Epoch: 2/20... Training Step: 4798... Training loss: 1.7898... 0.0766 sec/batch
Epoch: 2/20... Training Step: 4799... Training loss: 1.6863... 0.0669 sec/batch
Epoch: 2/20... Training Step: 4800... Training loss: 1.6516... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4801... Training loss: 1.8484... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4802... Training loss: 1.7575... 0.0782 sec/batch
Epoch: 2/20... Training Step: 4803... Training loss: 1.6307... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4804... Training loss: 1.7349... 0.0654 sec/batch
Epoch: 2/20... Training Step: 4805... Training loss: 1.8500... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4806... Training loss: 1.6270... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4807... Training loss: 1.7361... 0.0696 sec/batch
Epoch: 2/20... Training Step: 4808... Training loss: 1.6177... 0.0633 sec/batch
Epoch: 2/20... Training Step: 4809... Training loss: 1.8332... 0.0656 sec/batch
Epoch: 2/20... Training Step: 4810... Training loss: 1.7045... 0.0638 sec/batch
Epoch: 2/20... Training Step: 4811... Training loss: 1.7732... 0.0616 sec/batch
Epoch: 2/20... Training Step: 4812... Training loss: 1.7389... 0.0843 sec/batch
Epoch: 2/20... Training Step: 4813... Training loss: 1.7342... 0.0570 sec/batch
Epoch: 2/20... Training Step: 4814... Training loss: 1.7529... 0.0714 sec/batch
Epoch: 2/20... Training Step: 4815... Training loss: 1.7113... 0.0552 sec/batch
Epoch: 2/20... Training Step: 4816... Training loss: 1.7958... 0.0647 sec/batch
Epoch: 2/20... Training Step: 4817... Training loss: 1.8086... 0.0676 sec/batch
Epoch: 2/20... Training Step: 4818... Training loss: 1.7797... 0.0573 sec/batch
Epoch: 2/20... Training Step: 4819... Training loss: 1.6518... 0.0637 sec/batch
Epoch: 2/20... Training Step: 4820... Training loss: 1.8456... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4821... Training loss: 1.9327... 0.0642 sec/batch
Epoch: 2/20... Training Step: 4822... Training loss: 1.5713... 0.0664 sec/batch
Epoch: 2/20... Training Step: 4823... Training loss: 1.7546... 0.0559 sec/batch
Epoch: 2/20... Training Step: 4824... Training loss: 1.6272... 0.0560 sec/batch
Epoch: 2/20... Training Step: 4825... Training loss: 1.8784... 0.0694 sec/batch
Epoch: 2/20... Training Step: 4826... Training loss: 1.8205... 0.0722 sec/batch
Epoch: 2/20... Training Step: 4827... Training loss: 1.8637... 0.0641 sec/batch
Epoch: 2/20... Training Step: 4828... Training loss: 1.7278... 0.0742 sec/batch
Epoch: 2/20... Training Step: 4829... Training loss: 1.7863... 0.0757 sec/batch
Epoch: 2/20... Training Step: 4830... Training loss: 1.9242... 0.0557 sec/batch
Epoch: 2/20... Training Step: 4831... Training loss: 1.7733... 0.0609 sec/batch
Epoch: 2/20... Training Step: 4832... Training loss: 1.8210... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4833... Training loss: 1.9426... 0.0565 sec/batch
Epoch: 2/20... Training Step: 4834... Training loss: 1.7450... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4835... Training loss: 1.8950... 0.0580 sec/batch
Epoch: 2/20... Training Step: 4836... Training loss: 2.0211... 0.0606 sec/batch
Epoch: 2/20... Training Step: 4837... Training loss: 1.9006... 0.0627 sec/batch
Epoch: 2/20... Training Step: 4838... Training loss: 1.8581... 0.0548 sec/batch
Epoch: 2/20... Training Step: 4839... Training loss: 1.8075... 0.0805 sec/batch
Epoch: 2/20... Training Step: 4840... Training loss: 1.8417... 0.0655 sec/batch
Epoch: 2/20... Training Step: 4841... Training loss: 1.6890... 0.0581 sec/batch
Epoch: 2/20... Training Step: 4842... Training loss: 1.6691... 0.0649 sec/batch
Epoch: 2/20... Training Step: 4843... Training loss: 1.7560... 0.0719 sec/batch
Epoch: 2/20... Training Step: 4844... Training loss: 1.6956... 0.0657 sec/batch
Epoch: 2/20... Training Step: 4845... Training loss: 1.8879... 0.0593 sec/batch
Epoch: 2/20... Training Step: 4846... Training loss: 1.7541... 0.0661 sec/batch
Epoch: 2/20... Training Step: 4847... Training loss: 1.7340... 0.0645 sec/batch
Epoch: 2/20... Training Step: 4848... Training loss: 1.9203... 0.0565 sec/batch
Epoch: 2/20... Training Step: 4849... Training loss: 1.6724... 0.0629 sec/batch
Epoch: 2/20... Training Step: 4850... Training loss: 1.6558... 0.0628 sec/batch
Epoch: 2/20... Training Step: 4851... Training loss: 2.0105... 0.0493 sec/batch
Epoch: 2/20... Training Step: 4852... Training loss: 1.9157... 0.0641 sec/batch
Epoch: 2/20... Training Step: 4853... Training loss: 1.8113... 0.0682 sec/batch
Epoch: 2/20... Training Step: 4854... Training loss: 1.8074... 0.0702 sec/batch
Epoch: 2/20... Training Step: 4855... Training loss: 1.6482... 0.0566 sec/batch
Epoch: 2/20... Training Step: 4856... Training loss: 1.7630... 0.0704 sec/batch
Epoch: 2/20... Training Step: 4857... Training loss: 1.8173... 0.0728 sec/batch
Epoch: 2/20... Training Step: 4858... Training loss: 1.6066... 0.0650 sec/batch
Epoch: 2/20... Training Step: 4859... Training loss: 1.6779... 0.0518 sec/batch
Epoch: 2/20... Training Step: 4860... Training loss: 1.7827... 0.0586 sec/batch
Epoch: 2/20... Training Step: 4861... Training loss: 1.7588... 0.0622 sec/batch
Epoch: 2/20... Training Step: 4862... Training loss: 1.8360... 0.0843 sec/batch
Epoch: 2/20... Training Step: 4863... Training loss: 1.8996... 0.0577 sec/batch
Epoch: 2/20... Training Step: 4864... Training loss: 1.8123... 0.0520 sec/batch
Epoch: 2/20... Training Step: 4865... Training loss: 1.7940... 0.0575 sec/batch
Epoch: 2/20... Training Step: 4866... Training loss: 1.7520... 0.0731 sec/batch
Epoch: 2/20... Training Step: 4867... Training loss: 1.9612... 0.0599 sec/batch
Epoch: 2/20... Training Step: 4868... Training loss: 1.6945... 0.0608 sec/batch
Epoch: 2/20... Training Step: 4869... Training loss: 1.8486... 0.0620 sec/batch
Epoch: 2/20... Training Step: 4870... Training loss: 1.7399... 0.0669 sec/batch
Epoch: 2/20... Training Step: 4871... Training loss: 1.6300... 0.0833 sec/batch
Epoch: 2/20... Training Step: 4872... Training loss: 1.6336... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4873... Training loss: 1.7257... 0.0757 sec/batch
Epoch: 2/20... Training Step: 4874... Training loss: 1.7045... 0.0497 sec/batch
Epoch: 2/20... Training Step: 4875... Training loss: 1.7226... 0.0711 sec/batch
Epoch: 2/20... Training Step: 4876... Training loss: 1.8422... 0.0562 sec/batch
Epoch: 2/20... Training Step: 4877... Training loss: 1.8956... 0.0631 sec/batch
Epoch: 2/20... Training Step: 4878... Training loss: 1.9004... 0.0621 sec/batch
Epoch: 2/20... Training Step: 4879... Training loss: 1.9085... 0.0869 sec/batch
Epoch: 2/20... Training Step: 4880... Training loss: 1.8317... 0.0791 sec/batch
Epoch: 2/20... Training Step: 4881... Training loss: 1.7472... 0.0784 sec/batch
Epoch: 2/20... Training Step: 4882... Training loss: 1.7467... 0.0581 sec/batch
Epoch: 2/20... Training Step: 4883... Training loss: 1.8868... 0.0773 sec/batch
Epoch: 2/20... Training Step: 4884... Training loss: 1.7720... 0.0706 sec/batch
Epoch: 2/20... Training Step: 4885... Training loss: 1.7380... 0.0596 sec/batch
Epoch: 2/20... Training Step: 4886... Training loss: 1.7281... 0.0646 sec/batch
Epoch: 2/20... Training Step: 4887... Training loss: 1.8148... 0.0610 sec/batch
Epoch: 2/20... Training Step: 4888... Training loss: 1.6182... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4889... Training loss: 1.6709... 0.0797 sec/batch
Epoch: 2/20... Training Step: 4890... Training loss: 1.6713... 0.0667 sec/batch
Epoch: 2/20... Training Step: 4891... Training loss: 1.6596... 0.0643 sec/batch
Epoch: 2/20... Training Step: 4892... Training loss: 1.7422... 0.0585 sec/batch
Epoch: 2/20... Training Step: 4893... Training loss: 1.8286... 0.0796 sec/batch
Epoch: 2/20... Training Step: 4894... Training loss: 1.7499... 0.0675 sec/batch
Epoch: 2/20... Training Step: 4895... Training loss: 1.7481... 0.0607 sec/batch
Epoch: 2/20... Training Step: 4896... Training loss: 1.7380... 0.0655 sec/batch
Epoch: 2/20... Training Step: 4897... Training loss: 1.7370... 0.0656 sec/batch
Epoch: 2/20... Training Step: 4898... Training loss: 1.6509... 0.0917 sec/batch
Epoch: 2/20... Training Step: 4899... Training loss: 1.6956... 0.0754 sec/batch
Epoch: 2/20... Training Step: 4900... Training loss: 1.7604... 0.0588 sec/batch
Epoch: 2/20... Training Step: 4901... Training loss: 1.7468... 0.0603 sec/batch
Epoch: 2/20... Training Step: 4902... Training loss: 1.6067... 0.0720 sec/batch
Epoch: 2/20... Training Step: 4903... Training loss: 1.7137... 0.0660 sec/batch
Epoch: 2/20... Training Step: 4904... Training loss: 1.7523... 0.0653 sec/batch
Epoch: 2/20... Training Step: 4905... Training loss: 1.7120... 0.0845 sec/batch
Epoch: 2/20... Training Step: 4906... Training loss: 1.7050... 0.0743 sec/batch
Epoch: 2/20... Training Step: 4907... Training loss: 1.6452... 0.0614 sec/batch
Epoch: 2/20... Training Step: 4908... Training loss: 1.6542... 0.0683 sec/batch
Epoch: 2/20... Training Step: 4909... Training loss: 1.6783... 0.0684 sec/batch
Epoch: 2/20... Training Step: 4910... Training loss: 1.7479... 0.0668 sec/batch
Epoch: 2/20... Training Step: 4911... Training loss: 1.6975... 0.0664 sec/batch
Epoch: 2/20... Training Step: 4912... Training loss: 1.6374... 0.0702 sec/batch
Epoch: 2/20... Training Step: 4913... Training loss: 1.6888... 0.0603 sec/batch
Epoch: 2/20... Training Step: 4914... Training loss: 1.8514... 0.0752 sec/batch
Epoch: 2/20... Training Step: 4915... Training loss: 1.7323... 0.0822 sec/batch
Epoch: 2/20... Training Step: 4916... Training loss: 1.6967... 0.0767 sec/batch
Epoch: 2/20... Training Step: 4917... Training loss: 1.6954... 0.0598 sec/batch
Epoch: 2/20... Training Step: 4918... Training loss: 1.7730... 0.0664 sec/batch
Epoch: 2/20... Training Step: 4919... Training loss: 1.8255... 0.0554 sec/batch
Epoch: 2/20... Training Step: 4920... Training loss: 1.7644... 0.0706 sec/batch
Epoch: 2/20... Training Step: 4921... Training loss: 1.7784... 0.0634 sec/batch
Epoch: 2/20... Training Step: 4922... Training loss: 1.8366... 0.0636 sec/batch
Epoch: 2/20... Training Step: 4923... Training loss: 1.7646... 0.0615 sec/batch
Epoch: 2/20... Training Step: 4924... Training loss: 1.7656... 0.0701 sec/batch
Epoch: 2/20... Training Step: 4925... Training loss: 1.7427... 0.0660 sec/batch
Epoch: 2/20... Training Step: 4926... Training loss: 1.6434... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4927... Training loss: 1.8406... 0.0653 sec/batch
Epoch: 2/20... Training Step: 4928... Training loss: 1.7358... 0.0666 sec/batch
Epoch: 2/20... Training Step: 4929... Training loss: 1.6459... 0.0746 sec/batch
Epoch: 2/20... Training Step: 4930... Training loss: 1.8281... 0.0744 sec/batch
Epoch: 2/20... Training Step: 4931... Training loss: 1.6082... 0.0697 sec/batch
Epoch: 2/20... Training Step: 4932... Training loss: 1.7132... 0.0524 sec/batch
Epoch: 2/20... Training Step: 4933... Training loss: 1.6636... 0.0635 sec/batch
Epoch: 2/20... Training Step: 4934... Training loss: 1.6882... 0.0561 sec/batch
Epoch: 2/20... Training Step: 4935... Training loss: 1.8109... 0.0578 sec/batch
Epoch: 2/20... Training Step: 4936... Training loss: 1.8105... 0.0740 sec/batch
Epoch: 2/20... Training Step: 4937... Training loss: 1.6852... 0.0564 sec/batch
Epoch: 2/20... Training Step: 4938... Training loss: 1.7169... 0.0656 sec/batch
Epoch: 2/20... Training Step: 4939... Training loss: 1.6872... 0.0758 sec/batch
Epoch: 2/20... Training Step: 4940... Training loss: 1.6519... 0.0749 sec/batch
Epoch: 2/20... Training Step: 4941... Training loss: 1.6868... 0.0537 sec/batch
Epoch: 2/20... Training Step: 4942... Training loss: 1.8189... 0.0748 sec/batch
Epoch: 2/20... Training Step: 4943... Training loss: 1.7260... 0.0716 sec/batch
Epoch: 2/20... Training Step: 4944... Training loss: 1.7266... 0.0637 sec/batch
Epoch: 2/20... Training Step: 4945... Training loss: 1.6602... 0.0872 sec/batch
Epoch: 2/20... Training Step: 4946... Training loss: 1.8511... 0.0720 sec/batch
Epoch: 2/20... Training Step: 4947... Training loss: 1.9723... 0.0674 sec/batch
Epoch: 2/20... Training Step: 4948... Training loss: 1.6876... 0.0634 sec/batch
Epoch: 2/20... Training Step: 4949... Training loss: 1.8072... 0.0662 sec/batch
Epoch: 2/20... Training Step: 4950... Training loss: 1.7287... 0.0641 sec/batch
Epoch: 2/20... Training Step: 4951... Training loss: 1.8254... 0.0733 sec/batch
Epoch: 2/20... Training Step: 4952... Training loss: 1.6600... 0.0709 sec/batch
Epoch: 2/20... Training Step: 4953... Training loss: 1.9045... 0.0698 sec/batch
Epoch: 2/20... Training Step: 4954... Training loss: 1.7252... 0.0537 sec/batch
Epoch: 2/20... Training Step: 4955... Training loss: 1.6531... 0.0647 sec/batch
Epoch: 2/20... Training Step: 4956... Training loss: 1.6796... 0.0721 sec/batch
Epoch: 2/20... Training Step: 4957... Training loss: 1.6102... 0.0671 sec/batch
Epoch: 2/20... Training Step: 4958... Training loss: 1.6525... 0.0494 sec/batch
Epoch: 2/20... Training Step: 4959... Training loss: 1.7114... 0.0559 sec/batch
Epoch: 2/20... Training Step: 4960... Training loss: 1.6647... 0.0728 sec/batch
Epoch: 2/20... Training Step: 4961... Training loss: 1.5838... 0.0764 sec/batch
Epoch: 2/20... Training Step: 4962... Training loss: 1.8225... 0.0630 sec/batch
Epoch: 2/20... Training Step: 4963... Training loss: 1.8258... 0.0813 sec/batch
Epoch: 2/20... Training Step: 4964... Training loss: 1.7658... 0.0678 sec/batch
Epoch: 2/20... Training Step: 4965... Training loss: 1.8065... 0.0527 sec/batch
Epoch: 2/20... Training Step: 4966... Training loss: 1.8360... 0.0707 sec/batch
Epoch: 2/20... Training Step: 4967... Training loss: 1.7275... 0.0662 sec/batch
Epoch: 2/20... Training Step: 4968... Training loss: 1.8625... 0.0611 sec/batch
Epoch: 2/20... Training Step: 4969... Training loss: 1.7166... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4970... Training loss: 1.7152... 0.0642 sec/batch
Epoch: 2/20... Training Step: 4971... Training loss: 1.7478... 0.0600 sec/batch
Epoch: 2/20... Training Step: 4972... Training loss: 1.6212... 0.0545 sec/batch
Epoch: 2/20... Training Step: 4973... Training loss: 1.6879... 0.0638 sec/batch
Epoch: 2/20... Training Step: 4974... Training loss: 1.5813... 0.0781 sec/batch
Epoch: 2/20... Training Step: 4975... Training loss: 1.6064... 0.0772 sec/batch
Epoch: 2/20... Training Step: 4976... Training loss: 1.6560... 0.0647 sec/batch
Epoch: 2/20... Training Step: 4977... Training loss: 1.7526... 0.0718 sec/batch
Epoch: 2/20... Training Step: 4978... Training loss: 1.8493... 0.0771 sec/batch
Epoch: 2/20... Training Step: 4979... Training loss: 1.8005... 0.0547 sec/batch
Epoch: 2/20... Training Step: 4980... Training loss: 1.8319... 0.0715 sec/batch
Epoch: 2/20... Training Step: 4981... Training loss: 1.7162... 0.0619 sec/batch
Epoch: 2/20... Training Step: 4982... Training loss: 1.7727... 0.0729 sec/batch
Epoch: 2/20... Training Step: 4983... Training loss: 1.7467... 0.0553 sec/batch
Epoch: 2/20... Training Step: 4984... Training loss: 1.8666... 0.0708 sec/batch
Epoch: 2/20... Training Step: 4985... Training loss: 1.6332... 0.0640 sec/batch
Epoch: 2/20... Training Step: 4986... Training loss: 1.8446... 0.0763 sec/batch
Epoch: 2/20... Training Step: 4987... Training loss: 1.7910... 0.0670 sec/batch
Epoch: 2/20... Training Step: 4988... Training loss: 1.7811... 0.0686 sec/batch
Epoch: 2/20... Training Step: 4989... Training loss: 1.7186... 0.0604 sec/batch
Epoch: 2/20... Training Step: 4990... Training loss: 1.5824... 0.0744 sec/batch
Epoch: 2/20... Training Step: 4991... Training loss: 1.6816... 0.0737 sec/batch
Epoch: 2/20... Training Step: 4992... Training loss: 1.7666... 0.0717 sec/batch
Epoch: 2/20... Training Step: 4993... Training loss: 1.5948... 0.0534 sec/batch
Epoch: 2/20... Training Step: 4994... Training loss: 1.7226... 0.0601 sec/batch
Epoch: 2/20... Training Step: 4995... Training loss: 1.8022... 0.0814 sec/batch
Epoch: 2/20... Training Step: 4996... Training loss: 1.7801... 0.0778 sec/batch
Epoch: 2/20... Training Step: 4997... Training loss: 1.7188... 0.0615 sec/batch
Epoch: 2/20... Training Step: 4998... Training loss: 1.6893... 0.0802 sec/batch
Epoch: 2/20... Training Step: 4999... Training loss: 1.8484... 0.0713 sec/batch
Epoch: 2/20... Training Step: 5000... Training loss: 1.7524... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5001... Training loss: 1.6712... 0.0659 sec/batch
Epoch: 2/20... Training Step: 5002... Training loss: 1.6335... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5003... Training loss: 1.6365... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5004... Training loss: 1.7399... 0.0715 sec/batch
Epoch: 2/20... Training Step: 5005... Training loss: 1.7975... 0.0707 sec/batch
Epoch: 2/20... Training Step: 5006... Training loss: 1.8458... 0.0797 sec/batch
Epoch: 2/20... Training Step: 5007... Training loss: 1.6603... 0.0800 sec/batch
Epoch: 2/20... Training Step: 5008... Training loss: 1.7042... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5009... Training loss: 1.7459... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5010... Training loss: 1.7395... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5011... Training loss: 1.7350... 0.0529 sec/batch
Epoch: 2/20... Training Step: 5012... Training loss: 1.6974... 0.0740 sec/batch
Epoch: 2/20... Training Step: 5013... Training loss: 1.7996... 0.0702 sec/batch
Epoch: 2/20... Training Step: 5014... Training loss: 1.8323... 0.0688 sec/batch
Epoch: 2/20... Training Step: 5015... Training loss: 1.8364... 0.0561 sec/batch
Epoch: 2/20... Training Step: 5016... Training loss: 1.7156... 0.0783 sec/batch
Epoch: 2/20... Training Step: 5017... Training loss: 1.6616... 0.0733 sec/batch
Epoch: 2/20... Training Step: 5018... Training loss: 1.6734... 0.0766 sec/batch
Epoch: 2/20... Training Step: 5019... Training loss: 1.5978... 0.0596 sec/batch
Epoch: 2/20... Training Step: 5020... Training loss: 1.7402... 0.0758 sec/batch
Epoch: 2/20... Training Step: 5021... Training loss: 1.8267... 0.0856 sec/batch
Epoch: 2/20... Training Step: 5022... Training loss: 1.7730... 0.0762 sec/batch
Epoch: 2/20... Training Step: 5023... Training loss: 1.6688... 0.0823 sec/batch
Epoch: 2/20... Training Step: 5024... Training loss: 1.6845... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5025... Training loss: 1.7124... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5026... Training loss: 1.6550... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5027... Training loss: 1.5515... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5028... Training loss: 1.7490... 0.0717 sec/batch
Epoch: 2/20... Training Step: 5029... Training loss: 1.7722... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5030... Training loss: 1.7757... 0.0845 sec/batch
Epoch: 2/20... Training Step: 5031... Training loss: 1.7609... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5032... Training loss: 1.7418... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5033... Training loss: 1.7022... 0.0739 sec/batch
Epoch: 2/20... Training Step: 5034... Training loss: 1.7529... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5035... Training loss: 1.7044... 0.0843 sec/batch
Epoch: 2/20... Training Step: 5036... Training loss: 1.7583... 0.0595 sec/batch
Epoch: 2/20... Training Step: 5037... Training loss: 1.6349... 0.0677 sec/batch
Epoch: 2/20... Training Step: 5038... Training loss: 1.7902... 0.0692 sec/batch
Epoch: 2/20... Training Step: 5039... Training loss: 1.6820... 0.0787 sec/batch
Epoch: 2/20... Training Step: 5040... Training loss: 1.7417... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5041... Training loss: 1.7237... 0.0744 sec/batch
Epoch: 2/20... Training Step: 5042... Training loss: 1.6233... 0.0715 sec/batch
Epoch: 2/20... Training Step: 5043... Training loss: 1.5240... 0.0609 sec/batch
Epoch: 2/20... Training Step: 5044... Training loss: 1.6544... 0.0807 sec/batch
Epoch: 2/20... Training Step: 5045... Training loss: 1.7200... 0.0735 sec/batch
Epoch: 2/20... Training Step: 5046... Training loss: 1.6496... 0.0574 sec/batch
Epoch: 2/20... Training Step: 5047... Training loss: 1.7091... 0.0736 sec/batch
Epoch: 2/20... Training Step: 5048... Training loss: 1.7952... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5049... Training loss: 1.6921... 0.0596 sec/batch
Epoch: 2/20... Training Step: 5050... Training loss: 1.6334... 0.0774 sec/batch
Epoch: 2/20... Training Step: 5051... Training loss: 1.6772... 0.0869 sec/batch
Epoch: 2/20... Training Step: 5052... Training loss: 1.6509... 0.0645 sec/batch
Epoch: 2/20... Training Step: 5053... Training loss: 1.8578... 0.0602 sec/batch
Epoch: 2/20... Training Step: 5054... Training loss: 1.4707... 0.0765 sec/batch
Epoch: 2/20... Training Step: 5055... Training loss: 1.6461... 0.0776 sec/batch
Epoch: 2/20... Training Step: 5056... Training loss: 1.6930... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5057... Training loss: 1.6754... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5058... Training loss: 1.5518... 0.0620 sec/batch
Epoch: 2/20... Training Step: 5059... Training loss: 1.7393... 0.0636 sec/batch
Epoch: 2/20... Training Step: 5060... Training loss: 1.9303... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5061... Training loss: 1.8322... 0.0727 sec/batch
Epoch: 2/20... Training Step: 5062... Training loss: 1.9508... 0.0743 sec/batch
Epoch: 2/20... Training Step: 5063... Training loss: 1.7308... 0.0571 sec/batch
Epoch: 2/20... Training Step: 5064... Training loss: 1.7206... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5065... Training loss: 1.7612... 0.0662 sec/batch
Epoch: 2/20... Training Step: 5066... Training loss: 1.7906... 0.0731 sec/batch
Epoch: 2/20... Training Step: 5067... Training loss: 1.6713... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5068... Training loss: 1.5972... 0.0620 sec/batch
Epoch: 2/20... Training Step: 5069... Training loss: 1.7823... 0.0690 sec/batch
Epoch: 2/20... Training Step: 5070... Training loss: 1.7546... 0.0601 sec/batch
Epoch: 2/20... Training Step: 5071... Training loss: 1.6871... 0.0671 sec/batch
Epoch: 2/20... Training Step: 5072... Training loss: 1.6496... 0.0711 sec/batch
Epoch: 2/20... Training Step: 5073... Training loss: 1.5784... 0.0720 sec/batch
Epoch: 2/20... Training Step: 5074... Training loss: 1.7896... 0.0761 sec/batch
Epoch: 2/20... Training Step: 5075... Training loss: 1.7369... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5076... Training loss: 1.6922... 0.0557 sec/batch
Epoch: 2/20... Training Step: 5077... Training loss: 1.7279... 0.0846 sec/batch
Epoch: 2/20... Training Step: 5078... Training loss: 1.7954... 0.0793 sec/batch
Epoch: 2/20... Training Step: 5079... Training loss: 1.6074... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5080... Training loss: 1.6901... 0.0630 sec/batch
Epoch: 2/20... Training Step: 5081... Training loss: 1.6936... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5082... Training loss: 1.7683... 0.0619 sec/batch
Epoch: 2/20... Training Step: 5083... Training loss: 1.6021... 0.0681 sec/batch
Epoch: 2/20... Training Step: 5084... Training loss: 1.7718... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5085... Training loss: 1.7722... 0.0743 sec/batch
Epoch: 2/20... Training Step: 5086... Training loss: 1.8664... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5087... Training loss: 1.7393... 0.0553 sec/batch
Epoch: 2/20... Training Step: 5088... Training loss: 1.7207... 0.0754 sec/batch
Epoch: 2/20... Training Step: 5089... Training loss: 1.6021... 0.0809 sec/batch
Epoch: 2/20... Training Step: 5090... Training loss: 1.5793... 0.0743 sec/batch
Epoch: 2/20... Training Step: 5091... Training loss: 1.7699... 0.0549 sec/batch
Epoch: 2/20... Training Step: 5092... Training loss: 1.7501... 0.0750 sec/batch
Epoch: 2/20... Training Step: 5093... Training loss: 1.6747... 0.0608 sec/batch
Epoch: 2/20... Training Step: 5094... Training loss: 1.6379... 0.0658 sec/batch
Epoch: 2/20... Training Step: 5095... Training loss: 1.7883... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5096... Training loss: 1.7120... 0.0695 sec/batch
Epoch: 2/20... Training Step: 5097... Training loss: 1.8429... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5098... Training loss: 1.6236... 0.0640 sec/batch
Epoch: 2/20... Training Step: 5099... Training loss: 1.5992... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5100... Training loss: 1.6742... 0.0767 sec/batch
Epoch: 2/20... Training Step: 5101... Training loss: 1.7105... 0.0672 sec/batch
Epoch: 2/20... Training Step: 5102... Training loss: 1.5551... 0.0744 sec/batch
Epoch: 2/20... Training Step: 5103... Training loss: 1.6775... 0.0808 sec/batch
Epoch: 2/20... Training Step: 5104... Training loss: 1.7864... 0.0624 sec/batch
Epoch: 2/20... Training Step: 5105... Training loss: 1.5970... 0.0826 sec/batch
Epoch: 2/20... Training Step: 5106... Training loss: 1.8796... 0.0601 sec/batch
Epoch: 2/20... Training Step: 5107... Training loss: 1.7867... 0.0599 sec/batch
Epoch: 2/20... Training Step: 5108... Training loss: 1.8830... 0.0706 sec/batch
Epoch: 2/20... Training Step: 5109... Training loss: 1.8189... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5110... Training loss: 1.9410... 0.0752 sec/batch
Epoch: 2/20... Training Step: 5111... Training loss: 1.6918... 0.0664 sec/batch
Epoch: 2/20... Training Step: 5112... Training loss: 1.7218... 0.0844 sec/batch
Epoch: 2/20... Training Step: 5113... Training loss: 1.6630... 0.0663 sec/batch
Epoch: 2/20... Training Step: 5114... Training loss: 1.9686... 0.0641 sec/batch
Epoch: 2/20... Training Step: 5115... Training loss: 1.9398... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5116... Training loss: 1.6882... 0.0778 sec/batch
Epoch: 2/20... Training Step: 5117... Training loss: 1.8036... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5118... Training loss: 1.7714... 0.0783 sec/batch
Epoch: 2/20... Training Step: 5119... Training loss: 1.8657... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5120... Training loss: 1.6996... 0.0614 sec/batch
Epoch: 2/20... Training Step: 5121... Training loss: 1.7508... 0.0615 sec/batch
Epoch: 2/20... Training Step: 5122... Training loss: 1.7822... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5123... Training loss: 1.7960... 0.0643 sec/batch
Epoch: 2/20... Training Step: 5124... Training loss: 1.7811... 0.0706 sec/batch
Epoch: 2/20... Training Step: 5125... Training loss: 1.7436... 0.0613 sec/batch
Epoch: 2/20... Training Step: 5126... Training loss: 1.5861... 0.0719 sec/batch
Epoch: 2/20... Training Step: 5127... Training loss: 1.7082... 0.0761 sec/batch
Epoch: 2/20... Training Step: 5128... Training loss: 1.6179... 0.0779 sec/batch
Epoch: 2/20... Training Step: 5129... Training loss: 1.8067... 0.0615 sec/batch
Epoch: 2/20... Training Step: 5130... Training loss: 1.6740... 0.0710 sec/batch
Epoch: 2/20... Training Step: 5131... Training loss: 1.7268... 0.0778 sec/batch
Epoch: 2/20... Training Step: 5132... Training loss: 1.8339... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5133... Training loss: 1.7589... 0.0889 sec/batch
Epoch: 2/20... Training Step: 5134... Training loss: 1.7332... 0.0774 sec/batch
Epoch: 2/20... Training Step: 5135... Training loss: 1.7707... 0.0776 sec/batch
Epoch: 2/20... Training Step: 5136... Training loss: 1.8254... 0.0766 sec/batch
Epoch: 2/20... Training Step: 5137... Training loss: 1.7250... 0.0760 sec/batch
Epoch: 2/20... Training Step: 5138... Training loss: 1.8301... 0.0586 sec/batch
Epoch: 2/20... Training Step: 5139... Training loss: 1.7567... 0.0722 sec/batch
Epoch: 2/20... Training Step: 5140... Training loss: 1.8154... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5141... Training loss: 1.6300... 0.0691 sec/batch
Epoch: 2/20... Training Step: 5142... Training loss: 1.8711... 0.0700 sec/batch
Epoch: 2/20... Training Step: 5143... Training loss: 1.6773... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5144... Training loss: 1.6558... 0.0625 sec/batch
Epoch: 2/20... Training Step: 5145... Training loss: 1.7962... 0.0688 sec/batch
Epoch: 2/20... Training Step: 5146... Training loss: 1.6754... 0.0805 sec/batch
Epoch: 2/20... Training Step: 5147... Training loss: 1.6736... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5148... Training loss: 1.6071... 0.0663 sec/batch
Epoch: 2/20... Training Step: 5149... Training loss: 1.5577... 0.0836 sec/batch
Epoch: 2/20... Training Step: 5150... Training loss: 1.7592... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5151... Training loss: 1.8000... 0.0644 sec/batch
Epoch: 2/20... Training Step: 5152... Training loss: 1.8128... 0.0616 sec/batch
Epoch: 2/20... Training Step: 5153... Training loss: 1.7700... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5154... Training loss: 1.8570... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5155... Training loss: 1.7243... 0.0748 sec/batch
Epoch: 2/20... Training Step: 5156... Training loss: 1.7377... 0.0983 sec/batch
Epoch: 2/20... Training Step: 5157... Training loss: 1.8470... 0.1217 sec/batch
Epoch: 2/20... Training Step: 5158... Training loss: 1.8177... 0.0965 sec/batch
Epoch: 2/20... Training Step: 5159... Training loss: 1.6307... 0.1059 sec/batch
Epoch: 2/20... Training Step: 5160... Training loss: 1.6872... 0.1135 sec/batch
Epoch: 2/20... Training Step: 5161... Training loss: 1.7591... 0.1130 sec/batch
Epoch: 2/20... Training Step: 5162... Training loss: 1.7599... 0.0862 sec/batch
Epoch: 2/20... Training Step: 5163... Training loss: 1.7448... 0.1053 sec/batch
Epoch: 2/20... Training Step: 5164... Training loss: 1.7938... 0.1258 sec/batch
Epoch: 2/20... Training Step: 5165... Training loss: 1.8530... 0.1146 sec/batch
Epoch: 2/20... Training Step: 5166... Training loss: 1.7485... 0.0886 sec/batch
Epoch: 2/20... Training Step: 5167... Training loss: 1.6099... 0.1161 sec/batch
Epoch: 2/20... Training Step: 5168... Training loss: 1.7184... 0.1096 sec/batch
Epoch: 2/20... Training Step: 5169... Training loss: 1.7266... 0.1081 sec/batch
Epoch: 2/20... Training Step: 5170... Training loss: 1.7032... 0.1108 sec/batch
Epoch: 2/20... Training Step: 5171... Training loss: 1.6972... 0.0919 sec/batch
Epoch: 2/20... Training Step: 5172... Training loss: 1.8035... 0.1054 sec/batch
Epoch: 2/20... Training Step: 5173... Training loss: 1.7189... 0.0914 sec/batch
Epoch: 2/20... Training Step: 5174... Training loss: 1.6668... 0.0918 sec/batch
Epoch: 2/20... Training Step: 5175... Training loss: 1.7197... 0.0957 sec/batch
Epoch: 2/20... Training Step: 5176... Training loss: 1.7049... 0.1203 sec/batch
Epoch: 2/20... Training Step: 5177... Training loss: 1.6775... 0.1090 sec/batch
Epoch: 2/20... Training Step: 5178... Training loss: 1.8105... 0.0955 sec/batch
Epoch: 2/20... Training Step: 5179... Training loss: 1.8798... 0.1239 sec/batch
Epoch: 2/20... Training Step: 5180... Training loss: 1.8002... 0.0925 sec/batch
Epoch: 2/20... Training Step: 5181... Training loss: 1.7476... 0.1216 sec/batch
Epoch: 2/20... Training Step: 5182... Training loss: 1.8150... 0.0951 sec/batch
Epoch: 2/20... Training Step: 5183... Training loss: 2.0021... 0.0975 sec/batch
Epoch: 2/20... Training Step: 5184... Training loss: 1.7236... 0.1118 sec/batch
Epoch: 2/20... Training Step: 5185... Training loss: 1.6330... 0.1016 sec/batch
Epoch: 2/20... Training Step: 5186... Training loss: 1.9013... 0.0994 sec/batch
Epoch: 2/20... Training Step: 5187... Training loss: 1.8932... 0.1103 sec/batch
Epoch: 2/20... Training Step: 5188... Training loss: 1.8554... 0.1101 sec/batch
Epoch: 2/20... Training Step: 5189... Training loss: 1.7162... 0.0977 sec/batch
Epoch: 2/20... Training Step: 5190... Training loss: 1.6880... 0.1065 sec/batch
Epoch: 2/20... Training Step: 5191... Training loss: 1.7964... 0.1102 sec/batch
Epoch: 2/20... Training Step: 5192... Training loss: 1.7879... 0.0903 sec/batch
Epoch: 2/20... Training Step: 5193... Training loss: 1.5742... 0.1026 sec/batch
Epoch: 2/20... Training Step: 5194... Training loss: 1.6556... 0.0876 sec/batch
Epoch: 2/20... Training Step: 5195... Training loss: 1.7975... 0.0737 sec/batch
Epoch: 2/20... Training Step: 5196... Training loss: 1.6898... 0.0708 sec/batch
Epoch: 2/20... Training Step: 5197... Training loss: 1.8712... 0.0671 sec/batch
Epoch: 2/20... Training Step: 5198... Training loss: 1.7756... 0.0893 sec/batch
Epoch: 2/20... Training Step: 5199... Training loss: 1.7082... 0.0998 sec/batch
Epoch: 2/20... Training Step: 5200... Training loss: 1.8778... 0.0804 sec/batch
Epoch: 2/20... Training Step: 5201... Training loss: 1.6051... 0.0900 sec/batch
Epoch: 2/20... Training Step: 5202... Training loss: 1.8310... 0.0667 sec/batch
Epoch: 2/20... Training Step: 5203... Training loss: 1.7178... 0.0892 sec/batch
Epoch: 2/20... Training Step: 5204... Training loss: 1.8973... 0.0863 sec/batch
Epoch: 2/20... Training Step: 5205... Training loss: 1.6542... 0.0755 sec/batch
Epoch: 2/20... Training Step: 5206... Training loss: 1.6931... 0.0949 sec/batch
Epoch: 2/20... Training Step: 5207... Training loss: 1.7687... 0.0868 sec/batch
Epoch: 2/20... Training Step: 5208... Training loss: 1.5423... 0.0834 sec/batch
Epoch: 2/20... Training Step: 5209... Training loss: 1.6063... 0.0683 sec/batch
Epoch: 2/20... Training Step: 5210... Training loss: 1.6615... 0.0860 sec/batch
Epoch: 2/20... Training Step: 5211... Training loss: 1.6407... 0.0756 sec/batch
Epoch: 2/20... Training Step: 5212... Training loss: 1.6804... 0.0977 sec/batch
Epoch: 2/20... Training Step: 5213... Training loss: 1.6273... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5214... Training loss: 1.7726... 0.0748 sec/batch
Epoch: 2/20... Training Step: 5215... Training loss: 1.7022... 0.0701 sec/batch
Epoch: 2/20... Training Step: 5216... Training loss: 1.7624... 0.0837 sec/batch
Epoch: 2/20... Training Step: 5217... Training loss: 1.7699... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5218... Training loss: 1.7011... 0.0814 sec/batch
Epoch: 2/20... Training Step: 5219... Training loss: 1.7814... 0.0817 sec/batch
Epoch: 2/20... Training Step: 5220... Training loss: 1.5655... 0.0838 sec/batch
Epoch: 2/20... Training Step: 5221... Training loss: 1.7634... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5222... Training loss: 1.6741... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5223... Training loss: 1.6202... 0.0691 sec/batch
Epoch: 2/20... Training Step: 5224... Training loss: 1.7709... 0.0803 sec/batch
Epoch: 2/20... Training Step: 5225... Training loss: 1.6963... 0.0839 sec/batch
Epoch: 2/20... Training Step: 5226... Training loss: 1.7236... 0.0907 sec/batch
Epoch: 2/20... Training Step: 5227... Training loss: 1.7991... 0.0848 sec/batch
Epoch: 2/20... Training Step: 5228... Training loss: 1.6394... 0.0817 sec/batch
Epoch: 2/20... Training Step: 5229... Training loss: 1.6999... 0.0697 sec/batch
Epoch: 2/20... Training Step: 5230... Training loss: 1.7254... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5231... Training loss: 1.8558... 0.0621 sec/batch
Epoch: 2/20... Training Step: 5232... Training loss: 1.8465... 0.0794 sec/batch
Epoch: 2/20... Training Step: 5233... Training loss: 1.7255... 0.0898 sec/batch
Epoch: 2/20... Training Step: 5234... Training loss: 1.5458... 0.0860 sec/batch
Epoch: 2/20... Training Step: 5235... Training loss: 1.6613... 0.0835 sec/batch
Epoch: 2/20... Training Step: 5236... Training loss: 1.6342... 0.0851 sec/batch
Epoch: 2/20... Training Step: 5237... Training loss: 1.8704... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5238... Training loss: 1.7045... 0.0676 sec/batch
Epoch: 2/20... Training Step: 5239... Training loss: 1.6913... 0.0719 sec/batch
Epoch: 2/20... Training Step: 5240... Training loss: 1.7594... 0.0757 sec/batch
Epoch: 2/20... Training Step: 5241... Training loss: 1.6345... 0.0820 sec/batch
Epoch: 2/20... Training Step: 5242... Training loss: 1.7203... 0.0761 sec/batch
Epoch: 2/20... Training Step: 5243... Training loss: 1.7621... 0.0783 sec/batch
Epoch: 2/20... Training Step: 5244... Training loss: 1.6807... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5245... Training loss: 1.7951... 0.0817 sec/batch
Epoch: 2/20... Training Step: 5246... Training loss: 1.7109... 0.0807 sec/batch
Epoch: 2/20... Training Step: 5247... Training loss: 1.8532... 0.0814 sec/batch
Epoch: 2/20... Training Step: 5248... Training loss: 1.8933... 0.0678 sec/batch
Epoch: 2/20... Training Step: 5249... Training loss: 1.7830... 0.0606 sec/batch
Epoch: 2/20... Training Step: 5250... Training loss: 1.5902... 0.0601 sec/batch
Epoch: 2/20... Training Step: 5251... Training loss: 1.7121... 0.0687 sec/batch
Epoch: 2/20... Training Step: 5252... Training loss: 1.8959... 0.0856 sec/batch
Epoch: 2/20... Training Step: 5253... Training loss: 1.7039... 0.0560 sec/batch
Epoch: 2/20... Training Step: 5254... Training loss: 1.8984... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5255... Training loss: 1.8326... 0.0767 sec/batch
Epoch: 2/20... Training Step: 5256... Training loss: 1.7259... 0.0688 sec/batch
Epoch: 2/20... Training Step: 5257... Training loss: 1.8116... 0.0802 sec/batch
Epoch: 2/20... Training Step: 5258... Training loss: 1.6823... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5259... Training loss: 1.7089... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5260... Training loss: 1.9198... 0.0739 sec/batch
Epoch: 2/20... Training Step: 5261... Training loss: 1.8088... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5262... Training loss: 1.9043... 0.0662 sec/batch
Epoch: 2/20... Training Step: 5263... Training loss: 1.7769... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5264... Training loss: 1.7470... 0.0777 sec/batch
Epoch: 2/20... Training Step: 5265... Training loss: 1.7949... 0.0803 sec/batch
Epoch: 2/20... Training Step: 5266... Training loss: 1.8275... 0.0572 sec/batch
Epoch: 2/20... Training Step: 5267... Training loss: 1.7024... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5268... Training loss: 1.5599... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5269... Training loss: 1.6611... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5270... Training loss: 1.7335... 0.0744 sec/batch
Epoch: 2/20... Training Step: 5271... Training loss: 1.7640... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5272... Training loss: 1.7524... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5273... Training loss: 1.7624... 0.0626 sec/batch
Epoch: 2/20... Training Step: 5274... Training loss: 1.7892... 0.0556 sec/batch
Epoch: 2/20... Training Step: 5275... Training loss: 1.7895... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5276... Training loss: 1.8523... 0.0642 sec/batch
Epoch: 2/20... Training Step: 5277... Training loss: 1.8925... 0.0675 sec/batch
Epoch: 2/20... Training Step: 5278... Training loss: 1.8799... 0.0658 sec/batch
Epoch: 2/20... Training Step: 5279... Training loss: 1.7798... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5280... Training loss: 1.8404... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5281... Training loss: 1.7595... 0.0685 sec/batch
Epoch: 2/20... Training Step: 5282... Training loss: 1.8352... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5283... Training loss: 1.7140... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5284... Training loss: 1.8845... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5285... Training loss: 1.8021... 0.0712 sec/batch
Epoch: 2/20... Training Step: 5286... Training loss: 1.8093... 0.0718 sec/batch
Epoch: 2/20... Training Step: 5287... Training loss: 1.7439... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5288... Training loss: 1.6666... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5289... Training loss: 1.7082... 0.0613 sec/batch
Epoch: 2/20... Training Step: 5290... Training loss: 1.8000... 0.0760 sec/batch
Epoch: 2/20... Training Step: 5291... Training loss: 1.7663... 0.0800 sec/batch
Epoch: 2/20... Training Step: 5292... Training loss: 1.8305... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5293... Training loss: 1.8827... 0.0509 sec/batch
Epoch: 2/20... Training Step: 5294... Training loss: 1.8064... 0.0777 sec/batch
Epoch: 2/20... Training Step: 5295... Training loss: 1.7753... 0.0693 sec/batch
Epoch: 2/20... Training Step: 5296... Training loss: 1.7879... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5297... Training loss: 1.7693... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5298... Training loss: 1.7723... 0.0778 sec/batch
Epoch: 2/20... Training Step: 5299... Training loss: 1.8096... 0.0664 sec/batch
Epoch: 2/20... Training Step: 5300... Training loss: 1.7980... 0.0720 sec/batch
Epoch: 2/20... Training Step: 5301... Training loss: 1.8290... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5302... Training loss: 1.7542... 0.0764 sec/batch
Epoch: 2/20... Training Step: 5303... Training loss: 1.7960... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5304... Training loss: 1.7644... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5305... Training loss: 1.6712... 0.0776 sec/batch
Epoch: 2/20... Training Step: 5306... Training loss: 1.6744... 0.0596 sec/batch
Epoch: 2/20... Training Step: 5307... Training loss: 1.6105... 0.0779 sec/batch
Epoch: 2/20... Training Step: 5308... Training loss: 1.7768... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5309... Training loss: 1.8197... 0.0611 sec/batch
Epoch: 2/20... Training Step: 5310... Training loss: 1.7459... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5311... Training loss: 1.5918... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5312... Training loss: 1.6361... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5313... Training loss: 1.7745... 0.0604 sec/batch
Epoch: 2/20... Training Step: 5314... Training loss: 1.8200... 0.0624 sec/batch
Epoch: 2/20... Training Step: 5315... Training loss: 1.8268... 0.0572 sec/batch
Epoch: 2/20... Training Step: 5316... Training loss: 1.7339... 0.0555 sec/batch
Epoch: 2/20... Training Step: 5317... Training loss: 1.7753... 0.0625 sec/batch
Epoch: 2/20... Training Step: 5318... Training loss: 1.7802... 0.0476 sec/batch
Epoch: 2/20... Training Step: 5319... Training loss: 1.7616... 0.0601 sec/batch
Epoch: 2/20... Training Step: 5320... Training loss: 1.6184... 0.0709 sec/batch
Epoch: 2/20... Training Step: 5321... Training loss: 1.5689... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5322... Training loss: 1.6271... 0.0584 sec/batch
Epoch: 2/20... Training Step: 5323... Training loss: 1.6995... 0.0595 sec/batch
Epoch: 2/20... Training Step: 5324... Training loss: 1.7500... 0.0804 sec/batch
Epoch: 2/20... Training Step: 5325... Training loss: 1.6943... 0.0519 sec/batch
Epoch: 2/20... Training Step: 5326... Training loss: 1.7203... 0.0605 sec/batch
Epoch: 2/20... Training Step: 5327... Training loss: 1.7635... 0.0543 sec/batch
Epoch: 2/20... Training Step: 5328... Training loss: 1.6756... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5329... Training loss: 1.8886... 0.0641 sec/batch
Epoch: 2/20... Training Step: 5330... Training loss: 1.7028... 0.0565 sec/batch
Epoch: 2/20... Training Step: 5331... Training loss: 1.6650... 0.0592 sec/batch
Epoch: 2/20... Training Step: 5332... Training loss: 1.5999... 0.0654 sec/batch
Epoch: 2/20... Training Step: 5333... Training loss: 1.5639... 0.0597 sec/batch
Epoch: 2/20... Training Step: 5334... Training loss: 1.6926... 0.0671 sec/batch
Epoch: 2/20... Training Step: 5335... Training loss: 1.6575... 0.0695 sec/batch
Epoch: 2/20... Training Step: 5336... Training loss: 1.7579... 0.0642 sec/batch
Epoch: 2/20... Training Step: 5337... Training loss: 1.6834... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5338... Training loss: 1.7159... 0.0668 sec/batch
Epoch: 2/20... Training Step: 5339... Training loss: 1.8968... 0.0631 sec/batch
Epoch: 2/20... Training Step: 5340... Training loss: 1.7729... 0.0678 sec/batch
Epoch: 2/20... Training Step: 5341... Training loss: 1.7757... 0.0785 sec/batch
Epoch: 2/20... Training Step: 5342... Training loss: 1.6476... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5343... Training loss: 1.7370... 0.0642 sec/batch
Epoch: 2/20... Training Step: 5344... Training loss: 1.7874... 0.0551 sec/batch
Epoch: 2/20... Training Step: 5345... Training loss: 1.5884... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5346... Training loss: 1.7388... 0.0729 sec/batch
Epoch: 2/20... Training Step: 5347... Training loss: 1.7550... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5348... Training loss: 1.7082... 0.0607 sec/batch
Epoch: 2/20... Training Step: 5349... Training loss: 1.6209... 0.0676 sec/batch
Epoch: 2/20... Training Step: 5350... Training loss: 1.7079... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5351... Training loss: 1.8270... 0.0602 sec/batch
Epoch: 2/20... Training Step: 5352... Training loss: 1.8265... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5353... Training loss: 1.8913... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5354... Training loss: 1.7632... 0.0616 sec/batch
Epoch: 2/20... Training Step: 5355... Training loss: 1.8114... 0.0506 sec/batch
Epoch: 2/20... Training Step: 5356... Training loss: 1.9328... 0.0598 sec/batch
Epoch: 2/20... Training Step: 5357... Training loss: 1.7558... 0.0624 sec/batch
Epoch: 2/20... Training Step: 5358... Training loss: 1.7512... 0.0646 sec/batch
Epoch: 2/20... Training Step: 5359... Training loss: 1.7948... 0.0533 sec/batch
Epoch: 2/20... Training Step: 5360... Training loss: 1.7545... 0.0583 sec/batch
Epoch: 2/20... Training Step: 5361... Training loss: 1.8099... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5362... Training loss: 1.7199... 0.0705 sec/batch
Epoch: 2/20... Training Step: 5363... Training loss: 1.8139... 0.0683 sec/batch
Epoch: 2/20... Training Step: 5364... Training loss: 1.7289... 0.0675 sec/batch
Epoch: 2/20... Training Step: 5365... Training loss: 1.7456... 0.0843 sec/batch
Epoch: 2/20... Training Step: 5366... Training loss: 1.6901... 0.0805 sec/batch
Epoch: 2/20... Training Step: 5367... Training loss: 1.7183... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5368... Training loss: 1.6434... 0.0780 sec/batch
Epoch: 2/20... Training Step: 5369... Training loss: 1.8443... 0.0674 sec/batch
Epoch: 2/20... Training Step: 5370... Training loss: 1.8701... 0.0756 sec/batch
Epoch: 2/20... Training Step: 5371... Training loss: 1.6387... 0.0849 sec/batch
Epoch: 2/20... Training Step: 5372... Training loss: 1.8503... 0.0857 sec/batch
Epoch: 2/20... Training Step: 5373... Training loss: 1.7822... 0.0662 sec/batch
Epoch: 2/20... Training Step: 5374... Training loss: 1.8403... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5375... Training loss: 1.7788... 0.0578 sec/batch
Epoch: 2/20... Training Step: 5376... Training loss: 1.6083... 0.0676 sec/batch
Epoch: 2/20... Training Step: 5377... Training loss: 1.7961... 0.0774 sec/batch
Epoch: 2/20... Training Step: 5378... Training loss: 1.8227... 0.0765 sec/batch
Epoch: 2/20... Training Step: 5379... Training loss: 1.7487... 0.0534 sec/batch
Epoch: 2/20... Training Step: 5380... Training loss: 1.7596... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5381... Training loss: 1.6190... 0.0738 sec/batch
Epoch: 2/20... Training Step: 5382... Training loss: 1.7717... 0.0628 sec/batch
Epoch: 2/20... Training Step: 5383... Training loss: 1.6840... 0.0722 sec/batch
Epoch: 2/20... Training Step: 5384... Training loss: 2.0204... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5385... Training loss: 1.7450... 0.0803 sec/batch
Epoch: 2/20... Training Step: 5386... Training loss: 1.7407... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5387... Training loss: 1.8911... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5388... Training loss: 1.8308... 0.0631 sec/batch
Epoch: 2/20... Training Step: 5389... Training loss: 1.6591... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5390... Training loss: 1.7608... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5391... Training loss: 1.8786... 0.0708 sec/batch
Epoch: 2/20... Training Step: 5392... Training loss: 1.7257... 0.0760 sec/batch
Epoch: 2/20... Training Step: 5393... Training loss: 1.7929... 0.0706 sec/batch
Epoch: 2/20... Training Step: 5394... Training loss: 1.6729... 0.0801 sec/batch
Epoch: 2/20... Training Step: 5395... Training loss: 1.7095... 0.0723 sec/batch
Epoch: 2/20... Training Step: 5396... Training loss: 1.7138... 0.0878 sec/batch
Epoch: 2/20... Training Step: 5397... Training loss: 1.6128... 0.0810 sec/batch
Epoch: 2/20... Training Step: 5398... Training loss: 1.8206... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5399... Training loss: 1.7173... 0.0914 sec/batch
Epoch: 2/20... Training Step: 5400... Training loss: 1.7983... 0.0835 sec/batch
Epoch: 2/20... Training Step: 5401... Training loss: 1.7722... 0.0577 sec/batch
Epoch: 2/20... Training Step: 5402... Training loss: 1.7845... 0.0757 sec/batch
Epoch: 2/20... Training Step: 5403... Training loss: 1.9684... 0.0739 sec/batch
Epoch: 2/20... Training Step: 5404... Training loss: 1.7233... 0.0731 sec/batch
Epoch: 2/20... Training Step: 5405... Training loss: 1.8262... 0.0781 sec/batch
Epoch: 2/20... Training Step: 5406... Training loss: 1.7719... 0.0659 sec/batch
Epoch: 2/20... Training Step: 5407... Training loss: 1.7402... 0.0664 sec/batch
Epoch: 2/20... Training Step: 5408... Training loss: 1.7482... 0.0715 sec/batch
Epoch: 2/20... Training Step: 5409... Training loss: 1.8290... 0.0812 sec/batch
Epoch: 2/20... Training Step: 5410... Training loss: 1.6922... 0.0707 sec/batch
Epoch: 2/20... Training Step: 5411... Training loss: 1.6314... 0.0899 sec/batch
Epoch: 2/20... Training Step: 5412... Training loss: 1.8189... 0.0874 sec/batch
Epoch: 2/20... Training Step: 5413... Training loss: 1.8050... 0.0813 sec/batch
Epoch: 2/20... Training Step: 5414... Training loss: 1.7538... 0.0700 sec/batch
Epoch: 2/20... Training Step: 5415... Training loss: 1.7832... 0.0635 sec/batch
Epoch: 2/20... Training Step: 5416... Training loss: 1.9175... 0.0690 sec/batch
Epoch: 2/20... Training Step: 5417... Training loss: 1.9311... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5418... Training loss: 1.7275... 0.0749 sec/batch
Epoch: 2/20... Training Step: 5419... Training loss: 1.8185... 0.0766 sec/batch
Epoch: 2/20... Training Step: 5420... Training loss: 1.8064... 0.0739 sec/batch
Epoch: 2/20... Training Step: 5421... Training loss: 1.7574... 0.0786 sec/batch
Epoch: 2/20... Training Step: 5422... Training loss: 1.6484... 0.0745 sec/batch
Epoch: 2/20... Training Step: 5423... Training loss: 1.8811... 0.0784 sec/batch
Epoch: 2/20... Training Step: 5424... Training loss: 1.8675... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5425... Training loss: 1.8697... 0.0722 sec/batch
Epoch: 2/20... Training Step: 5426... Training loss: 1.6119... 0.0847 sec/batch
Epoch: 2/20... Training Step: 5427... Training loss: 1.6830... 0.0813 sec/batch
Epoch: 2/20... Training Step: 5428... Training loss: 1.7625... 0.0598 sec/batch
Epoch: 2/20... Training Step: 5429... Training loss: 1.6403... 0.0624 sec/batch
Epoch: 2/20... Training Step: 5430... Training loss: 1.7623... 0.0594 sec/batch
Epoch: 2/20... Training Step: 5431... Training loss: 1.7053... 0.0595 sec/batch
Epoch: 2/20... Training Step: 5432... Training loss: 1.7011... 0.0576 sec/batch
Epoch: 2/20... Training Step: 5433... Training loss: 1.7133... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5434... Training loss: 1.7345... 0.0732 sec/batch
Epoch: 2/20... Training Step: 5435... Training loss: 1.7510... 0.0641 sec/batch
Epoch: 2/20... Training Step: 5436... Training loss: 1.7660... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5437... Training loss: 1.6665... 0.0642 sec/batch
Epoch: 2/20... Training Step: 5438... Training loss: 1.6640... 0.0672 sec/batch
Epoch: 2/20... Training Step: 5439... Training loss: 1.6794... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5440... Training loss: 1.6807... 0.0597 sec/batch
Epoch: 2/20... Training Step: 5441... Training loss: 1.7716... 0.0646 sec/batch
Epoch: 2/20... Training Step: 5442... Training loss: 1.8567... 0.0603 sec/batch
Epoch: 2/20... Training Step: 5443... Training loss: 1.7470... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5444... Training loss: 1.8270... 0.0859 sec/batch
Epoch: 2/20... Training Step: 5445... Training loss: 1.7146... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5446... Training loss: 1.6344... 0.0656 sec/batch
Epoch: 2/20... Training Step: 5447... Training loss: 1.8792... 0.0712 sec/batch
Epoch: 2/20... Training Step: 5448... Training loss: 1.8803... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5449... Training loss: 1.8333... 0.0605 sec/batch
Epoch: 2/20... Training Step: 5450... Training loss: 1.7900... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5451... Training loss: 1.8060... 0.0839 sec/batch
Epoch: 2/20... Training Step: 5452... Training loss: 1.6926... 0.0581 sec/batch
Epoch: 2/20... Training Step: 5453... Training loss: 1.7325... 0.0565 sec/batch
Epoch: 2/20... Training Step: 5454... Training loss: 1.8030... 0.0714 sec/batch
Epoch: 2/20... Training Step: 5455... Training loss: 1.6966... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5456... Training loss: 1.8603... 0.0628 sec/batch
Epoch: 2/20... Training Step: 5457... Training loss: 1.7460... 0.0551 sec/batch
Epoch: 2/20... Training Step: 5458... Training loss: 1.7534... 0.0727 sec/batch
Epoch: 2/20... Training Step: 5459... Training loss: 1.8036... 0.0629 sec/batch
Epoch: 2/20... Training Step: 5460... Training loss: 1.7887... 0.0629 sec/batch
Epoch: 2/20... Training Step: 5461... Training loss: 1.8257... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5462... Training loss: 1.6861... 0.0590 sec/batch
Epoch: 2/20... Training Step: 5463... Training loss: 1.9077... 0.0564 sec/batch
Epoch: 2/20... Training Step: 5464... Training loss: 1.6124... 0.0610 sec/batch
Epoch: 2/20... Training Step: 5465... Training loss: 1.7215... 0.0683 sec/batch
Epoch: 2/20... Training Step: 5466... Training loss: 1.6780... 0.0677 sec/batch
Epoch: 2/20... Training Step: 5467... Training loss: 1.7829... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5468... Training loss: 1.7386... 0.0787 sec/batch
Epoch: 2/20... Training Step: 5469... Training loss: 1.7966... 0.0736 sec/batch
Epoch: 2/20... Training Step: 5470... Training loss: 1.8334... 0.0710 sec/batch
Epoch: 2/20... Training Step: 5471... Training loss: 1.7423... 0.0709 sec/batch
Epoch: 2/20... Training Step: 5472... Training loss: 1.7750... 0.0705 sec/batch
Epoch: 2/20... Training Step: 5473... Training loss: 1.8339... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5474... Training loss: 1.9494... 0.0829 sec/batch
Epoch: 2/20... Training Step: 5475... Training loss: 1.7009... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5476... Training loss: 1.6595... 0.0925 sec/batch
Epoch: 2/20... Training Step: 5477... Training loss: 1.7955... 0.0764 sec/batch
Epoch: 2/20... Training Step: 5478... Training loss: 1.7177... 0.0552 sec/batch
Epoch: 2/20... Training Step: 5479... Training loss: 1.6634... 0.0660 sec/batch
Epoch: 2/20... Training Step: 5480... Training loss: 1.6649... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5481... Training loss: 1.7108... 0.0840 sec/batch
Epoch: 2/20... Training Step: 5482... Training loss: 1.5431... 0.0758 sec/batch
Epoch: 2/20... Training Step: 5483... Training loss: 1.6548... 0.0652 sec/batch
Epoch: 2/20... Training Step: 5484... Training loss: 1.6233... 0.0559 sec/batch
Epoch: 2/20... Training Step: 5485... Training loss: 1.6953... 0.0504 sec/batch
Epoch: 2/20... Training Step: 5486... Training loss: 1.8521... 0.0801 sec/batch
Epoch: 2/20... Training Step: 5487... Training loss: 1.7891... 0.0745 sec/batch
Epoch: 2/20... Training Step: 5488... Training loss: 1.7579... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5489... Training loss: 1.7796... 0.0818 sec/batch
Epoch: 2/20... Training Step: 5490... Training loss: 1.8042... 0.0554 sec/batch
Epoch: 2/20... Training Step: 5491... Training loss: 1.7457... 0.0635 sec/batch
Epoch: 2/20... Training Step: 5492... Training loss: 1.8813... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5493... Training loss: 1.7644... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5494... Training loss: 1.7812... 0.0696 sec/batch
Epoch: 2/20... Training Step: 5495... Training loss: 1.8103... 0.0758 sec/batch
Epoch: 2/20... Training Step: 5496... Training loss: 1.7634... 0.0676 sec/batch
Epoch: 2/20... Training Step: 5497... Training loss: 1.5359... 0.0566 sec/batch
Epoch: 2/20... Training Step: 5498... Training loss: 1.7781... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5499... Training loss: 1.7981... 0.0656 sec/batch
Epoch: 2/20... Training Step: 5500... Training loss: 1.7641... 0.0683 sec/batch
Epoch: 2/20... Training Step: 5501... Training loss: 1.8718... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5502... Training loss: 1.8615... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5503... Training loss: 1.7023... 0.0749 sec/batch
Epoch: 2/20... Training Step: 5504... Training loss: 1.7336... 0.0737 sec/batch
Epoch: 2/20... Training Step: 5505... Training loss: 1.7669... 0.0794 sec/batch
Epoch: 2/20... Training Step: 5506... Training loss: 1.6241... 0.0583 sec/batch
Epoch: 2/20... Training Step: 5507... Training loss: 1.6715... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5508... Training loss: 1.7694... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5509... Training loss: 2.0172... 0.0717 sec/batch
Epoch: 2/20... Training Step: 5510... Training loss: 1.8441... 0.0924 sec/batch
Epoch: 2/20... Training Step: 5511... Training loss: 1.8380... 0.0643 sec/batch
Epoch: 2/20... Training Step: 5512... Training loss: 1.7199... 0.0740 sec/batch
Epoch: 2/20... Training Step: 5513... Training loss: 1.6973... 0.0773 sec/batch
Epoch: 2/20... Training Step: 5514... Training loss: 1.9068... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5515... Training loss: 1.5471... 0.0657 sec/batch
Epoch: 2/20... Training Step: 5516... Training loss: 1.8330... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5517... Training loss: 1.6343... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5518... Training loss: 1.5137... 0.0746 sec/batch
Epoch: 2/20... Training Step: 5519... Training loss: 1.5208... 0.0844 sec/batch
Epoch: 2/20... Training Step: 5520... Training loss: 1.6590... 0.0894 sec/batch
Epoch: 2/20... Training Step: 5521... Training loss: 1.6528... 0.0761 sec/batch
Epoch: 2/20... Training Step: 5522... Training loss: 1.7010... 0.0707 sec/batch
Epoch: 2/20... Training Step: 5523... Training loss: 1.8596... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5524... Training loss: 1.8036... 0.0723 sec/batch
Epoch: 2/20... Training Step: 5525... Training loss: 1.8651... 0.0778 sec/batch
Epoch: 2/20... Training Step: 5526... Training loss: 1.7998... 0.0829 sec/batch
Epoch: 2/20... Training Step: 5527... Training loss: 1.8842... 0.0896 sec/batch
Epoch: 2/20... Training Step: 5528... Training loss: 1.7745... 0.0901 sec/batch
Epoch: 2/20... Training Step: 5529... Training loss: 1.7559... 0.1071 sec/batch
Epoch: 2/20... Training Step: 5530... Training loss: 1.7908... 0.1090 sec/batch
Epoch: 2/20... Training Step: 5531... Training loss: 1.6841... 0.1310 sec/batch
Epoch: 2/20... Training Step: 5532... Training loss: 1.7807... 0.1282 sec/batch
Epoch: 2/20... Training Step: 5533... Training loss: 1.7537... 0.0929 sec/batch
Epoch: 2/20... Training Step: 5534... Training loss: 1.7273... 0.1298 sec/batch
Epoch: 2/20... Training Step: 5535... Training loss: 1.8277... 0.0902 sec/batch
Epoch: 2/20... Training Step: 5536... Training loss: 1.9690... 0.0948 sec/batch
Epoch: 2/20... Training Step: 5537... Training loss: 1.7808... 0.1314 sec/batch
Epoch: 2/20... Training Step: 5538... Training loss: 1.6658... 0.1249 sec/batch
Epoch: 2/20... Training Step: 5539... Training loss: 1.7230... 0.1254 sec/batch
Epoch: 2/20... Training Step: 5540... Training loss: 1.8002... 0.1240 sec/batch
Epoch: 2/20... Training Step: 5541... Training loss: 1.7087... 0.1202 sec/batch
Epoch: 2/20... Training Step: 5542... Training loss: 1.6477... 0.1076 sec/batch
Epoch: 2/20... Training Step: 5543... Training loss: 1.7711... 0.0940 sec/batch
Epoch: 2/20... Training Step: 5544... Training loss: 1.6193... 0.0945 sec/batch
Epoch: 2/20... Training Step: 5545... Training loss: 1.7810... 0.0967 sec/batch
Epoch: 2/20... Training Step: 5546... Training loss: 1.7221... 0.1138 sec/batch
Epoch: 2/20... Training Step: 5547... Training loss: 1.6624... 0.1000 sec/batch
Epoch: 2/20... Training Step: 5548... Training loss: 1.8997... 0.0850 sec/batch
Epoch: 2/20... Training Step: 5549... Training loss: 1.8079... 0.1337 sec/batch
Epoch: 2/20... Training Step: 5550... Training loss: 1.6757... 0.1083 sec/batch
Epoch: 2/20... Training Step: 5551... Training loss: 1.9142... 0.1171 sec/batch
Epoch: 2/20... Training Step: 5552... Training loss: 1.8427... 0.0926 sec/batch
Epoch: 2/20... Training Step: 5553... Training loss: 1.7822... 0.0987 sec/batch
Epoch: 2/20... Training Step: 5554... Training loss: 1.7026... 0.0815 sec/batch
Epoch: 2/20... Training Step: 5555... Training loss: 1.7136... 0.1221 sec/batch
Epoch: 2/20... Training Step: 5556... Training loss: 1.6892... 0.1320 sec/batch
Epoch: 2/20... Training Step: 5557... Training loss: 1.6270... 0.0887 sec/batch
Epoch: 2/20... Training Step: 5558... Training loss: 1.6955... 0.1093 sec/batch
Epoch: 2/20... Training Step: 5559... Training loss: 1.8266... 0.0944 sec/batch
Epoch: 2/20... Training Step: 5560... Training loss: 1.7621... 0.1161 sec/batch
Epoch: 2/20... Training Step: 5561... Training loss: 1.6780... 0.0934 sec/batch
Epoch: 2/20... Training Step: 5562... Training loss: 1.8179... 0.0937 sec/batch
Epoch: 2/20... Training Step: 5563... Training loss: 1.6959... 0.0997 sec/batch
Epoch: 2/20... Training Step: 5564... Training loss: 1.8335... 0.1086 sec/batch
Epoch: 2/20... Training Step: 5565... Training loss: 1.8153... 0.0912 sec/batch
Epoch: 2/20... Training Step: 5566... Training loss: 1.6807... 0.0801 sec/batch
Epoch: 2/20... Training Step: 5567... Training loss: 1.6313... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5568... Training loss: 1.8177... 0.0678 sec/batch
Epoch: 2/20... Training Step: 5569... Training loss: 1.7980... 0.0891 sec/batch
Epoch: 2/20... Training Step: 5570... Training loss: 1.7328... 0.0746 sec/batch
Epoch: 2/20... Training Step: 5571... Training loss: 1.8530... 0.0740 sec/batch
Epoch: 2/20... Training Step: 5572... Training loss: 1.8041... 0.0727 sec/batch
Epoch: 2/20... Training Step: 5573... Training loss: 1.7830... 0.0822 sec/batch
Epoch: 2/20... Training Step: 5574... Training loss: 1.8494... 0.0916 sec/batch
Epoch: 2/20... Training Step: 5575... Training loss: 1.9515... 0.0891 sec/batch
Epoch: 2/20... Training Step: 5576... Training loss: 1.8142... 0.0743 sec/batch
Epoch: 2/20... Training Step: 5577... Training loss: 1.9756... 0.0833 sec/batch
Epoch: 2/20... Training Step: 5578... Training loss: 1.9520... 0.0978 sec/batch
Epoch: 2/20... Training Step: 5579... Training loss: 1.8662... 0.0807 sec/batch
Epoch: 2/20... Training Step: 5580... Training loss: 1.7845... 0.0811 sec/batch
Epoch: 2/20... Training Step: 5581... Training loss: 1.9523... 0.0776 sec/batch
Epoch: 2/20... Training Step: 5582... Training loss: 1.7511... 0.0892 sec/batch
Epoch: 2/20... Training Step: 5583... Training loss: 1.8191... 0.0815 sec/batch
Epoch: 2/20... Training Step: 5584... Training loss: 1.8039... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5585... Training loss: 1.8498... 0.0605 sec/batch
Epoch: 2/20... Training Step: 5586... Training loss: 1.7445... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5587... Training loss: 1.7848... 0.0919 sec/batch
Epoch: 2/20... Training Step: 5588... Training loss: 1.9270... 0.0674 sec/batch
Epoch: 2/20... Training Step: 5589... Training loss: 1.8403... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5590... Training loss: 1.8592... 0.0806 sec/batch
Epoch: 2/20... Training Step: 5591... Training loss: 1.8318... 0.0835 sec/batch
Epoch: 2/20... Training Step: 5592... Training loss: 1.7317... 0.0903 sec/batch
Epoch: 2/20... Training Step: 5593... Training loss: 1.7192... 0.0749 sec/batch
Epoch: 2/20... Training Step: 5594... Training loss: 1.6905... 0.0989 sec/batch
Epoch: 2/20... Training Step: 5595... Training loss: 1.8049... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5596... Training loss: 1.7725... 0.0934 sec/batch
Epoch: 2/20... Training Step: 5597... Training loss: 1.6265... 0.0932 sec/batch
Epoch: 2/20... Training Step: 5598... Training loss: 1.8232... 0.0887 sec/batch
Epoch: 2/20... Training Step: 5599... Training loss: 1.7716... 0.0850 sec/batch
Epoch: 2/20... Training Step: 5600... Training loss: 1.8127... 0.0940 sec/batch
Epoch: 2/20... Training Step: 5601... Training loss: 1.7514... 0.0983 sec/batch
Epoch: 2/20... Training Step: 5602... Training loss: 1.6262... 0.0718 sec/batch
Epoch: 2/20... Training Step: 5603... Training loss: 1.7228... 0.0732 sec/batch
Epoch: 2/20... Training Step: 5604... Training loss: 1.6914... 0.0759 sec/batch
Epoch: 2/20... Training Step: 5605... Training loss: 1.7660... 0.0779 sec/batch
Epoch: 2/20... Training Step: 5606... Training loss: 1.8354... 0.0819 sec/batch
Epoch: 2/20... Training Step: 5607... Training loss: 1.7868... 0.0867 sec/batch
Epoch: 2/20... Training Step: 5608... Training loss: 1.6973... 0.0868 sec/batch
Epoch: 2/20... Training Step: 5609... Training loss: 1.6916... 0.0901 sec/batch
Epoch: 2/20... Training Step: 5610... Training loss: 1.7424... 0.0792 sec/batch
Epoch: 2/20... Training Step: 5611... Training loss: 1.5970... 0.0746 sec/batch
Epoch: 2/20... Training Step: 5612... Training loss: 1.7272... 0.0767 sec/batch
Epoch: 2/20... Training Step: 5613... Training loss: 1.6887... 0.0843 sec/batch
Epoch: 2/20... Training Step: 5614... Training loss: 1.6671... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5615... Training loss: 1.7144... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5616... Training loss: 1.7351... 0.0922 sec/batch
Epoch: 2/20... Training Step: 5617... Training loss: 1.7135... 0.0767 sec/batch
Epoch: 2/20... Training Step: 5618... Training loss: 1.8545... 0.0714 sec/batch
Epoch: 2/20... Training Step: 5619... Training loss: 1.6098... 0.0671 sec/batch
Epoch: 2/20... Training Step: 5620... Training loss: 1.6612... 0.0677 sec/batch
Epoch: 2/20... Training Step: 5621... Training loss: 1.7786... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5622... Training loss: 1.8752... 0.0738 sec/batch
Epoch: 2/20... Training Step: 5623... Training loss: 1.7577... 0.0874 sec/batch
Epoch: 2/20... Training Step: 5624... Training loss: 1.6767... 0.0802 sec/batch
Epoch: 2/20... Training Step: 5625... Training loss: 1.6340... 0.0710 sec/batch
Epoch: 2/20... Training Step: 5626... Training loss: 1.6076... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5627... Training loss: 1.7414... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5628... Training loss: 1.7793... 0.0671 sec/batch
Epoch: 2/20... Training Step: 5629... Training loss: 1.7852... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5630... Training loss: 1.7142... 0.0690 sec/batch
Epoch: 2/20... Training Step: 5631... Training loss: 1.8157... 0.0690 sec/batch
Epoch: 2/20... Training Step: 5632... Training loss: 1.6690... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5633... Training loss: 1.8247... 0.0759 sec/batch
Epoch: 2/20... Training Step: 5634... Training loss: 1.8178... 0.0587 sec/batch
Epoch: 2/20... Training Step: 5635... Training loss: 2.0684... 0.0722 sec/batch
Epoch: 2/20... Training Step: 5636... Training loss: 1.8122... 0.0609 sec/batch
Epoch: 2/20... Training Step: 5637... Training loss: 1.7501... 0.0781 sec/batch
Epoch: 2/20... Training Step: 5638... Training loss: 1.8603... 0.0578 sec/batch
Epoch: 2/20... Training Step: 5639... Training loss: 1.7546... 0.0927 sec/batch
Epoch: 2/20... Training Step: 5640... Training loss: 1.7999... 0.0693 sec/batch
Epoch: 2/20... Training Step: 5641... Training loss: 1.6972... 0.0875 sec/batch
Epoch: 2/20... Training Step: 5642... Training loss: 1.8091... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5643... Training loss: 1.7718... 0.0645 sec/batch
Epoch: 2/20... Training Step: 5644... Training loss: 1.7409... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5645... Training loss: 1.9611... 0.0576 sec/batch
Epoch: 2/20... Training Step: 5646... Training loss: 1.7818... 0.0871 sec/batch
Epoch: 2/20... Training Step: 5647... Training loss: 1.7929... 0.0709 sec/batch
Epoch: 2/20... Training Step: 5648... Training loss: 1.8148... 0.0947 sec/batch
Epoch: 2/20... Training Step: 5649... Training loss: 1.8228... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5650... Training loss: 1.6577... 0.0772 sec/batch
Epoch: 2/20... Training Step: 5651... Training loss: 1.8425... 0.0848 sec/batch
Epoch: 2/20... Training Step: 5652... Training loss: 1.7248... 0.0715 sec/batch
Epoch: 2/20... Training Step: 5653... Training loss: 1.8139... 0.0869 sec/batch
Epoch: 2/20... Training Step: 5654... Training loss: 1.7428... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5655... Training loss: 1.7357... 0.0820 sec/batch
Epoch: 2/20... Training Step: 5656... Training loss: 1.7776... 0.0779 sec/batch
Epoch: 2/20... Training Step: 5657... Training loss: 1.7260... 0.0707 sec/batch
Epoch: 2/20... Training Step: 5658... Training loss: 1.8479... 0.0620 sec/batch
Epoch: 2/20... Training Step: 5659... Training loss: 1.7932... 0.0684 sec/batch
Epoch: 2/20... Training Step: 5660... Training loss: 1.6879... 0.0649 sec/batch
Epoch: 2/20... Training Step: 5661... Training loss: 1.7450... 0.0705 sec/batch
Epoch: 2/20... Training Step: 5662... Training loss: 1.7394... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5663... Training loss: 1.6166... 0.0616 sec/batch
Epoch: 2/20... Training Step: 5664... Training loss: 1.7086... 0.0919 sec/batch
Epoch: 2/20... Training Step: 5665... Training loss: 1.8294... 0.0859 sec/batch
Epoch: 2/20... Training Step: 5666... Training loss: 1.7277... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5667... Training loss: 1.6389... 0.0816 sec/batch
Epoch: 2/20... Training Step: 5668... Training loss: 1.8463... 0.0584 sec/batch
Epoch: 2/20... Training Step: 5669... Training loss: 1.7897... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5670... Training loss: 1.7747... 0.0649 sec/batch
Epoch: 2/20... Training Step: 5671... Training loss: 1.5838... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5672... Training loss: 1.7212... 0.0686 sec/batch
Epoch: 2/20... Training Step: 5673... Training loss: 1.7454... 0.0670 sec/batch
Epoch: 2/20... Training Step: 5674... Training loss: 1.7917... 0.0603 sec/batch
Epoch: 2/20... Training Step: 5675... Training loss: 1.8486... 0.0725 sec/batch
Epoch: 2/20... Training Step: 5676... Training loss: 1.7939... 0.0493 sec/batch
Epoch: 2/20... Training Step: 5677... Training loss: 1.7310... 0.0737 sec/batch
Epoch: 2/20... Training Step: 5678... Training loss: 1.6896... 0.0562 sec/batch
Epoch: 2/20... Training Step: 5679... Training loss: 1.8161... 0.0718 sec/batch
Epoch: 2/20... Training Step: 5680... Training loss: 1.7273... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5681... Training loss: 1.7274... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5682... Training loss: 1.7615... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5683... Training loss: 1.6429... 0.0694 sec/batch
Epoch: 2/20... Training Step: 5684... Training loss: 1.8328... 0.0626 sec/batch
Epoch: 2/20... Training Step: 5685... Training loss: 1.8252... 0.0664 sec/batch
Epoch: 2/20... Training Step: 5686... Training loss: 1.8767... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5687... Training loss: 1.6245... 0.0557 sec/batch
Epoch: 2/20... Training Step: 5688... Training loss: 1.7710... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5689... Training loss: 1.7027... 0.0783 sec/batch
Epoch: 2/20... Training Step: 5690... Training loss: 1.6945... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5691... Training loss: 1.7879... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5692... Training loss: 1.7618... 0.0621 sec/batch
Epoch: 2/20... Training Step: 5693... Training loss: 1.6734... 0.0659 sec/batch
Epoch: 2/20... Training Step: 5694... Training loss: 1.5757... 0.0614 sec/batch
Epoch: 2/20... Training Step: 5695... Training loss: 1.6847... 0.0691 sec/batch
Epoch: 2/20... Training Step: 5696... Training loss: 1.7495... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5697... Training loss: 1.6246... 0.0632 sec/batch
Epoch: 2/20... Training Step: 5698... Training loss: 1.7416... 0.0746 sec/batch
Epoch: 2/20... Training Step: 5699... Training loss: 1.7523... 0.0574 sec/batch
Epoch: 2/20... Training Step: 5700... Training loss: 1.6628... 0.0562 sec/batch
Epoch: 2/20... Training Step: 5701... Training loss: 1.7872... 0.0576 sec/batch
Epoch: 2/20... Training Step: 5702... Training loss: 1.8645... 0.0625 sec/batch
Epoch: 2/20... Training Step: 5703... Training loss: 1.8152... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5704... Training loss: 1.8905... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5705... Training loss: 1.7932... 0.0769 sec/batch
Epoch: 2/20... Training Step: 5706... Training loss: 1.7216... 0.0606 sec/batch
Epoch: 2/20... Training Step: 5707... Training loss: 1.7207... 0.0768 sec/batch
Epoch: 2/20... Training Step: 5708... Training loss: 1.7233... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5709... Training loss: 1.7651... 0.0730 sec/batch
Epoch: 2/20... Training Step: 5710... Training loss: 1.7516... 0.0751 sec/batch
Epoch: 2/20... Training Step: 5711... Training loss: 1.7143... 0.0602 sec/batch
Epoch: 2/20... Training Step: 5712... Training loss: 1.7402... 0.0642 sec/batch
Epoch: 2/20... Training Step: 5713... Training loss: 1.8472... 0.0643 sec/batch
Epoch: 2/20... Training Step: 5714... Training loss: 1.6597... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5715... Training loss: 1.7248... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5716... Training loss: 1.5717... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5717... Training loss: 1.8108... 0.0638 sec/batch
Epoch: 2/20... Training Step: 5718... Training loss: 1.8520... 0.0706 sec/batch
Epoch: 2/20... Training Step: 5719... Training loss: 1.5918... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5720... Training loss: 1.7865... 0.0641 sec/batch
Epoch: 2/20... Training Step: 5721... Training loss: 1.7711... 0.0848 sec/batch
Epoch: 2/20... Training Step: 5722... Training loss: 1.6666... 0.0794 sec/batch
Epoch: 2/20... Training Step: 5723... Training loss: 1.6542... 0.0696 sec/batch
Epoch: 2/20... Training Step: 5724... Training loss: 1.8110... 0.0701 sec/batch
Epoch: 2/20... Training Step: 5725... Training loss: 1.7019... 0.0806 sec/batch
Epoch: 2/20... Training Step: 5726... Training loss: 1.5694... 0.0694 sec/batch
Epoch: 2/20... Training Step: 5727... Training loss: 1.6783... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5728... Training loss: 1.7568... 0.0686 sec/batch
Epoch: 2/20... Training Step: 5729... Training loss: 1.7908... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5730... Training loss: 1.6624... 0.0677 sec/batch
Epoch: 2/20... Training Step: 5731... Training loss: 1.8025... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5732... Training loss: 1.7866... 0.0715 sec/batch
Epoch: 2/20... Training Step: 5733... Training loss: 1.6933... 0.0856 sec/batch
Epoch: 2/20... Training Step: 5734... Training loss: 1.7287... 0.0738 sec/batch
Epoch: 2/20... Training Step: 5735... Training loss: 1.7343... 0.0765 sec/batch
Epoch: 2/20... Training Step: 5736... Training loss: 1.7724... 0.0762 sec/batch
Epoch: 2/20... Training Step: 5737... Training loss: 1.5935... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5738... Training loss: 1.7590... 0.0769 sec/batch
Epoch: 2/20... Training Step: 5739... Training loss: 1.8716... 0.0753 sec/batch
Epoch: 2/20... Training Step: 5740... Training loss: 1.7563... 0.0638 sec/batch
Epoch: 2/20... Training Step: 5741... Training loss: 1.6379... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5742... Training loss: 1.6438... 0.0593 sec/batch
Epoch: 2/20... Training Step: 5743... Training loss: 1.7079... 0.0750 sec/batch
Epoch: 2/20... Training Step: 5744... Training loss: 1.8666... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5745... Training loss: 1.8209... 0.0786 sec/batch
Epoch: 2/20... Training Step: 5746... Training loss: 1.6180... 0.0687 sec/batch
Epoch: 2/20... Training Step: 5747... Training loss: 1.7335... 0.0734 sec/batch
Epoch: 2/20... Training Step: 5748... Training loss: 1.6487... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5749... Training loss: 1.6615... 0.0795 sec/batch
Epoch: 2/20... Training Step: 5750... Training loss: 1.7549... 0.0609 sec/batch
Epoch: 2/20... Training Step: 5751... Training loss: 1.7592... 0.0580 sec/batch
Epoch: 2/20... Training Step: 5752... Training loss: 1.6334... 0.0770 sec/batch
Epoch: 2/20... Training Step: 5753... Training loss: 1.7383... 0.0874 sec/batch
Epoch: 2/20... Training Step: 5754... Training loss: 1.6338... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5755... Training loss: 1.7355... 0.0773 sec/batch
Epoch: 2/20... Training Step: 5756... Training loss: 1.8067... 0.0657 sec/batch
Epoch: 2/20... Training Step: 5757... Training loss: 1.7464... 0.0741 sec/batch
Epoch: 2/20... Training Step: 5758... Training loss: 1.6519... 0.0668 sec/batch
Epoch: 2/20... Training Step: 5759... Training loss: 1.7757... 0.0661 sec/batch
Epoch: 2/20... Training Step: 5760... Training loss: 1.6785... 0.0822 sec/batch
Epoch: 2/20... Training Step: 5761... Training loss: 1.7264... 0.0605 sec/batch
Epoch: 2/20... Training Step: 5762... Training loss: 1.9254... 0.0682 sec/batch
Epoch: 2/20... Training Step: 5763... Training loss: 1.6471... 0.0636 sec/batch
Epoch: 2/20... Training Step: 5764... Training loss: 1.6064... 0.0703 sec/batch
Epoch: 2/20... Training Step: 5765... Training loss: 1.5466... 0.0687 sec/batch
Epoch: 2/20... Training Step: 5766... Training loss: 1.7799... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5767... Training loss: 1.5092... 0.0694 sec/batch
Epoch: 2/20... Training Step: 5768... Training loss: 1.6842... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5769... Training loss: 1.6796... 0.0617 sec/batch
Epoch: 2/20... Training Step: 5770... Training loss: 1.7183... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5771... Training loss: 1.7108... 0.0846 sec/batch
Epoch: 2/20... Training Step: 5772... Training loss: 1.7196... 0.0807 sec/batch
Epoch: 2/20... Training Step: 5773... Training loss: 1.7443... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5774... Training loss: 1.6957... 0.0768 sec/batch
Epoch: 2/20... Training Step: 5775... Training loss: 1.6397... 0.0614 sec/batch
Epoch: 2/20... Training Step: 5776... Training loss: 1.7021... 0.0771 sec/batch
Epoch: 2/20... Training Step: 5777... Training loss: 1.6280... 0.0899 sec/batch
Epoch: 2/20... Training Step: 5778... Training loss: 1.7142... 0.0843 sec/batch
Epoch: 2/20... Training Step: 5779... Training loss: 1.6683... 0.0626 sec/batch
Epoch: 2/20... Training Step: 5780... Training loss: 1.5112... 0.0610 sec/batch
Epoch: 2/20... Training Step: 5781... Training loss: 1.6459... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5782... Training loss: 1.7840... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5783... Training loss: 1.6499... 0.0528 sec/batch
Epoch: 2/20... Training Step: 5784... Training loss: 1.7544... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5785... Training loss: 1.7455... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5786... Training loss: 1.8714... 0.0621 sec/batch
Epoch: 2/20... Training Step: 5787... Training loss: 1.7142... 0.0654 sec/batch
Epoch: 2/20... Training Step: 5788... Training loss: 1.6841... 0.0681 sec/batch
Epoch: 2/20... Training Step: 5789... Training loss: 1.6993... 0.0635 sec/batch
Epoch: 2/20... Training Step: 5790... Training loss: 1.7527... 0.0576 sec/batch
Epoch: 2/20... Training Step: 5791... Training loss: 1.7508... 0.0712 sec/batch
Epoch: 2/20... Training Step: 5792... Training loss: 1.8006... 0.0725 sec/batch
Epoch: 2/20... Training Step: 5793... Training loss: 1.7134... 0.0657 sec/batch
Epoch: 2/20... Training Step: 5794... Training loss: 1.6831... 0.0870 sec/batch
Epoch: 2/20... Training Step: 5795... Training loss: 1.8135... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5796... Training loss: 1.6604... 0.0566 sec/batch
Epoch: 2/20... Training Step: 5797... Training loss: 1.8245... 0.0741 sec/batch
Epoch: 2/20... Training Step: 5798... Training loss: 1.6781... 0.0626 sec/batch
Epoch: 2/20... Training Step: 5799... Training loss: 1.8449... 0.0614 sec/batch
Epoch: 2/20... Training Step: 5800... Training loss: 1.7400... 0.0624 sec/batch
Epoch: 2/20... Training Step: 5801... Training loss: 1.8094... 0.0722 sec/batch
Epoch: 2/20... Training Step: 5802... Training loss: 1.7818... 0.0679 sec/batch
Epoch: 2/20... Training Step: 5803... Training loss: 1.6997... 0.0763 sec/batch
Epoch: 2/20... Training Step: 5804... Training loss: 1.8420... 0.0631 sec/batch
Epoch: 2/20... Training Step: 5805... Training loss: 1.7191... 0.0565 sec/batch
Epoch: 2/20... Training Step: 5806... Training loss: 1.7255... 0.0587 sec/batch
Epoch: 2/20... Training Step: 5807... Training loss: 1.7395... 0.0535 sec/batch
Epoch: 2/20... Training Step: 5808... Training loss: 1.6919... 0.0595 sec/batch
Epoch: 2/20... Training Step: 5809... Training loss: 1.6303... 0.0553 sec/batch
Epoch: 2/20... Training Step: 5810... Training loss: 1.7059... 0.0626 sec/batch
Epoch: 2/20... Training Step: 5811... Training loss: 1.6837... 0.0528 sec/batch
Epoch: 2/20... Training Step: 5812... Training loss: 1.7979... 0.0739 sec/batch
Epoch: 2/20... Training Step: 5813... Training loss: 1.5459... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5814... Training loss: 1.5409... 0.0667 sec/batch
Epoch: 2/20... Training Step: 5815... Training loss: 1.7614... 0.0599 sec/batch
Epoch: 2/20... Training Step: 5816... Training loss: 1.7928... 0.0804 sec/batch
Epoch: 2/20... Training Step: 5817... Training loss: 1.7402... 0.0633 sec/batch
Epoch: 2/20... Training Step: 5818... Training loss: 1.7780... 0.0537 sec/batch
Epoch: 2/20... Training Step: 5819... Training loss: 1.6060... 0.0685 sec/batch
Epoch: 2/20... Training Step: 5820... Training loss: 1.8439... 0.0621 sec/batch
Epoch: 2/20... Training Step: 5821... Training loss: 1.7058... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5822... Training loss: 1.8223... 0.0740 sec/batch
Epoch: 2/20... Training Step: 5823... Training loss: 2.0034... 0.0690 sec/batch
Epoch: 2/20... Training Step: 5824... Training loss: 1.8605... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5825... Training loss: 1.7366... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5826... Training loss: 1.5680... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5827... Training loss: 2.0175... 0.0593 sec/batch
Epoch: 2/20... Training Step: 5828... Training loss: 1.8520... 0.0890 sec/batch
Epoch: 2/20... Training Step: 5829... Training loss: 1.6835... 0.0784 sec/batch
Epoch: 2/20... Training Step: 5830... Training loss: 1.7626... 0.0627 sec/batch
Epoch: 2/20... Training Step: 5831... Training loss: 1.8585... 0.0666 sec/batch
Epoch: 2/20... Training Step: 5832... Training loss: 1.7190... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5833... Training loss: 1.7514... 0.0601 sec/batch
Epoch: 2/20... Training Step: 5834... Training loss: 1.8030... 0.0590 sec/batch
Epoch: 2/20... Training Step: 5835... Training loss: 1.6691... 0.0716 sec/batch
Epoch: 2/20... Training Step: 5836... Training loss: 1.6292... 0.0777 sec/batch
Epoch: 2/20... Training Step: 5837... Training loss: 1.8454... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5838... Training loss: 1.6614... 0.0854 sec/batch
Epoch: 2/20... Training Step: 5839... Training loss: 1.7320... 0.0703 sec/batch
Epoch: 2/20... Training Step: 5840... Training loss: 1.6401... 0.0699 sec/batch
Epoch: 2/20... Training Step: 5841... Training loss: 1.6956... 0.0828 sec/batch
Epoch: 2/20... Training Step: 5842... Training loss: 1.5394... 0.0820 sec/batch
Epoch: 2/20... Training Step: 5843... Training loss: 1.6952... 0.0728 sec/batch
Epoch: 2/20... Training Step: 5844... Training loss: 1.6655... 0.0745 sec/batch
Epoch: 2/20... Training Step: 5845... Training loss: 1.7255... 0.0754 sec/batch
Epoch: 2/20... Training Step: 5846... Training loss: 1.8817... 0.0750 sec/batch
Epoch: 2/20... Training Step: 5847... Training loss: 1.7269... 0.0672 sec/batch
Epoch: 2/20... Training Step: 5848... Training loss: 1.6932... 0.0802 sec/batch
Epoch: 2/20... Training Step: 5849... Training loss: 1.6333... 0.0754 sec/batch
Epoch: 2/20... Training Step: 5850... Training loss: 1.8843... 0.0784 sec/batch
Epoch: 2/20... Training Step: 5851... Training loss: 1.8636... 0.0736 sec/batch
Epoch: 2/20... Training Step: 5852... Training loss: 1.6112... 0.0756 sec/batch
Epoch: 2/20... Training Step: 5853... Training loss: 1.8557... 0.0891 sec/batch
Epoch: 2/20... Training Step: 5854... Training loss: 1.8756... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5855... Training loss: 1.5806... 0.0717 sec/batch
Epoch: 2/20... Training Step: 5856... Training loss: 1.7703... 0.0753 sec/batch
Epoch: 2/20... Training Step: 5857... Training loss: 1.8093... 0.0795 sec/batch
Epoch: 2/20... Training Step: 5858... Training loss: 1.8565... 0.0821 sec/batch
Epoch: 2/20... Training Step: 5859... Training loss: 1.7289... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5860... Training loss: 1.7399... 0.0660 sec/batch
Epoch: 2/20... Training Step: 5861... Training loss: 1.8947... 0.0634 sec/batch
Epoch: 2/20... Training Step: 5862... Training loss: 1.7943... 0.0966 sec/batch
Epoch: 2/20... Training Step: 5863... Training loss: 1.7579... 0.0824 sec/batch
Epoch: 2/20... Training Step: 5864... Training loss: 1.7085... 0.0751 sec/batch
Epoch: 2/20... Training Step: 5865... Training loss: 1.6292... 0.0866 sec/batch
Epoch: 2/20... Training Step: 5866... Training loss: 1.7110... 0.0782 sec/batch
Epoch: 2/20... Training Step: 5867... Training loss: 1.7833... 0.0891 sec/batch
Epoch: 2/20... Training Step: 5868... Training loss: 1.7873... 0.0723 sec/batch
Epoch: 2/20... Training Step: 5869... Training loss: 1.7149... 0.0688 sec/batch
Epoch: 2/20... Training Step: 5870... Training loss: 1.6300... 0.0708 sec/batch
Epoch: 2/20... Training Step: 5871... Training loss: 1.7450... 0.0650 sec/batch
Epoch: 2/20... Training Step: 5872... Training loss: 1.7875... 0.0821 sec/batch
Epoch: 2/20... Training Step: 5873... Training loss: 1.6915... 0.0708 sec/batch
Epoch: 2/20... Training Step: 5874... Training loss: 1.5225... 0.0791 sec/batch
Epoch: 2/20... Training Step: 5875... Training loss: 1.6356... 0.0804 sec/batch
Epoch: 2/20... Training Step: 5876... Training loss: 1.8176... 0.0691 sec/batch
Epoch: 2/20... Training Step: 5877... Training loss: 1.7316... 0.0719 sec/batch
Epoch: 2/20... Training Step: 5878... Training loss: 1.8113... 0.0801 sec/batch
Epoch: 2/20... Training Step: 5879... Training loss: 1.7568... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5880... Training loss: 1.6406... 0.0638 sec/batch
Epoch: 2/20... Training Step: 5881... Training loss: 1.6913... 0.0724 sec/batch
Epoch: 2/20... Training Step: 5882... Training loss: 1.8125... 0.0678 sec/batch
Epoch: 2/20... Training Step: 5883... Training loss: 1.7764... 0.0674 sec/batch
Epoch: 2/20... Training Step: 5884... Training loss: 1.7562... 0.0704 sec/batch
Epoch: 2/20... Training Step: 5885... Training loss: 1.7151... 0.0805 sec/batch
Epoch: 2/20... Training Step: 5886... Training loss: 1.8067... 0.0672 sec/batch
Epoch: 2/20... Training Step: 5887... Training loss: 1.6936... 0.0635 sec/batch
Epoch: 2/20... Training Step: 5888... Training loss: 1.8597... 0.0825 sec/batch
Epoch: 2/20... Training Step: 5889... Training loss: 1.6553... 0.0773 sec/batch
Epoch: 2/20... Training Step: 5890... Training loss: 1.7515... 0.0680 sec/batch
Epoch: 2/20... Training Step: 5891... Training loss: 1.8768... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5892... Training loss: 1.5651... 0.0639 sec/batch
Epoch: 2/20... Training Step: 5893... Training loss: 1.6830... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5894... Training loss: 1.6375... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5895... Training loss: 1.6970... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5896... Training loss: 1.6135... 0.0585 sec/batch
Epoch: 2/20... Training Step: 5897... Training loss: 1.7078... 0.0635 sec/batch
Epoch: 2/20... Training Step: 5898... Training loss: 1.7809... 0.0544 sec/batch
Epoch: 2/20... Training Step: 5899... Training loss: 1.7898... 0.0793 sec/batch
Epoch: 2/20... Training Step: 5900... Training loss: 1.7044... 0.0557 sec/batch
Epoch: 2/20... Training Step: 5901... Training loss: 1.6654... 0.0621 sec/batch
Epoch: 2/20... Training Step: 5902... Training loss: 1.7367... 0.0679 sec/batch
Epoch: 2/20... Training Step: 5903... Training loss: 1.7088... 0.0658 sec/batch
Epoch: 2/20... Training Step: 5904... Training loss: 1.9708... 0.0539 sec/batch
Epoch: 2/20... Training Step: 5905... Training loss: 1.7756... 0.0622 sec/batch
Epoch: 2/20... Training Step: 5906... Training loss: 1.7351... 0.0646 sec/batch
Epoch: 2/20... Training Step: 5907... Training loss: 1.8575... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5908... Training loss: 2.0828... 0.0748 sec/batch
Epoch: 2/20... Training Step: 5909... Training loss: 1.8125... 0.0563 sec/batch
Epoch: 2/20... Training Step: 5910... Training loss: 1.9643... 0.0714 sec/batch
Epoch: 2/20... Training Step: 5911... Training loss: 1.7336... 0.0847 sec/batch
Epoch: 2/20... Training Step: 5912... Training loss: 1.6626... 0.0662 sec/batch
Epoch: 2/20... Training Step: 5913... Training loss: 1.7363... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5914... Training loss: 1.6573... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5915... Training loss: 1.8695... 0.0831 sec/batch
Epoch: 2/20... Training Step: 5916... Training loss: 1.6321... 0.0764 sec/batch
Epoch: 2/20... Training Step: 5917... Training loss: 1.7455... 0.0634 sec/batch
Epoch: 2/20... Training Step: 5918... Training loss: 1.6590... 0.0563 sec/batch
Epoch: 2/20... Training Step: 5919... Training loss: 1.7801... 0.0742 sec/batch
Epoch: 2/20... Training Step: 5920... Training loss: 1.6970... 0.0712 sec/batch
Epoch: 2/20... Training Step: 5921... Training loss: 1.7703... 0.0837 sec/batch
Epoch: 2/20... Training Step: 5922... Training loss: 1.7482... 0.0612 sec/batch
Epoch: 2/20... Training Step: 5923... Training loss: 1.7856... 0.0756 sec/batch
Epoch: 2/20... Training Step: 5924... Training loss: 1.7179... 0.0702 sec/batch
Epoch: 2/20... Training Step: 5925... Training loss: 1.8857... 0.0726 sec/batch
Epoch: 2/20... Training Step: 5926... Training loss: 1.8869... 0.0698 sec/batch
Epoch: 2/20... Training Step: 5927... Training loss: 1.7548... 0.0634 sec/batch
Epoch: 2/20... Training Step: 5928... Training loss: 1.6944... 0.0694 sec/batch
Epoch: 2/20... Training Step: 5929... Training loss: 1.7314... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5930... Training loss: 1.7980... 0.0720 sec/batch
Epoch: 2/20... Training Step: 5931... Training loss: 1.6684... 0.0794 sec/batch
Epoch: 2/20... Training Step: 5932... Training loss: 1.7909... 0.0720 sec/batch
Epoch: 2/20... Training Step: 5933... Training loss: 1.5819... 0.0679 sec/batch
Epoch: 2/20... Training Step: 5934... Training loss: 1.5195... 0.0775 sec/batch
Epoch: 2/20... Training Step: 5935... Training loss: 1.7158... 0.0737 sec/batch
Epoch: 2/20... Training Step: 5936... Training loss: 1.6182... 0.0783 sec/batch
Epoch: 2/20... Training Step: 5937... Training loss: 1.8304... 0.0720 sec/batch
Epoch: 2/20... Training Step: 5938... Training loss: 1.8360... 0.0793 sec/batch
Epoch: 2/20... Training Step: 5939... Training loss: 1.8922... 0.0779 sec/batch
Epoch: 2/20... Training Step: 5940... Training loss: 1.8033... 0.0689 sec/batch
Epoch: 2/20... Training Step: 5941... Training loss: 1.6949... 0.0665 sec/batch
Epoch: 2/20... Training Step: 5942... Training loss: 1.7539... 0.0750 sec/batch
Epoch: 2/20... Training Step: 5943... Training loss: 1.8340... 0.0676 sec/batch
Epoch: 2/20... Training Step: 5944... Training loss: 1.7542... 0.0754 sec/batch
Epoch: 2/20... Training Step: 5945... Training loss: 1.8672... 0.0768 sec/batch
Epoch: 2/20... Training Step: 5946... Training loss: 1.6560... 0.0723 sec/batch
Epoch: 2/20... Training Step: 5947... Training loss: 1.7682... 0.0788 sec/batch
Epoch: 2/20... Training Step: 5948... Training loss: 1.7035... 0.0892 sec/batch
Epoch: 2/20... Training Step: 5949... Training loss: 1.7658... 0.0674 sec/batch
Epoch: 2/20... Training Step: 5950... Training loss: 1.6807... 0.0831 sec/batch
Epoch: 2/20... Training Step: 5951... Training loss: 1.7916... 0.0751 sec/batch
Epoch: 2/20... Training Step: 5952... Training loss: 1.6561... 0.0629 sec/batch
Epoch: 2/20... Training Step: 5953... Training loss: 1.7517... 0.0683 sec/batch
Epoch: 2/20... Training Step: 5954... Training loss: 1.9493... 0.0669 sec/batch
Epoch: 2/20... Training Step: 5955... Training loss: 1.9728... 0.0718 sec/batch
Epoch: 2/20... Training Step: 5956... Training loss: 1.7156... 0.0657 sec/batch
Epoch: 2/20... Training Step: 5957... Training loss: 1.6002... 0.0868 sec/batch
Epoch: 2/20... Training Step: 5958... Training loss: 1.7506... 0.0792 sec/batch
Epoch: 2/20... Training Step: 5959... Training loss: 1.6940... 0.0575 sec/batch
Epoch: 2/20... Training Step: 5960... Training loss: 1.6960... 0.0719 sec/batch
Epoch: 2/20... Training Step: 5961... Training loss: 1.6035... 0.0623 sec/batch
Epoch: 2/20... Training Step: 5962... Training loss: 1.7925... 0.0651 sec/batch
Epoch: 2/20... Training Step: 5963... Training loss: 1.9585... 0.0889 sec/batch
Epoch: 2/20... Training Step: 5964... Training loss: 1.8967... 0.0748 sec/batch
Epoch: 2/20... Training Step: 5965... Training loss: 1.7746... 0.0888 sec/batch
Epoch: 2/20... Training Step: 5966... Training loss: 1.8791... 0.0780 sec/batch
Epoch: 2/20... Training Step: 5967... Training loss: 1.7357... 0.0777 sec/batch
Epoch: 2/20... Training Step: 5968... Training loss: 1.7257... 0.0721 sec/batch
Epoch: 2/20... Training Step: 5969... Training loss: 1.7267... 0.0647 sec/batch
Epoch: 2/20... Training Step: 5970... Training loss: 1.7477... 0.0847 sec/batch
Epoch: 2/20... Training Step: 5971... Training loss: 1.6510... 0.0829 sec/batch
Epoch: 2/20... Training Step: 5972... Training loss: 1.5060... 0.0844 sec/batch
Epoch: 2/20... Training Step: 5973... Training loss: 1.5616... 0.0787 sec/batch
Epoch: 2/20... Training Step: 5974... Training loss: 1.5861... 0.0813 sec/batch
Epoch: 2/20... Training Step: 5975... Training loss: 1.7356... 0.0879 sec/batch
Epoch: 2/20... Training Step: 5976... Training loss: 1.5986... 0.0811 sec/batch
Epoch: 2/20... Training Step: 5977... Training loss: 1.7921... 0.0798 sec/batch
Epoch: 2/20... Training Step: 5978... Training loss: 1.6788... 0.0853 sec/batch
Epoch: 2/20... Training Step: 5979... Training loss: 1.5983... 0.0748 sec/batch
Epoch: 2/20... Training Step: 5980... Training loss: 1.8009... 0.0920 sec/batch
Epoch: 2/20... Training Step: 5981... Training loss: 1.6302... 0.0753 sec/batch
Epoch: 2/20... Training Step: 5982... Training loss: 1.6745... 0.0685 sec/batch
Epoch: 2/20... Training Step: 5983... Training loss: 1.6839... 0.0708 sec/batch
Epoch: 2/20... Training Step: 5984... Training loss: 1.6390... 0.0835 sec/batch
Epoch: 2/20... Training Step: 5985... Training loss: 1.7276... 0.0927 sec/batch
Epoch: 2/20... Training Step: 5986... Training loss: 1.7628... 0.0865 sec/batch
Epoch: 2/20... Training Step: 5987... Training loss: 1.6214... 0.0819 sec/batch
Epoch: 2/20... Training Step: 5988... Training loss: 1.6153... 0.0667 sec/batch
Epoch: 2/20... Training Step: 5989... Training loss: 1.5878... 0.0673 sec/batch
Epoch: 2/20... Training Step: 5990... Training loss: 1.6177... 0.0893 sec/batch
Epoch: 2/20... Training Step: 5991... Training loss: 1.7855... 0.0653 sec/batch
Epoch: 2/20... Training Step: 5992... Training loss: 1.6362... 0.0712 sec/batch
Epoch: 2/20... Training Step: 5993... Training loss: 1.8526... 0.0747 sec/batch
Epoch: 2/20... Training Step: 5994... Training loss: 1.6903... 0.0659 sec/batch
Epoch: 2/20... Training Step: 5995... Training loss: 1.6752... 0.0828 sec/batch
Epoch: 2/20... Training Step: 5996... Training loss: 1.6393... 0.0752 sec/batch
Epoch: 2/20... Training Step: 5997... Training loss: 1.7431... 0.0878 sec/batch
Epoch: 2/20... Training Step: 5998... Training loss: 1.5105... 0.0937 sec/batch
Epoch: 2/20... Training Step: 5999... Training loss: 1.7794... 0.0621 sec/batch
Epoch: 2/20... Training Step: 6000... Training loss: 1.5658... 0.0607 sec/batch
Epoch: 2/20... Training Step: 6001... Training loss: 1.6981... 0.0564 sec/batch
Epoch: 2/20... Training Step: 6002... Training loss: 1.7731... 0.0679 sec/batch
Epoch: 2/20... Training Step: 6003... Training loss: 1.7329... 0.0658 sec/batch
Epoch: 2/20... Training Step: 6004... Training loss: 1.8072... 0.0589 sec/batch
Epoch: 2/20... Training Step: 6005... Training loss: 1.7253... 0.0689 sec/batch
Epoch: 2/20... Training Step: 6006... Training loss: 1.7906... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6007... Training loss: 1.5968... 0.0803 sec/batch
Epoch: 2/20... Training Step: 6008... Training loss: 1.6643... 0.0852 sec/batch
Epoch: 2/20... Training Step: 6009... Training loss: 1.7306... 0.0768 sec/batch
Epoch: 2/20... Training Step: 6010... Training loss: 1.7889... 0.0665 sec/batch
Epoch: 2/20... Training Step: 6011... Training loss: 1.6726... 0.0615 sec/batch
Epoch: 2/20... Training Step: 6012... Training loss: 1.8056... 0.0647 sec/batch
Epoch: 2/20... Training Step: 6013... Training loss: 1.7731... 0.0651 sec/batch
Epoch: 2/20... Training Step: 6014... Training loss: 1.7863... 0.0617 sec/batch
Epoch: 2/20... Training Step: 6015... Training loss: 1.8407... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6016... Training loss: 1.7005... 0.0729 sec/batch
Epoch: 2/20... Training Step: 6017... Training loss: 1.7965... 0.0690 sec/batch
Epoch: 2/20... Training Step: 6018... Training loss: 1.5731... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6019... Training loss: 1.7889... 0.0769 sec/batch
Epoch: 2/20... Training Step: 6020... Training loss: 1.9504... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6021... Training loss: 1.7429... 0.0680 sec/batch
Epoch: 2/20... Training Step: 6022... Training loss: 1.6738... 0.0616 sec/batch
Epoch: 2/20... Training Step: 6023... Training loss: 1.6221... 0.0735 sec/batch
Epoch: 2/20... Training Step: 6024... Training loss: 1.6199... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6025... Training loss: 1.6497... 0.0708 sec/batch
Epoch: 2/20... Training Step: 6026... Training loss: 1.6854... 0.0633 sec/batch
Epoch: 2/20... Training Step: 6027... Training loss: 1.6952... 0.0782 sec/batch
Epoch: 2/20... Training Step: 6028... Training loss: 1.6467... 0.0581 sec/batch
Epoch: 2/20... Training Step: 6029... Training loss: 1.7045... 0.0776 sec/batch
Epoch: 2/20... Training Step: 6030... Training loss: 1.6894... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6031... Training loss: 1.7540... 0.0589 sec/batch
Epoch: 2/20... Training Step: 6032... Training loss: 1.7313... 0.0743 sec/batch
Epoch: 2/20... Training Step: 6033... Training loss: 1.6702... 0.0808 sec/batch
Epoch: 2/20... Training Step: 6034... Training loss: 1.8114... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6035... Training loss: 1.6674... 0.0697 sec/batch
Epoch: 2/20... Training Step: 6036... Training loss: 1.7969... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6037... Training loss: 1.6801... 0.0645 sec/batch
Epoch: 2/20... Training Step: 6038... Training loss: 1.7843... 0.0636 sec/batch
Epoch: 2/20... Training Step: 6039... Training loss: 1.8195... 0.0700 sec/batch
Epoch: 2/20... Training Step: 6040... Training loss: 1.6565... 0.0676 sec/batch
Epoch: 2/20... Training Step: 6041... Training loss: 1.8448... 0.0722 sec/batch
Epoch: 2/20... Training Step: 6042... Training loss: 1.8738... 0.0674 sec/batch
Epoch: 2/20... Training Step: 6043... Training loss: 1.6555... 0.0623 sec/batch
Epoch: 2/20... Training Step: 6044... Training loss: 1.8372... 0.0697 sec/batch
Epoch: 2/20... Training Step: 6045... Training loss: 1.6875... 0.0631 sec/batch
Epoch: 2/20... Training Step: 6046... Training loss: 1.7366... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6047... Training loss: 1.6600... 0.0654 sec/batch
Epoch: 2/20... Training Step: 6048... Training loss: 1.7643... 0.0794 sec/batch
Epoch: 2/20... Training Step: 6049... Training loss: 1.7997... 0.0722 sec/batch
Epoch: 2/20... Training Step: 6050... Training loss: 1.7149... 0.0706 sec/batch
Epoch: 2/20... Training Step: 6051... Training loss: 1.7085... 0.0548 sec/batch
Epoch: 2/20... Training Step: 6052... Training loss: 1.7248... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6053... Training loss: 1.7560... 0.0793 sec/batch
Epoch: 2/20... Training Step: 6054... Training loss: 1.7032... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6055... Training loss: 1.6684... 0.0711 sec/batch
Epoch: 2/20... Training Step: 6056... Training loss: 1.7174... 0.0849 sec/batch
Epoch: 2/20... Training Step: 6057... Training loss: 1.7168... 0.0906 sec/batch
Epoch: 2/20... Training Step: 6058... Training loss: 1.8533... 0.0806 sec/batch
Epoch: 2/20... Training Step: 6059... Training loss: 1.6956... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6060... Training loss: 1.6917... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6061... Training loss: 1.7883... 0.0691 sec/batch
Epoch: 2/20... Training Step: 6062... Training loss: 1.6968... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6063... Training loss: 1.6987... 0.0660 sec/batch
Epoch: 2/20... Training Step: 6064... Training loss: 1.7892... 0.0723 sec/batch
Epoch: 2/20... Training Step: 6065... Training loss: 1.6428... 0.0779 sec/batch
Epoch: 2/20... Training Step: 6066... Training loss: 1.7940... 0.0828 sec/batch
Epoch: 2/20... Training Step: 6067... Training loss: 1.7599... 0.0906 sec/batch
Epoch: 2/20... Training Step: 6068... Training loss: 1.8169... 0.0819 sec/batch
Epoch: 2/20... Training Step: 6069... Training loss: 1.6266... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6070... Training loss: 1.7346... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6071... Training loss: 1.6391... 0.0774 sec/batch
Epoch: 2/20... Training Step: 6072... Training loss: 1.6664... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6073... Training loss: 1.8317... 0.0582 sec/batch
Epoch: 2/20... Training Step: 6074... Training loss: 1.8150... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6075... Training loss: 1.8066... 0.0807 sec/batch
Epoch: 2/20... Training Step: 6076... Training loss: 1.7877... 0.0719 sec/batch
Epoch: 2/20... Training Step: 6077... Training loss: 1.7236... 0.0784 sec/batch
Epoch: 2/20... Training Step: 6078... Training loss: 1.8244... 0.0822 sec/batch
Epoch: 2/20... Training Step: 6079... Training loss: 1.7436... 0.0726 sec/batch
Epoch: 2/20... Training Step: 6080... Training loss: 1.7082... 0.0796 sec/batch
Epoch: 2/20... Training Step: 6081... Training loss: 1.6182... 0.0782 sec/batch
Epoch: 2/20... Training Step: 6082... Training loss: 1.7730... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6083... Training loss: 1.6582... 0.0692 sec/batch
Epoch: 2/20... Training Step: 6084... Training loss: 1.6760... 0.0698 sec/batch
Epoch: 2/20... Training Step: 6085... Training loss: 1.6617... 0.0791 sec/batch
Epoch: 2/20... Training Step: 6086... Training loss: 1.6394... 0.0668 sec/batch
Epoch: 2/20... Training Step: 6087... Training loss: 1.6956... 0.0693 sec/batch
Epoch: 2/20... Training Step: 6088... Training loss: 1.7023... 0.0713 sec/batch
Epoch: 2/20... Training Step: 6089... Training loss: 1.6632... 0.0828 sec/batch
Epoch: 2/20... Training Step: 6090... Training loss: 1.5620... 0.0660 sec/batch
Epoch: 2/20... Training Step: 6091... Training loss: 1.8068... 0.0731 sec/batch
Epoch: 2/20... Training Step: 6092... Training loss: 1.8492... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6093... Training loss: 1.6361... 0.0787 sec/batch
Epoch: 2/20... Training Step: 6094... Training loss: 1.7043... 0.0797 sec/batch
Epoch: 2/20... Training Step: 6095... Training loss: 1.7607... 0.0789 sec/batch
Epoch: 2/20... Training Step: 6096... Training loss: 1.7482... 0.0626 sec/batch
Epoch: 2/20... Training Step: 6097... Training loss: 1.7977... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6098... Training loss: 1.6804... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6099... Training loss: 1.6094... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6100... Training loss: 1.8476... 0.0826 sec/batch
Epoch: 2/20... Training Step: 6101... Training loss: 1.7027... 0.0669 sec/batch
Epoch: 2/20... Training Step: 6102... Training loss: 1.6837... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6103... Training loss: 1.7791... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6104... Training loss: 1.7153... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6105... Training loss: 1.7382... 0.0674 sec/batch
Epoch: 2/20... Training Step: 6106... Training loss: 1.6606... 0.0625 sec/batch
Epoch: 2/20... Training Step: 6107... Training loss: 1.6728... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6108... Training loss: 1.7974... 0.0635 sec/batch
Epoch: 2/20... Training Step: 6109... Training loss: 1.7873... 0.0792 sec/batch
Epoch: 2/20... Training Step: 6110... Training loss: 1.7456... 0.0581 sec/batch
Epoch: 2/20... Training Step: 6111... Training loss: 1.7260... 0.0614 sec/batch
Epoch: 2/20... Training Step: 6112... Training loss: 1.7156... 0.0775 sec/batch
Epoch: 2/20... Training Step: 6113... Training loss: 1.7629... 0.0717 sec/batch
Epoch: 2/20... Training Step: 6114... Training loss: 1.7039... 0.0779 sec/batch
Epoch: 2/20... Training Step: 6115... Training loss: 1.7543... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6116... Training loss: 1.7037... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6117... Training loss: 1.5982... 0.0713 sec/batch
Epoch: 2/20... Training Step: 6118... Training loss: 2.0616... 0.0705 sec/batch
Epoch: 2/20... Training Step: 6119... Training loss: 1.7770... 0.0746 sec/batch
Epoch: 2/20... Training Step: 6120... Training loss: 1.6865... 0.0721 sec/batch
Epoch: 2/20... Training Step: 6121... Training loss: 1.8958... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6122... Training loss: 1.7361... 0.0758 sec/batch
Epoch: 2/20... Training Step: 6123... Training loss: 1.6048... 0.0806 sec/batch
Epoch: 2/20... Training Step: 6124... Training loss: 1.7609... 0.0688 sec/batch
Epoch: 2/20... Training Step: 6125... Training loss: 1.8002... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6126... Training loss: 1.5593... 0.0645 sec/batch
Epoch: 2/20... Training Step: 6127... Training loss: 1.7820... 0.0819 sec/batch
Epoch: 2/20... Training Step: 6128... Training loss: 1.6795... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6129... Training loss: 1.8697... 0.0617 sec/batch
Epoch: 2/20... Training Step: 6130... Training loss: 1.7585... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6131... Training loss: 1.6945... 0.0847 sec/batch
Epoch: 2/20... Training Step: 6132... Training loss: 1.7079... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6133... Training loss: 1.6278... 0.0710 sec/batch
Epoch: 2/20... Training Step: 6134... Training loss: 1.6860... 0.0649 sec/batch
Epoch: 2/20... Training Step: 6135... Training loss: 1.7412... 0.0673 sec/batch
Epoch: 2/20... Training Step: 6136... Training loss: 1.6632... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6137... Training loss: 1.7503... 0.0725 sec/batch
Epoch: 2/20... Training Step: 6138... Training loss: 1.6121... 0.0621 sec/batch
Epoch: 2/20... Training Step: 6139... Training loss: 1.7728... 0.0641 sec/batch
Epoch: 2/20... Training Step: 6140... Training loss: 1.6639... 0.0782 sec/batch
Epoch: 2/20... Training Step: 6141... Training loss: 1.6581... 0.0736 sec/batch
Epoch: 2/20... Training Step: 6142... Training loss: 1.7345... 0.0643 sec/batch
Epoch: 2/20... Training Step: 6143... Training loss: 1.9144... 0.0751 sec/batch
Epoch: 2/20... Training Step: 6144... Training loss: 1.6380... 0.0716 sec/batch
Epoch: 2/20... Training Step: 6145... Training loss: 1.7516... 0.0592 sec/batch
Epoch: 2/20... Training Step: 6146... Training loss: 1.7674... 0.0757 sec/batch
Epoch: 2/20... Training Step: 6147... Training loss: 1.8089... 0.0777 sec/batch
Epoch: 2/20... Training Step: 6148... Training loss: 1.8489... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6149... Training loss: 1.7057... 0.0635 sec/batch
Epoch: 2/20... Training Step: 6150... Training loss: 1.7915... 0.0644 sec/batch
Epoch: 2/20... Training Step: 6151... Training loss: 1.7136... 0.0775 sec/batch
Epoch: 2/20... Training Step: 6152... Training loss: 1.9237... 0.0744 sec/batch
Epoch: 2/20... Training Step: 6153... Training loss: 1.6068... 0.0810 sec/batch
Epoch: 2/20... Training Step: 6154... Training loss: 1.8017... 0.0731 sec/batch
Epoch: 2/20... Training Step: 6155... Training loss: 1.8413... 0.0710 sec/batch
Epoch: 2/20... Training Step: 6156... Training loss: 1.7424... 0.0727 sec/batch
Epoch: 2/20... Training Step: 6157... Training loss: 1.8143... 0.0629 sec/batch
Epoch: 2/20... Training Step: 6158... Training loss: 1.6473... 0.0611 sec/batch
Epoch: 2/20... Training Step: 6159... Training loss: 1.7013... 0.0788 sec/batch
Epoch: 2/20... Training Step: 6160... Training loss: 1.7177... 0.0811 sec/batch
Epoch: 2/20... Training Step: 6161... Training loss: 1.7767... 0.0737 sec/batch
Epoch: 2/20... Training Step: 6162... Training loss: 1.7018... 0.0792 sec/batch
Epoch: 2/20... Training Step: 6163... Training loss: 1.6548... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6164... Training loss: 1.6403... 0.0699 sec/batch
Epoch: 2/20... Training Step: 6165... Training loss: 1.7183... 0.0611 sec/batch
Epoch: 2/20... Training Step: 6166... Training loss: 1.7697... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6167... Training loss: 1.7794... 0.0703 sec/batch
Epoch: 2/20... Training Step: 6168... Training loss: 1.7816... 0.0803 sec/batch
Epoch: 2/20... Training Step: 6169... Training loss: 1.6719... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6170... Training loss: 1.7525... 0.0783 sec/batch
Epoch: 2/20... Training Step: 6171... Training loss: 1.8700... 0.0775 sec/batch
Epoch: 2/20... Training Step: 6172... Training loss: 1.8493... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6173... Training loss: 1.8832... 0.0659 sec/batch
Epoch: 2/20... Training Step: 6174... Training loss: 1.6468... 0.0635 sec/batch
Epoch: 2/20... Training Step: 6175... Training loss: 1.8021... 0.0735 sec/batch
Epoch: 2/20... Training Step: 6176... Training loss: 1.7364... 0.0842 sec/batch
Epoch: 2/20... Training Step: 6177... Training loss: 1.7729... 0.0609 sec/batch
Epoch: 2/20... Training Step: 6178... Training loss: 1.7603... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6179... Training loss: 1.7474... 0.0641 sec/batch
Epoch: 2/20... Training Step: 6180... Training loss: 1.7177... 0.0818 sec/batch
Epoch: 2/20... Training Step: 6181... Training loss: 1.8101... 0.0859 sec/batch
Epoch: 2/20... Training Step: 6182... Training loss: 1.6038... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6183... Training loss: 1.6465... 0.0677 sec/batch
Epoch: 2/20... Training Step: 6184... Training loss: 1.8371... 0.0784 sec/batch
Epoch: 2/20... Training Step: 6185... Training loss: 1.7753... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6186... Training loss: 2.0574... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6187... Training loss: 1.7809... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6188... Training loss: 1.6977... 0.0742 sec/batch
Epoch: 2/20... Training Step: 6189... Training loss: 1.7262... 0.0794 sec/batch
Epoch: 2/20... Training Step: 6190... Training loss: 1.8036... 0.0749 sec/batch
Epoch: 2/20... Training Step: 6191... Training loss: 1.7532... 0.0848 sec/batch
Epoch: 2/20... Training Step: 6192... Training loss: 1.7472... 0.0867 sec/batch
Epoch: 2/20... Training Step: 6193... Training loss: 1.6270... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6194... Training loss: 1.6078... 0.0832 sec/batch
Epoch: 2/20... Training Step: 6195... Training loss: 1.8594... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6196... Training loss: 1.5883... 0.0787 sec/batch
Epoch: 2/20... Training Step: 6197... Training loss: 1.6568... 0.0694 sec/batch
Epoch: 2/20... Training Step: 6198... Training loss: 1.7543... 0.0662 sec/batch
Epoch: 2/20... Training Step: 6199... Training loss: 1.7909... 0.0912 sec/batch
Epoch: 2/20... Training Step: 6200... Training loss: 1.7394... 0.0791 sec/batch
Epoch: 2/20... Training Step: 6201... Training loss: 1.7799... 0.0890 sec/batch
Epoch: 2/20... Training Step: 6202... Training loss: 1.7000... 0.0743 sec/batch
Epoch: 2/20... Training Step: 6203... Training loss: 1.6863... 0.0607 sec/batch
Epoch: 2/20... Training Step: 6204... Training loss: 1.7121... 0.0578 sec/batch
Epoch: 2/20... Training Step: 6205... Training loss: 1.5816... 0.0925 sec/batch
Epoch: 2/20... Training Step: 6206... Training loss: 1.5782... 0.0824 sec/batch
Epoch: 2/20... Training Step: 6207... Training loss: 1.7357... 0.0825 sec/batch
Epoch: 2/20... Training Step: 6208... Training loss: 1.5955... 0.0858 sec/batch
Epoch: 2/20... Training Step: 6209... Training loss: 1.8174... 0.0799 sec/batch
Epoch: 2/20... Training Step: 6210... Training loss: 1.9992... 0.0829 sec/batch
Epoch: 2/20... Training Step: 6211... Training loss: 1.8352... 0.0860 sec/batch
Epoch: 2/20... Training Step: 6212... Training loss: 1.7776... 0.0960 sec/batch
Epoch: 2/20... Training Step: 6213... Training loss: 1.8103... 0.0708 sec/batch
Epoch: 2/20... Training Step: 6214... Training loss: 1.8024... 0.0778 sec/batch
Epoch: 2/20... Training Step: 6215... Training loss: 2.1098... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6216... Training loss: 1.9368... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6217... Training loss: 1.6388... 0.0577 sec/batch
Epoch: 2/20... Training Step: 6218... Training loss: 1.8379... 0.0747 sec/batch
Epoch: 2/20... Training Step: 6219... Training loss: 1.6129... 0.0652 sec/batch
Epoch: 2/20... Training Step: 6220... Training loss: 1.7372... 0.0706 sec/batch
Epoch: 2/20... Training Step: 6221... Training loss: 1.6164... 0.0611 sec/batch
Epoch: 2/20... Training Step: 6222... Training loss: 1.8136... 0.0741 sec/batch
Epoch: 2/20... Training Step: 6223... Training loss: 1.5932... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6224... Training loss: 1.6519... 0.0605 sec/batch
Epoch: 2/20... Training Step: 6225... Training loss: 1.7310... 0.0658 sec/batch
Epoch: 2/20... Training Step: 6226... Training loss: 1.6376... 0.0613 sec/batch
Epoch: 2/20... Training Step: 6227... Training loss: 1.7953... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6228... Training loss: 1.6908... 0.0618 sec/batch
Epoch: 2/20... Training Step: 6229... Training loss: 1.6427... 0.0582 sec/batch
Epoch: 2/20... Training Step: 6230... Training loss: 1.9536... 0.0702 sec/batch
Epoch: 2/20... Training Step: 6231... Training loss: 1.8970... 0.0625 sec/batch
Epoch: 2/20... Training Step: 6232... Training loss: 1.5326... 0.0670 sec/batch
Epoch: 2/20... Training Step: 6233... Training loss: 1.6418... 0.0860 sec/batch
Epoch: 2/20... Training Step: 6234... Training loss: 1.7828... 0.0754 sec/batch
Epoch: 2/20... Training Step: 6235... Training loss: 1.6377... 0.0584 sec/batch
Epoch: 2/20... Training Step: 6236... Training loss: 1.6446... 0.0583 sec/batch
Epoch: 2/20... Training Step: 6237... Training loss: 1.7543... 0.0676 sec/batch
Epoch: 2/20... Training Step: 6238... Training loss: 1.7961... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6239... Training loss: 1.7706... 0.0668 sec/batch
Epoch: 2/20... Training Step: 6240... Training loss: 1.6478... 0.0683 sec/batch
Epoch: 2/20... Training Step: 6241... Training loss: 1.7024... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6242... Training loss: 1.7588... 0.0730 sec/batch
Epoch: 2/20... Training Step: 6243... Training loss: 1.5869... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6244... Training loss: 1.6054... 0.0737 sec/batch
Epoch: 2/20... Training Step: 6245... Training loss: 1.5798... 0.0534 sec/batch
Epoch: 2/20... Training Step: 6246... Training loss: 1.6597... 0.0682 sec/batch
Epoch: 2/20... Training Step: 6247... Training loss: 1.6547... 0.0785 sec/batch
Epoch: 2/20... Training Step: 6248... Training loss: 1.7799... 0.0630 sec/batch
Epoch: 2/20... Training Step: 6249... Training loss: 1.8146... 0.0561 sec/batch
Epoch: 2/20... Training Step: 6250... Training loss: 1.6772... 0.0717 sec/batch
Epoch: 2/20... Training Step: 6251... Training loss: 1.7255... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6252... Training loss: 1.7882... 0.0816 sec/batch
Epoch: 2/20... Training Step: 6253... Training loss: 1.8115... 0.0744 sec/batch
Epoch: 2/20... Training Step: 6254... Training loss: 1.5660... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6255... Training loss: 1.6628... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6256... Training loss: 1.6053... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6257... Training loss: 1.6681... 0.0772 sec/batch
Epoch: 2/20... Training Step: 6258... Training loss: 1.6766... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6259... Training loss: 1.7039... 0.0797 sec/batch
Epoch: 2/20... Training Step: 6260... Training loss: 1.7434... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6261... Training loss: 1.6054... 0.0637 sec/batch
Epoch: 2/20... Training Step: 6262... Training loss: 1.7296... 0.0661 sec/batch
Epoch: 2/20... Training Step: 6263... Training loss: 1.6051... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6264... Training loss: 1.7316... 0.0627 sec/batch
Epoch: 2/20... Training Step: 6265... Training loss: 1.7133... 0.0708 sec/batch
Epoch: 2/20... Training Step: 6266... Training loss: 1.6586... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6267... Training loss: 1.7122... 0.0553 sec/batch
Epoch: 2/20... Training Step: 6268... Training loss: 1.5875... 0.0765 sec/batch
Epoch: 2/20... Training Step: 6269... Training loss: 1.8209... 0.0599 sec/batch
Epoch: 2/20... Training Step: 6270... Training loss: 1.5957... 0.0727 sec/batch
Epoch: 2/20... Training Step: 6271... Training loss: 1.7714... 0.0569 sec/batch
Epoch: 2/20... Training Step: 6272... Training loss: 1.8690... 0.0531 sec/batch
Epoch: 2/20... Training Step: 6273... Training loss: 1.7128... 0.0598 sec/batch
Epoch: 2/20... Training Step: 6274... Training loss: 1.6930... 0.0606 sec/batch
Epoch: 2/20... Training Step: 6275... Training loss: 1.6993... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6276... Training loss: 1.8103... 0.0724 sec/batch
Epoch: 2/20... Training Step: 6277... Training loss: 1.6915... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6278... Training loss: 1.7935... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6279... Training loss: 1.5854... 0.0658 sec/batch
Epoch: 2/20... Training Step: 6280... Training loss: 1.7872... 0.0757 sec/batch
Epoch: 2/20... Training Step: 6281... Training loss: 1.6488... 0.0718 sec/batch
Epoch: 2/20... Training Step: 6282... Training loss: 1.5740... 0.0658 sec/batch
Epoch: 2/20... Training Step: 6283... Training loss: 1.7213... 0.0682 sec/batch
Epoch: 2/20... Training Step: 6284... Training loss: 1.7448... 0.0668 sec/batch
Epoch: 2/20... Training Step: 6285... Training loss: 1.6413... 0.0653 sec/batch
Epoch: 2/20... Training Step: 6286... Training loss: 1.7241... 0.0846 sec/batch
Epoch: 2/20... Training Step: 6287... Training loss: 1.5331... 0.0649 sec/batch
Epoch: 2/20... Training Step: 6288... Training loss: 1.8086... 0.0690 sec/batch
Epoch: 2/20... Training Step: 6289... Training loss: 1.6580... 0.0843 sec/batch
Epoch: 2/20... Training Step: 6290... Training loss: 1.6821... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6291... Training loss: 1.6354... 0.0688 sec/batch
Epoch: 2/20... Training Step: 6292... Training loss: 1.6562... 0.0840 sec/batch
Epoch: 2/20... Training Step: 6293... Training loss: 1.5897... 0.0813 sec/batch
Epoch: 2/20... Training Step: 6294... Training loss: 1.8253... 0.0663 sec/batch
Epoch: 2/20... Training Step: 6295... Training loss: 1.9093... 0.0732 sec/batch
Epoch: 2/20... Training Step: 6296... Training loss: 1.7984... 0.0633 sec/batch
Epoch: 2/20... Training Step: 6297... Training loss: 1.7961... 0.0780 sec/batch
Epoch: 2/20... Training Step: 6298... Training loss: 1.7310... 0.0803 sec/batch
Epoch: 2/20... Training Step: 6299... Training loss: 1.7758... 0.0878 sec/batch
Epoch: 2/20... Training Step: 6300... Training loss: 1.6443... 0.0741 sec/batch
Epoch: 2/20... Training Step: 6301... Training loss: 1.5763... 0.0645 sec/batch
Epoch: 2/20... Training Step: 6302... Training loss: 1.7964... 0.0835 sec/batch
Epoch: 2/20... Training Step: 6303... Training loss: 1.6112... 0.0873 sec/batch
Epoch: 2/20... Training Step: 6304... Training loss: 1.6019... 0.0806 sec/batch
Epoch: 2/20... Training Step: 6305... Training loss: 1.7247... 0.0775 sec/batch
Epoch: 2/20... Training Step: 6306... Training loss: 1.7101... 0.0804 sec/batch
Epoch: 2/20... Training Step: 6307... Training loss: 1.9159... 0.0841 sec/batch
Epoch: 2/20... Training Step: 6308... Training loss: 1.9659... 0.0815 sec/batch
Epoch: 2/20... Training Step: 6309... Training loss: 1.6917... 0.0848 sec/batch
Epoch: 2/20... Training Step: 6310... Training loss: 1.8292... 0.0655 sec/batch
Epoch: 2/20... Training Step: 6311... Training loss: 1.7304... 0.0869 sec/batch
Epoch: 2/20... Training Step: 6312... Training loss: 1.6711... 0.0889 sec/batch
Epoch: 2/20... Training Step: 6313... Training loss: 1.7343... 0.0672 sec/batch
Epoch: 2/20... Training Step: 6314... Training loss: 1.6585... 0.0719 sec/batch
Epoch: 2/20... Training Step: 6315... Training loss: 1.7067... 0.0707 sec/batch
Epoch: 2/20... Training Step: 6316... Training loss: 1.5271... 0.0734 sec/batch
Epoch: 2/20... Training Step: 6317... Training loss: 1.8186... 0.0833 sec/batch
Epoch: 2/20... Training Step: 6318... Training loss: 1.6942... 0.0744 sec/batch
Epoch: 2/20... Training Step: 6319... Training loss: 1.7197... 0.0663 sec/batch
Epoch: 2/20... Training Step: 6320... Training loss: 1.8503... 0.0694 sec/batch
Epoch: 2/20... Training Step: 6321... Training loss: 1.9726... 0.0824 sec/batch
Epoch: 2/20... Training Step: 6322... Training loss: 1.7988... 0.0742 sec/batch
Epoch: 2/20... Training Step: 6323... Training loss: 1.8592... 0.0740 sec/batch
Epoch: 2/20... Training Step: 6324... Training loss: 1.7486... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6325... Training loss: 1.7331... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6326... Training loss: 1.7472... 0.0657 sec/batch
Epoch: 2/20... Training Step: 6327... Training loss: 1.7306... 0.0838 sec/batch
Epoch: 2/20... Training Step: 6328... Training loss: 1.7418... 0.0703 sec/batch
Epoch: 2/20... Training Step: 6329... Training loss: 1.6809... 0.0902 sec/batch
Epoch: 2/20... Training Step: 6330... Training loss: 1.8842... 0.1129 sec/batch
Epoch: 2/20... Training Step: 6331... Training loss: 1.6694... 0.1024 sec/batch
Epoch: 2/20... Training Step: 6332... Training loss: 1.8971... 0.0894 sec/batch
Epoch: 2/20... Training Step: 6333... Training loss: 1.6528... 0.1150 sec/batch
Epoch: 2/20... Training Step: 6334... Training loss: 1.8498... 0.1167 sec/batch
Epoch: 2/20... Training Step: 6335... Training loss: 1.6840... 0.0922 sec/batch
Epoch: 2/20... Training Step: 6336... Training loss: 1.6458... 0.0984 sec/batch
Epoch: 2/20... Training Step: 6337... Training loss: 1.6356... 0.0935 sec/batch
Epoch: 2/20... Training Step: 6338... Training loss: 1.6185... 0.1024 sec/batch
Epoch: 2/20... Training Step: 6339... Training loss: 1.7963... 0.1056 sec/batch
Epoch: 2/20... Training Step: 6340... Training loss: 1.6838... 0.1338 sec/batch
Epoch: 2/20... Training Step: 6341... Training loss: 1.8245... 0.1050 sec/batch
Epoch: 2/20... Training Step: 6342... Training loss: 1.7614... 0.0973 sec/batch
Epoch: 2/20... Training Step: 6343... Training loss: 1.5964... 0.1006 sec/batch
Epoch: 2/20... Training Step: 6344... Training loss: 1.7310... 0.1096 sec/batch
Epoch: 2/20... Training Step: 6345... Training loss: 1.7574... 0.1103 sec/batch
Epoch: 2/20... Training Step: 6346... Training loss: 1.8074... 0.0896 sec/batch
Epoch: 2/20... Training Step: 6347... Training loss: 1.8426... 0.1157 sec/batch
Epoch: 2/20... Training Step: 6348... Training loss: 1.7741... 0.0990 sec/batch
Epoch: 2/20... Training Step: 6349... Training loss: 1.7362... 0.0926 sec/batch
Epoch: 2/20... Training Step: 6350... Training loss: 1.8371... 0.1009 sec/batch
Epoch: 2/20... Training Step: 6351... Training loss: 1.6502... 0.0935 sec/batch
Epoch: 2/20... Training Step: 6352... Training loss: 1.8012... 0.0999 sec/batch
Epoch: 2/20... Training Step: 6353... Training loss: 1.7684... 0.0955 sec/batch
Epoch: 2/20... Training Step: 6354... Training loss: 1.7271... 0.1168 sec/batch
Epoch: 2/20... Training Step: 6355... Training loss: 1.8000... 0.1167 sec/batch
Epoch: 2/20... Training Step: 6356... Training loss: 1.6835... 0.0879 sec/batch
Epoch: 2/20... Training Step: 6357... Training loss: 1.6048... 0.1015 sec/batch
Epoch: 2/20... Training Step: 6358... Training loss: 1.8122... 0.0959 sec/batch
Epoch: 2/20... Training Step: 6359... Training loss: 1.8110... 0.0982 sec/batch
Epoch: 2/20... Training Step: 6360... Training loss: 1.7382... 0.1074 sec/batch
Epoch: 2/20... Training Step: 6361... Training loss: 1.7944... 0.1202 sec/batch
Epoch: 2/20... Training Step: 6362... Training loss: 1.7346... 0.0982 sec/batch
Epoch: 2/20... Training Step: 6363... Training loss: 1.8687... 0.0932 sec/batch
Epoch: 2/20... Training Step: 6364... Training loss: 1.8659... 0.1029 sec/batch
Epoch: 2/20... Training Step: 6365... Training loss: 1.7670... 0.1008 sec/batch
Epoch: 2/20... Training Step: 6366... Training loss: 1.7534... 0.0934 sec/batch
Epoch: 2/20... Training Step: 6367... Training loss: 1.7128... 0.0923 sec/batch
Epoch: 2/20... Training Step: 6368... Training loss: 1.6935... 0.0783 sec/batch
Epoch: 2/20... Training Step: 6369... Training loss: 1.7408... 0.0671 sec/batch
Epoch: 2/20... Training Step: 6370... Training loss: 1.8338... 0.0829 sec/batch
Epoch: 2/20... Training Step: 6371... Training loss: 1.7187... 0.0802 sec/batch
Epoch: 2/20... Training Step: 6372... Training loss: 1.6391... 0.0711 sec/batch
Epoch: 2/20... Training Step: 6373... Training loss: 1.6840... 0.0934 sec/batch
Epoch: 2/20... Training Step: 6374... Training loss: 1.7441... 0.0848 sec/batch
Epoch: 2/20... Training Step: 6375... Training loss: 1.5963... 0.0907 sec/batch
Epoch: 2/20... Training Step: 6376... Training loss: 1.8147... 0.0824 sec/batch
Epoch: 2/20... Training Step: 6377... Training loss: 1.6549... 0.0831 sec/batch
Epoch: 2/20... Training Step: 6378... Training loss: 1.6908... 0.0936 sec/batch
Epoch: 2/20... Training Step: 6379... Training loss: 1.5997... 0.0871 sec/batch
Epoch: 2/20... Training Step: 6380... Training loss: 1.6649... 0.0726 sec/batch
Epoch: 2/20... Training Step: 6381... Training loss: 1.7786... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6382... Training loss: 1.7515... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6383... Training loss: 1.7494... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6384... Training loss: 1.6718... 0.0874 sec/batch
Epoch: 2/20... Training Step: 6385... Training loss: 1.7107... 0.0784 sec/batch
Epoch: 2/20... Training Step: 6386... Training loss: 1.6393... 0.0979 sec/batch
Epoch: 2/20... Training Step: 6387... Training loss: 1.6701... 0.0843 sec/batch
Epoch: 2/20... Training Step: 6388... Training loss: 1.6580... 0.0828 sec/batch
Epoch: 2/20... Training Step: 6389... Training loss: 1.6425... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6390... Training loss: 1.9033... 0.0687 sec/batch
Epoch: 2/20... Training Step: 6391... Training loss: 1.8082... 0.0810 sec/batch
Epoch: 2/20... Training Step: 6392... Training loss: 1.7734... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6393... Training loss: 1.7030... 0.0929 sec/batch
Epoch: 2/20... Training Step: 6394... Training loss: 1.6906... 0.0911 sec/batch
Epoch: 2/20... Training Step: 6395... Training loss: 1.6364... 0.0904 sec/batch
Epoch: 2/20... Training Step: 6396... Training loss: 1.7417... 0.0801 sec/batch
Epoch: 2/20... Training Step: 6397... Training loss: 1.7057... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6398... Training loss: 1.8134... 0.0802 sec/batch
Epoch: 2/20... Training Step: 6399... Training loss: 1.8406... 0.1006 sec/batch
Epoch: 2/20... Training Step: 6400... Training loss: 1.6884... 0.0692 sec/batch
Epoch: 2/20... Training Step: 6401... Training loss: 1.7876... 0.0737 sec/batch
Epoch: 2/20... Training Step: 6402... Training loss: 1.8594... 0.0909 sec/batch
Epoch: 2/20... Training Step: 6403... Training loss: 1.8433... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6404... Training loss: 1.6772... 0.0904 sec/batch
Epoch: 2/20... Training Step: 6405... Training loss: 1.6307... 0.0872 sec/batch
Epoch: 2/20... Training Step: 6406... Training loss: 1.7452... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6407... Training loss: 1.6629... 0.0801 sec/batch
Epoch: 2/20... Training Step: 6408... Training loss: 1.7855... 0.0871 sec/batch
Epoch: 2/20... Training Step: 6409... Training loss: 1.8221... 0.0915 sec/batch
Epoch: 2/20... Training Step: 6410... Training loss: 1.7411... 0.0671 sec/batch
Epoch: 2/20... Training Step: 6411... Training loss: 1.7069... 0.0790 sec/batch
Epoch: 2/20... Training Step: 6412... Training loss: 1.7238... 0.0722 sec/batch
Epoch: 2/20... Training Step: 6413... Training loss: 1.7114... 0.0740 sec/batch
Epoch: 2/20... Training Step: 6414... Training loss: 1.8483... 0.0930 sec/batch
Epoch: 2/20... Training Step: 6415... Training loss: 1.6932... 0.0862 sec/batch
Epoch: 2/20... Training Step: 6416... Training loss: 1.7032... 0.0725 sec/batch
Epoch: 2/20... Training Step: 6417... Training loss: 1.8040... 0.0692 sec/batch
Epoch: 2/20... Training Step: 6418... Training loss: 1.6532... 0.0749 sec/batch
Epoch: 2/20... Training Step: 6419... Training loss: 1.7139... 0.0937 sec/batch
Epoch: 2/20... Training Step: 6420... Training loss: 1.8956... 0.0795 sec/batch
Epoch: 2/20... Training Step: 6421... Training loss: 1.9121... 0.0912 sec/batch
Epoch: 2/20... Training Step: 6422... Training loss: 1.7309... 0.0673 sec/batch
Epoch: 2/20... Training Step: 6423... Training loss: 1.7948... 0.0643 sec/batch
Epoch: 2/20... Training Step: 6424... Training loss: 1.6867... 0.0662 sec/batch
Epoch: 2/20... Training Step: 6425... Training loss: 1.6637... 0.0797 sec/batch
Epoch: 2/20... Training Step: 6426... Training loss: 1.7433... 0.0874 sec/batch
Epoch: 2/20... Training Step: 6427... Training loss: 1.8998... 0.0602 sec/batch
Epoch: 2/20... Training Step: 6428... Training loss: 1.7744... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6429... Training loss: 1.9298... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6430... Training loss: 1.7888... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6431... Training loss: 1.5987... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6432... Training loss: 1.7445... 0.0708 sec/batch
Epoch: 2/20... Training Step: 6433... Training loss: 1.5949... 0.0881 sec/batch
Epoch: 2/20... Training Step: 6434... Training loss: 1.5890... 0.0815 sec/batch
Epoch: 2/20... Training Step: 6435... Training loss: 1.8124... 0.0886 sec/batch
Epoch: 2/20... Training Step: 6436... Training loss: 1.6744... 0.0802 sec/batch
Epoch: 2/20... Training Step: 6437... Training loss: 1.7982... 0.0851 sec/batch
Epoch: 2/20... Training Step: 6438... Training loss: 1.7135... 0.0782 sec/batch
Epoch: 2/20... Training Step: 6439... Training loss: 1.6986... 0.0687 sec/batch
Epoch: 2/20... Training Step: 6440... Training loss: 1.7657... 0.0744 sec/batch
Epoch: 2/20... Training Step: 6441... Training loss: 1.6916... 0.0727 sec/batch
Epoch: 2/20... Training Step: 6442... Training loss: 1.7472... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6443... Training loss: 1.6449... 0.0893 sec/batch
Epoch: 2/20... Training Step: 6444... Training loss: 1.6772... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6445... Training loss: 1.5342... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6446... Training loss: 1.6058... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6447... Training loss: 1.6023... 0.0687 sec/batch
Epoch: 2/20... Training Step: 6448... Training loss: 1.6878... 0.0736 sec/batch
Epoch: 2/20... Training Step: 6449... Training loss: 1.7210... 0.0688 sec/batch
Epoch: 2/20... Training Step: 6450... Training loss: 1.7573... 0.0778 sec/batch
Epoch: 2/20... Training Step: 6451... Training loss: 1.6858... 0.0609 sec/batch
Epoch: 2/20... Training Step: 6452... Training loss: 1.7127... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6453... Training loss: 1.8935... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6454... Training loss: 1.6687... 0.0651 sec/batch
Epoch: 2/20... Training Step: 6455... Training loss: 1.7237... 0.0851 sec/batch
Epoch: 2/20... Training Step: 6456... Training loss: 1.7043... 0.0813 sec/batch
Epoch: 2/20... Training Step: 6457... Training loss: 1.6864... 0.0634 sec/batch
Epoch: 2/20... Training Step: 6458... Training loss: 1.6420... 0.0693 sec/batch
Epoch: 2/20... Training Step: 6459... Training loss: 1.8263... 0.0649 sec/batch
Epoch: 2/20... Training Step: 6460... Training loss: 1.7772... 0.0678 sec/batch
Epoch: 2/20... Training Step: 6461... Training loss: 1.8013... 0.0845 sec/batch
Epoch: 2/20... Training Step: 6462... Training loss: 1.8024... 0.0823 sec/batch
Epoch: 2/20... Training Step: 6463... Training loss: 1.7472... 0.0666 sec/batch
Epoch: 2/20... Training Step: 6464... Training loss: 1.7033... 0.0726 sec/batch
Epoch: 2/20... Training Step: 6465... Training loss: 1.7607... 0.0793 sec/batch
Epoch: 2/20... Training Step: 6466... Training loss: 1.7468... 0.0882 sec/batch
Epoch: 2/20... Training Step: 6467... Training loss: 1.7070... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6468... Training loss: 1.6792... 0.0768 sec/batch
Epoch: 2/20... Training Step: 6469... Training loss: 1.7847... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6470... Training loss: 1.6698... 0.0592 sec/batch
Epoch: 2/20... Training Step: 6471... Training loss: 1.6323... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6472... Training loss: 1.6825... 0.0635 sec/batch
Epoch: 2/20... Training Step: 6473... Training loss: 1.6069... 0.0573 sec/batch
Epoch: 2/20... Training Step: 6474... Training loss: 1.6429... 0.0661 sec/batch
Epoch: 2/20... Training Step: 6475... Training loss: 1.7793... 0.0551 sec/batch
Epoch: 2/20... Training Step: 6476... Training loss: 1.7583... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6477... Training loss: 1.6255... 0.0696 sec/batch
Epoch: 2/20... Training Step: 6478... Training loss: 1.7607... 0.0731 sec/batch
Epoch: 2/20... Training Step: 6479... Training loss: 1.7728... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6480... Training loss: 1.8824... 0.0607 sec/batch
Epoch: 2/20... Training Step: 6481... Training loss: 1.6731... 0.0666 sec/batch
Epoch: 2/20... Training Step: 6482... Training loss: 1.6341... 0.0627 sec/batch
Epoch: 2/20... Training Step: 6483... Training loss: 1.6593... 0.0698 sec/batch
Epoch: 2/20... Training Step: 6484... Training loss: 1.7932... 0.0641 sec/batch
Epoch: 2/20... Training Step: 6485... Training loss: 1.5823... 0.0800 sec/batch
Epoch: 2/20... Training Step: 6486... Training loss: 1.7799... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6487... Training loss: 1.7081... 0.0653 sec/batch
Epoch: 2/20... Training Step: 6488... Training loss: 1.7642... 0.0779 sec/batch
Epoch: 2/20... Training Step: 6489... Training loss: 1.7965... 0.0542 sec/batch
Epoch: 2/20... Training Step: 6490... Training loss: 1.7939... 0.0591 sec/batch
Epoch: 2/20... Training Step: 6491... Training loss: 1.6711... 0.0708 sec/batch
Epoch: 2/20... Training Step: 6492... Training loss: 1.7186... 0.0792 sec/batch
Epoch: 2/20... Training Step: 6493... Training loss: 1.6610... 0.0792 sec/batch
Epoch: 2/20... Training Step: 6494... Training loss: 1.8149... 0.0753 sec/batch
Epoch: 2/20... Training Step: 6495... Training loss: 1.8151... 0.0699 sec/batch
Epoch: 2/20... Training Step: 6496... Training loss: 1.6647... 0.0627 sec/batch
Epoch: 2/20... Training Step: 6497... Training loss: 1.8568... 0.0618 sec/batch
Epoch: 2/20... Training Step: 6498... Training loss: 1.6437... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6499... Training loss: 1.7733... 0.0657 sec/batch
Epoch: 2/20... Training Step: 6500... Training loss: 1.6993... 0.0584 sec/batch
Epoch: 2/20... Training Step: 6501... Training loss: 1.6723... 0.0682 sec/batch
Epoch: 2/20... Training Step: 6502... Training loss: 1.6565... 0.0927 sec/batch
Epoch: 2/20... Training Step: 6503... Training loss: 1.7750... 0.0557 sec/batch
Epoch: 2/20... Training Step: 6504... Training loss: 1.5828... 0.0629 sec/batch
Epoch: 2/20... Training Step: 6505... Training loss: 1.7635... 0.0579 sec/batch
Epoch: 2/20... Training Step: 6506... Training loss: 1.7357... 0.0676 sec/batch
Epoch: 2/20... Training Step: 6507... Training loss: 1.7076... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6508... Training loss: 1.7789... 0.0646 sec/batch
Epoch: 2/20... Training Step: 6509... Training loss: 1.8331... 0.0699 sec/batch
Epoch: 2/20... Training Step: 6510... Training loss: 1.8833... 0.0707 sec/batch
Epoch: 2/20... Training Step: 6511... Training loss: 1.7578... 0.0565 sec/batch
Epoch: 2/20... Training Step: 6512... Training loss: 1.8026... 0.0661 sec/batch
Epoch: 2/20... Training Step: 6513... Training loss: 1.6383... 0.0757 sec/batch
Epoch: 2/20... Training Step: 6514... Training loss: 1.7559... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6515... Training loss: 1.7047... 0.0807 sec/batch
Epoch: 2/20... Training Step: 6516... Training loss: 1.6835... 0.0802 sec/batch
Epoch: 2/20... Training Step: 6517... Training loss: 1.7673... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6518... Training loss: 1.6573... 0.0698 sec/batch
Epoch: 2/20... Training Step: 6519... Training loss: 1.7232... 0.0678 sec/batch
Epoch: 2/20... Training Step: 6520... Training loss: 1.7396... 0.0810 sec/batch
Epoch: 2/20... Training Step: 6521... Training loss: 1.6126... 0.0687 sec/batch
Epoch: 2/20... Training Step: 6522... Training loss: 1.8490... 0.0818 sec/batch
Epoch: 2/20... Training Step: 6523... Training loss: 1.5740... 0.0758 sec/batch
Epoch: 2/20... Training Step: 6524... Training loss: 1.6245... 0.0771 sec/batch
Epoch: 2/20... Training Step: 6525... Training loss: 1.6733... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6526... Training loss: 1.7945... 0.0787 sec/batch
Epoch: 2/20... Training Step: 6527... Training loss: 1.7527... 0.0811 sec/batch
Epoch: 2/20... Training Step: 6528... Training loss: 1.7564... 0.0635 sec/batch
Epoch: 2/20... Training Step: 6529... Training loss: 1.7159... 0.0816 sec/batch
Epoch: 2/20... Training Step: 6530... Training loss: 1.6582... 0.0633 sec/batch
Epoch: 2/20... Training Step: 6531... Training loss: 1.5531... 0.0819 sec/batch
Epoch: 2/20... Training Step: 6532... Training loss: 1.6550... 0.0746 sec/batch
Epoch: 2/20... Training Step: 6533... Training loss: 1.6531... 0.0587 sec/batch
Epoch: 2/20... Training Step: 6534... Training loss: 1.5688... 0.0633 sec/batch
Epoch: 2/20... Training Step: 6535... Training loss: 1.5319... 0.0934 sec/batch
Epoch: 2/20... Training Step: 6536... Training loss: 1.7386... 0.0813 sec/batch
Epoch: 2/20... Training Step: 6537... Training loss: 1.6759... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6538... Training loss: 1.7398... 0.0745 sec/batch
Epoch: 2/20... Training Step: 6539... Training loss: 1.6453... 0.0573 sec/batch
Epoch: 2/20... Training Step: 6540... Training loss: 1.6789... 0.0795 sec/batch
Epoch: 2/20... Training Step: 6541... Training loss: 1.6245... 0.0911 sec/batch
Epoch: 2/20... Training Step: 6542... Training loss: 1.7641... 0.0737 sec/batch
Epoch: 2/20... Training Step: 6543... Training loss: 1.7389... 0.0797 sec/batch
Epoch: 2/20... Training Step: 6544... Training loss: 1.7207... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6545... Training loss: 1.6775... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6546... Training loss: 1.6710... 0.0816 sec/batch
Epoch: 2/20... Training Step: 6547... Training loss: 1.7951... 0.0826 sec/batch
Epoch: 2/20... Training Step: 6548... Training loss: 1.7830... 0.0812 sec/batch
Epoch: 2/20... Training Step: 6549... Training loss: 1.7207... 0.0793 sec/batch
Epoch: 2/20... Training Step: 6550... Training loss: 1.7728... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6551... Training loss: 1.6226... 0.0686 sec/batch
Epoch: 2/20... Training Step: 6552... Training loss: 1.8496... 0.0604 sec/batch
Epoch: 2/20... Training Step: 6553... Training loss: 1.6145... 0.0617 sec/batch
Epoch: 2/20... Training Step: 6554... Training loss: 1.6540... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6555... Training loss: 1.6868... 0.0828 sec/batch
Epoch: 2/20... Training Step: 6556... Training loss: 1.9196... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6557... Training loss: 1.7438... 0.0620 sec/batch
Epoch: 2/20... Training Step: 6558... Training loss: 1.9548... 0.0774 sec/batch
Epoch: 2/20... Training Step: 6559... Training loss: 1.8145... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6560... Training loss: 1.7799... 0.0735 sec/batch
Epoch: 2/20... Training Step: 6561... Training loss: 1.7329... 0.0885 sec/batch
Epoch: 2/20... Training Step: 6562... Training loss: 1.6727... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6563... Training loss: 1.7621... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6564... Training loss: 1.7256... 0.0707 sec/batch
Epoch: 2/20... Training Step: 6565... Training loss: 1.6601... 0.0706 sec/batch
Epoch: 2/20... Training Step: 6566... Training loss: 1.7228... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6567... Training loss: 1.6057... 0.0705 sec/batch
Epoch: 2/20... Training Step: 6568... Training loss: 1.6683... 0.0833 sec/batch
Epoch: 2/20... Training Step: 6569... Training loss: 1.6909... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6570... Training loss: 1.6741... 0.0731 sec/batch
Epoch: 2/20... Training Step: 6571... Training loss: 1.6946... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6572... Training loss: 1.6537... 0.0734 sec/batch
Epoch: 2/20... Training Step: 6573... Training loss: 1.7978... 0.0774 sec/batch
Epoch: 2/20... Training Step: 6574... Training loss: 1.8021... 0.0944 sec/batch
Epoch: 2/20... Training Step: 6575... Training loss: 1.6699... 0.0804 sec/batch
Epoch: 2/20... Training Step: 6576... Training loss: 1.6324... 0.0598 sec/batch
Epoch: 2/20... Training Step: 6577... Training loss: 1.7255... 0.0787 sec/batch
Epoch: 2/20... Training Step: 6578... Training loss: 1.7775... 0.0742 sec/batch
Epoch: 2/20... Training Step: 6579... Training loss: 1.5914... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6580... Training loss: 1.8152... 0.0638 sec/batch
Epoch: 2/20... Training Step: 6581... Training loss: 1.8112... 0.0983 sec/batch
Epoch: 2/20... Training Step: 6582... Training loss: 1.6981... 0.1263 sec/batch
Epoch: 2/20... Training Step: 6583... Training loss: 1.6851... 0.0955 sec/batch
Epoch: 2/20... Training Step: 6584... Training loss: 1.7099... 0.0998 sec/batch
Epoch: 2/20... Training Step: 6585... Training loss: 1.5018... 0.1307 sec/batch
Epoch: 2/20... Training Step: 6586... Training loss: 1.6078... 0.1141 sec/batch
Epoch: 2/20... Training Step: 6587... Training loss: 1.7058... 0.0970 sec/batch
Epoch: 2/20... Training Step: 6588... Training loss: 1.7283... 0.1098 sec/batch
Epoch: 2/20... Training Step: 6589... Training loss: 1.5594... 0.1246 sec/batch
Epoch: 2/20... Training Step: 6590... Training loss: 1.5854... 0.1150 sec/batch
Epoch: 2/20... Training Step: 6591... Training loss: 1.8874... 0.1172 sec/batch
Epoch: 2/20... Training Step: 6592... Training loss: 1.5602... 0.1079 sec/batch
Epoch: 2/20... Training Step: 6593... Training loss: 1.7565... 0.1049 sec/batch
Epoch: 2/20... Training Step: 6594... Training loss: 1.8551... 0.1014 sec/batch
Epoch: 2/20... Training Step: 6595... Training loss: 1.7294... 0.1305 sec/batch
Epoch: 2/20... Training Step: 6596... Training loss: 1.7594... 0.1083 sec/batch
Epoch: 2/20... Training Step: 6597... Training loss: 1.7226... 0.0930 sec/batch
Epoch: 2/20... Training Step: 6598... Training loss: 1.7405... 0.1110 sec/batch
Epoch: 2/20... Training Step: 6599... Training loss: 1.6239... 0.1196 sec/batch
Epoch: 2/20... Training Step: 6600... Training loss: 1.9334... 0.1096 sec/batch
Epoch: 2/20... Training Step: 6601... Training loss: 1.7898... 0.1241 sec/batch
Epoch: 2/20... Training Step: 6602... Training loss: 1.6992... 0.1007 sec/batch
Epoch: 2/20... Training Step: 6603... Training loss: 1.7470... 0.1040 sec/batch
Epoch: 2/20... Training Step: 6604... Training loss: 1.7467... 0.1259 sec/batch
Epoch: 2/20... Training Step: 6605... Training loss: 1.5930... 0.0982 sec/batch
Epoch: 2/20... Training Step: 6606... Training loss: 1.7851... 0.0968 sec/batch
Epoch: 2/20... Training Step: 6607... Training loss: 1.7906... 0.1182 sec/batch
Epoch: 2/20... Training Step: 6608... Training loss: 1.6733... 0.0840 sec/batch
Epoch: 2/20... Training Step: 6609... Training loss: 1.7344... 0.1085 sec/batch
Epoch: 2/20... Training Step: 6610... Training loss: 1.6338... 0.1097 sec/batch
Epoch: 2/20... Training Step: 6611... Training loss: 1.7612... 0.0950 sec/batch
Epoch: 2/20... Training Step: 6612... Training loss: 1.7394... 0.1125 sec/batch
Epoch: 2/20... Training Step: 6613... Training loss: 1.6447... 0.1079 sec/batch
Epoch: 2/20... Training Step: 6614... Training loss: 1.6939... 0.0968 sec/batch
Epoch: 2/20... Training Step: 6615... Training loss: 1.7577... 0.0861 sec/batch
Epoch: 2/20... Training Step: 6616... Training loss: 1.7407... 0.0885 sec/batch
Epoch: 2/20... Training Step: 6617... Training loss: 1.7385... 0.0668 sec/batch
Epoch: 2/20... Training Step: 6618... Training loss: 1.7723... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6619... Training loss: 1.6906... 0.0781 sec/batch
Epoch: 2/20... Training Step: 6620... Training loss: 1.8515... 0.0845 sec/batch
Epoch: 2/20... Training Step: 6621... Training loss: 1.8222... 0.0893 sec/batch
Epoch: 2/20... Training Step: 6622... Training loss: 1.7789... 0.0783 sec/batch
Epoch: 2/20... Training Step: 6623... Training loss: 1.7753... 0.0781 sec/batch
Epoch: 2/20... Training Step: 6624... Training loss: 1.8273... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6625... Training loss: 1.7826... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6626... Training loss: 1.7692... 0.0845 sec/batch
Epoch: 2/20... Training Step: 6627... Training loss: 1.7804... 0.0842 sec/batch
Epoch: 2/20... Training Step: 6628... Training loss: 1.6787... 0.0658 sec/batch
Epoch: 2/20... Training Step: 6629... Training loss: 1.6533... 0.0653 sec/batch
Epoch: 2/20... Training Step: 6630... Training loss: 1.8980... 0.0691 sec/batch
Epoch: 2/20... Training Step: 6631... Training loss: 1.6032... 0.0698 sec/batch
Epoch: 2/20... Training Step: 6632... Training loss: 1.6690... 0.0844 sec/batch
Epoch: 2/20... Training Step: 6633... Training loss: 1.6342... 0.0915 sec/batch
Epoch: 2/20... Training Step: 6634... Training loss: 1.6994... 0.0830 sec/batch
Epoch: 2/20... Training Step: 6635... Training loss: 1.8887... 0.0694 sec/batch
Epoch: 2/20... Training Step: 6636... Training loss: 1.7585... 0.0846 sec/batch
Epoch: 2/20... Training Step: 6637... Training loss: 1.6496... 0.0965 sec/batch
Epoch: 2/20... Training Step: 6638... Training loss: 1.5956... 0.0739 sec/batch
Epoch: 2/20... Training Step: 6639... Training loss: 1.8405... 0.0821 sec/batch
Epoch: 2/20... Training Step: 6640... Training loss: 1.7116... 0.0772 sec/batch
Epoch: 2/20... Training Step: 6641... Training loss: 2.0186... 0.0918 sec/batch
Epoch: 2/20... Training Step: 6642... Training loss: 1.7127... 0.0855 sec/batch
Epoch: 2/20... Training Step: 6643... Training loss: 1.8292... 0.0918 sec/batch
Epoch: 2/20... Training Step: 6644... Training loss: 1.8578... 0.0840 sec/batch
Epoch: 2/20... Training Step: 6645... Training loss: 1.6891... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6646... Training loss: 1.7350... 0.0873 sec/batch
Epoch: 2/20... Training Step: 6647... Training loss: 1.6882... 0.0853 sec/batch
Epoch: 2/20... Training Step: 6648... Training loss: 1.6890... 0.0924 sec/batch
Epoch: 2/20... Training Step: 6649... Training loss: 1.8041... 0.0741 sec/batch
Epoch: 2/20... Training Step: 6650... Training loss: 1.5947... 0.0743 sec/batch
Epoch: 2/20... Training Step: 6651... Training loss: 1.6849... 0.0821 sec/batch
Epoch: 2/20... Training Step: 6652... Training loss: 1.6997... 0.0721 sec/batch
Epoch: 2/20... Training Step: 6653... Training loss: 1.8172... 0.0903 sec/batch
Epoch: 2/20... Training Step: 6654... Training loss: 1.6299... 0.0864 sec/batch
Epoch: 2/20... Training Step: 6655... Training loss: 1.6613... 0.0886 sec/batch
Epoch: 2/20... Training Step: 6656... Training loss: 1.6403... 0.0795 sec/batch
Epoch: 2/20... Training Step: 6657... Training loss: 1.6266... 0.0707 sec/batch
Epoch: 2/20... Training Step: 6658... Training loss: 1.6558... 0.0710 sec/batch
Epoch: 2/20... Training Step: 6659... Training loss: 1.6607... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6660... Training loss: 1.7713... 0.0825 sec/batch
Epoch: 2/20... Training Step: 6661... Training loss: 1.7592... 0.0798 sec/batch
Epoch: 2/20... Training Step: 6662... Training loss: 1.7310... 0.0833 sec/batch
Epoch: 2/20... Training Step: 6663... Training loss: 1.5877... 0.0881 sec/batch
Epoch: 2/20... Training Step: 6664... Training loss: 1.8273... 0.0786 sec/batch
Epoch: 2/20... Training Step: 6665... Training loss: 1.9548... 0.0725 sec/batch
Epoch: 2/20... Training Step: 6666... Training loss: 1.7217... 0.0886 sec/batch
Epoch: 2/20... Training Step: 6667... Training loss: 1.6954... 0.0795 sec/batch
Epoch: 2/20... Training Step: 6668... Training loss: 1.7056... 0.0711 sec/batch
Epoch: 2/20... Training Step: 6669... Training loss: 1.9343... 0.0796 sec/batch
Epoch: 2/20... Training Step: 6670... Training loss: 1.8286... 0.0683 sec/batch
Epoch: 2/20... Training Step: 6671... Training loss: 1.7703... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6672... Training loss: 1.7893... 0.0753 sec/batch
Epoch: 2/20... Training Step: 6673... Training loss: 1.8700... 0.0831 sec/batch
Epoch: 2/20... Training Step: 6674... Training loss: 1.8895... 0.0781 sec/batch
Epoch: 2/20... Training Step: 6675... Training loss: 1.8746... 0.0799 sec/batch
Epoch: 2/20... Training Step: 6676... Training loss: 1.7711... 0.0761 sec/batch
Epoch: 2/20... Training Step: 6677... Training loss: 1.6402... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6678... Training loss: 1.7197... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6679... Training loss: 1.6819... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6680... Training loss: 1.7007... 0.0704 sec/batch
Epoch: 2/20... Training Step: 6681... Training loss: 1.5363... 0.0718 sec/batch
Epoch: 2/20... Training Step: 6682... Training loss: 1.8056... 0.0950 sec/batch
Epoch: 2/20... Training Step: 6683... Training loss: 1.7202... 0.0830 sec/batch
Epoch: 2/20... Training Step: 6684... Training loss: 1.7628... 0.0645 sec/batch
Epoch: 2/20... Training Step: 6685... Training loss: 1.6577... 0.0776 sec/batch
Epoch: 2/20... Training Step: 6686... Training loss: 1.7149... 0.0756 sec/batch
Epoch: 2/20... Training Step: 6687... Training loss: 1.5741... 0.0704 sec/batch
Epoch: 2/20... Training Step: 6688... Training loss: 1.6623... 0.0550 sec/batch
Epoch: 2/20... Training Step: 6689... Training loss: 1.7172... 0.0750 sec/batch
Epoch: 2/20... Training Step: 6690... Training loss: 1.5760... 0.0621 sec/batch
Epoch: 2/20... Training Step: 6691... Training loss: 1.7149... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6692... Training loss: 1.6444... 0.0594 sec/batch
Epoch: 2/20... Training Step: 6693... Training loss: 1.6883... 0.0740 sec/batch
Epoch: 2/20... Training Step: 6694... Training loss: 1.7380... 0.0808 sec/batch
Epoch: 2/20... Training Step: 6695... Training loss: 1.7725... 0.0823 sec/batch
Epoch: 2/20... Training Step: 6696... Training loss: 1.8535... 0.0759 sec/batch
Epoch: 2/20... Training Step: 6697... Training loss: 1.7454... 0.0739 sec/batch
Epoch: 2/20... Training Step: 6698... Training loss: 1.8511... 0.0679 sec/batch
Epoch: 2/20... Training Step: 6699... Training loss: 1.7345... 0.0691 sec/batch
Epoch: 2/20... Training Step: 6700... Training loss: 1.6860... 0.0688 sec/batch
Epoch: 2/20... Training Step: 6701... Training loss: 1.7426... 0.0674 sec/batch
Epoch: 2/20... Training Step: 6702... Training loss: 1.6374... 0.0617 sec/batch
Epoch: 2/20... Training Step: 6703... Training loss: 1.7658... 0.0725 sec/batch
Epoch: 2/20... Training Step: 6704... Training loss: 1.7193... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6705... Training loss: 1.6612... 0.0672 sec/batch
Epoch: 2/20... Training Step: 6706... Training loss: 1.8200... 0.0706 sec/batch
Epoch: 2/20... Training Step: 6707... Training loss: 1.6736... 0.0590 sec/batch
Epoch: 2/20... Training Step: 6708... Training loss: 1.7279... 0.0643 sec/batch
Epoch: 2/20... Training Step: 6709... Training loss: 1.8703... 0.0679 sec/batch
Epoch: 2/20... Training Step: 6710... Training loss: 1.6654... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6711... Training loss: 1.7380... 0.0823 sec/batch
Epoch: 2/20... Training Step: 6712... Training loss: 1.7915... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6713... Training loss: 1.8824... 0.0643 sec/batch
Epoch: 2/20... Training Step: 6714... Training loss: 1.7872... 0.0694 sec/batch
Epoch: 2/20... Training Step: 6715... Training loss: 1.7531... 0.0696 sec/batch
Epoch: 2/20... Training Step: 6716... Training loss: 1.8856... 0.0814 sec/batch
Epoch: 2/20... Training Step: 6717... Training loss: 1.7153... 0.0710 sec/batch
Epoch: 2/20... Training Step: 6718... Training loss: 1.7976... 0.0655 sec/batch
Epoch: 2/20... Training Step: 6719... Training loss: 1.8255... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6720... Training loss: 1.7018... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6721... Training loss: 1.6783... 0.0776 sec/batch
Epoch: 2/20... Training Step: 6722... Training loss: 1.7670... 0.0636 sec/batch
Epoch: 2/20... Training Step: 6723... Training loss: 1.6751... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6724... Training loss: 1.7053... 0.0530 sec/batch
Epoch: 2/20... Training Step: 6725... Training loss: 1.8750... 0.0624 sec/batch
Epoch: 2/20... Training Step: 6726... Training loss: 1.7542... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6727... Training loss: 1.8065... 0.0577 sec/batch
Epoch: 2/20... Training Step: 6728... Training loss: 1.7260... 0.0569 sec/batch
Epoch: 2/20... Training Step: 6729... Training loss: 1.7376... 0.0574 sec/batch
Epoch: 2/20... Training Step: 6730... Training loss: 1.7820... 0.0632 sec/batch
Epoch: 2/20... Training Step: 6731... Training loss: 1.5795... 0.0606 sec/batch
Epoch: 2/20... Training Step: 6732... Training loss: 1.6856... 0.0807 sec/batch
Epoch: 2/20... Training Step: 6733... Training loss: 1.7437... 0.0693 sec/batch
Epoch: 2/20... Training Step: 6734... Training loss: 1.6042... 0.0755 sec/batch
Epoch: 2/20... Training Step: 6735... Training loss: 1.8690... 0.0644 sec/batch
Epoch: 2/20... Training Step: 6736... Training loss: 1.6616... 0.0716 sec/batch
Epoch: 2/20... Training Step: 6737... Training loss: 1.5743... 0.0798 sec/batch
Epoch: 2/20... Training Step: 6738... Training loss: 1.7591... 0.0730 sec/batch
Epoch: 2/20... Training Step: 6739... Training loss: 1.9008... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6740... Training loss: 1.8409... 0.0732 sec/batch
Epoch: 2/20... Training Step: 6741... Training loss: 1.7883... 0.0777 sec/batch
Epoch: 2/20... Training Step: 6742... Training loss: 1.8412... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6743... Training loss: 1.7244... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6744... Training loss: 1.7861... 0.0651 sec/batch
Epoch: 2/20... Training Step: 6745... Training loss: 1.7131... 0.0751 sec/batch
Epoch: 2/20... Training Step: 6746... Training loss: 1.6706... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6747... Training loss: 1.7584... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6748... Training loss: 1.6188... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6749... Training loss: 1.5751... 0.0769 sec/batch
Epoch: 2/20... Training Step: 6750... Training loss: 1.6686... 0.0806 sec/batch
Epoch: 2/20... Training Step: 6751... Training loss: 1.6839... 0.0665 sec/batch
Epoch: 2/20... Training Step: 6752... Training loss: 1.8667... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6753... Training loss: 1.6491... 0.0616 sec/batch
Epoch: 2/20... Training Step: 6754... Training loss: 1.6255... 0.0613 sec/batch
Epoch: 2/20... Training Step: 6755... Training loss: 1.6722... 0.0721 sec/batch
Epoch: 2/20... Training Step: 6756... Training loss: 1.8099... 0.0724 sec/batch
Epoch: 2/20... Training Step: 6757... Training loss: 1.6311... 0.0603 sec/batch
Epoch: 2/20... Training Step: 6758... Training loss: 1.6882... 0.0649 sec/batch
Epoch: 2/20... Training Step: 6759... Training loss: 1.6370... 0.0755 sec/batch
Epoch: 2/20... Training Step: 6760... Training loss: 1.6889... 0.0713 sec/batch
Epoch: 2/20... Training Step: 6761... Training loss: 1.6711... 0.0590 sec/batch
Epoch: 2/20... Training Step: 6762... Training loss: 1.6321... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6763... Training loss: 1.5415... 0.0608 sec/batch
Epoch: 2/20... Training Step: 6764... Training loss: 1.6116... 0.0609 sec/batch
Epoch: 2/20... Training Step: 6765... Training loss: 1.6735... 0.0651 sec/batch
Epoch: 2/20... Training Step: 6766... Training loss: 1.8289... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6767... Training loss: 1.6258... 0.0679 sec/batch
Epoch: 2/20... Training Step: 6768... Training loss: 1.6923... 0.0755 sec/batch
Epoch: 2/20... Training Step: 6769... Training loss: 1.6828... 0.0749 sec/batch
Epoch: 2/20... Training Step: 6770... Training loss: 1.5860... 0.0683 sec/batch
Epoch: 2/20... Training Step: 6771... Training loss: 1.7287... 0.0880 sec/batch
Epoch: 2/20... Training Step: 6772... Training loss: 1.7285... 0.0742 sec/batch
Epoch: 2/20... Training Step: 6773... Training loss: 1.7544... 0.0660 sec/batch
Epoch: 2/20... Training Step: 6774... Training loss: 1.7264... 0.0618 sec/batch
Epoch: 2/20... Training Step: 6775... Training loss: 1.7824... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6776... Training loss: 1.7870... 0.0722 sec/batch
Epoch: 2/20... Training Step: 6777... Training loss: 1.7556... 0.0607 sec/batch
Epoch: 2/20... Training Step: 6778... Training loss: 1.7752... 0.0873 sec/batch
Epoch: 2/20... Training Step: 6779... Training loss: 1.7768... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6780... Training loss: 1.7858... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6781... Training loss: 1.7245... 0.0617 sec/batch
Epoch: 2/20... Training Step: 6782... Training loss: 1.6878... 0.0732 sec/batch
Epoch: 2/20... Training Step: 6783... Training loss: 1.7255... 0.0744 sec/batch
Epoch: 2/20... Training Step: 6784... Training loss: 1.6270... 0.0787 sec/batch
Epoch: 2/20... Training Step: 6785... Training loss: 1.8473... 0.0837 sec/batch
Epoch: 2/20... Training Step: 6786... Training loss: 1.8649... 0.0811 sec/batch
Epoch: 2/20... Training Step: 6787... Training loss: 1.7228... 0.0697 sec/batch
Epoch: 2/20... Training Step: 6788... Training loss: 1.8458... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6789... Training loss: 1.6947... 0.0834 sec/batch
Epoch: 2/20... Training Step: 6790... Training loss: 1.8537... 0.0705 sec/batch
Epoch: 2/20... Training Step: 6791... Training loss: 1.8176... 0.0792 sec/batch
Epoch: 2/20... Training Step: 6792... Training loss: 1.7243... 0.0628 sec/batch
Epoch: 2/20... Training Step: 6793... Training loss: 1.8324... 0.0825 sec/batch
Epoch: 2/20... Training Step: 6794... Training loss: 1.6051... 0.0823 sec/batch
Epoch: 2/20... Training Step: 6795... Training loss: 1.8809... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6796... Training loss: 1.8004... 0.0799 sec/batch
Epoch: 2/20... Training Step: 6797... Training loss: 1.7691... 0.0736 sec/batch
Epoch: 2/20... Training Step: 6798... Training loss: 1.8347... 0.0652 sec/batch
Epoch: 2/20... Training Step: 6799... Training loss: 1.7631... 0.0829 sec/batch
Epoch: 2/20... Training Step: 6800... Training loss: 1.7839... 0.0699 sec/batch
Epoch: 2/20... Training Step: 6801... Training loss: 1.7514... 0.0788 sec/batch
Epoch: 2/20... Training Step: 6802... Training loss: 1.8264... 0.0882 sec/batch
Epoch: 2/20... Training Step: 6803... Training loss: 1.7324... 0.0800 sec/batch
Epoch: 2/20... Training Step: 6804... Training loss: 1.8360... 0.0614 sec/batch
Epoch: 2/20... Training Step: 6805... Training loss: 1.7431... 0.0762 sec/batch
Epoch: 2/20... Training Step: 6806... Training loss: 1.8728... 0.0820 sec/batch
Epoch: 2/20... Training Step: 6807... Training loss: 1.8202... 0.0596 sec/batch
Epoch: 2/20... Training Step: 6808... Training loss: 1.8073... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6809... Training loss: 1.7638... 0.0590 sec/batch
Epoch: 2/20... Training Step: 6810... Training loss: 1.7675... 0.0726 sec/batch
Epoch: 2/20... Training Step: 6811... Training loss: 1.8580... 0.0745 sec/batch
Epoch: 2/20... Training Step: 6812... Training loss: 1.8224... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6813... Training loss: 1.5858... 0.0740 sec/batch
Epoch: 2/20... Training Step: 6814... Training loss: 1.5955... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6815... Training loss: 1.7305... 0.0667 sec/batch
Epoch: 2/20... Training Step: 6816... Training loss: 1.7429... 0.0716 sec/batch
Epoch: 2/20... Training Step: 6817... Training loss: 1.7333... 0.0631 sec/batch
Epoch: 2/20... Training Step: 6818... Training loss: 1.7201... 0.0813 sec/batch
Epoch: 2/20... Training Step: 6819... Training loss: 1.7493... 0.0648 sec/batch
Epoch: 2/20... Training Step: 6820... Training loss: 1.9109... 0.0842 sec/batch
Epoch: 2/20... Training Step: 6821... Training loss: 1.7818... 0.0855 sec/batch
Epoch: 2/20... Training Step: 6822... Training loss: 1.8891... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6823... Training loss: 1.7981... 0.0734 sec/batch
Epoch: 2/20... Training Step: 6824... Training loss: 1.7824... 0.0659 sec/batch
Epoch: 2/20... Training Step: 6825... Training loss: 1.6752... 0.0672 sec/batch
Epoch: 2/20... Training Step: 6826... Training loss: 1.8784... 0.0895 sec/batch
Epoch: 2/20... Training Step: 6827... Training loss: 1.6902... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6828... Training loss: 1.9004... 0.0859 sec/batch
Epoch: 2/20... Training Step: 6829... Training loss: 1.4828... 0.0764 sec/batch
Epoch: 2/20... Training Step: 6830... Training loss: 1.6685... 0.0647 sec/batch
Epoch: 2/20... Training Step: 6831... Training loss: 1.8299... 0.0684 sec/batch
Epoch: 2/20... Training Step: 6832... Training loss: 1.5386... 0.0690 sec/batch
Epoch: 2/20... Training Step: 6833... Training loss: 1.7370... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6834... Training loss: 1.7121... 0.0653 sec/batch
Epoch: 2/20... Training Step: 6835... Training loss: 1.5991... 0.0691 sec/batch
Epoch: 2/20... Training Step: 6836... Training loss: 1.6259... 0.0690 sec/batch
Epoch: 2/20... Training Step: 6837... Training loss: 1.6504... 0.0730 sec/batch
Epoch: 2/20... Training Step: 6838... Training loss: 1.7588... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6839... Training loss: 1.8136... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6840... Training loss: 1.7289... 0.0623 sec/batch
Epoch: 2/20... Training Step: 6841... Training loss: 1.5938... 0.0704 sec/batch
Epoch: 2/20... Training Step: 6842... Training loss: 1.5938... 0.0736 sec/batch
Epoch: 2/20... Training Step: 6843... Training loss: 1.7821... 0.0732 sec/batch
Epoch: 2/20... Training Step: 6844... Training loss: 1.7056... 0.0696 sec/batch
Epoch: 2/20... Training Step: 6845... Training loss: 1.7710... 0.0814 sec/batch
Epoch: 2/20... Training Step: 6846... Training loss: 1.9230... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6847... Training loss: 1.7813... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6848... Training loss: 1.6024... 0.0612 sec/batch
Epoch: 2/20... Training Step: 6849... Training loss: 1.9168... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6850... Training loss: 1.6762... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6851... Training loss: 1.8346... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6852... Training loss: 1.8216... 0.0767 sec/batch
Epoch: 2/20... Training Step: 6853... Training loss: 1.8083... 0.0773 sec/batch
Epoch: 2/20... Training Step: 6854... Training loss: 1.6726... 0.0665 sec/batch
Epoch: 2/20... Training Step: 6855... Training loss: 1.7723... 0.0647 sec/batch
Epoch: 2/20... Training Step: 6856... Training loss: 1.6306... 0.0824 sec/batch
Epoch: 2/20... Training Step: 6857... Training loss: 1.6363... 0.0739 sec/batch
Epoch: 2/20... Training Step: 6858... Training loss: 1.7389... 0.0674 sec/batch
Epoch: 2/20... Training Step: 6859... Training loss: 1.7971... 0.0696 sec/batch
Epoch: 2/20... Training Step: 6860... Training loss: 1.6776... 0.0630 sec/batch
Epoch: 2/20... Training Step: 6861... Training loss: 1.8497... 0.0676 sec/batch
Epoch: 2/20... Training Step: 6862... Training loss: 1.6150... 0.0597 sec/batch
Epoch: 2/20... Training Step: 6863... Training loss: 1.8258... 0.0693 sec/batch
Epoch: 2/20... Training Step: 6864... Training loss: 1.7805... 0.0673 sec/batch
Epoch: 2/20... Training Step: 6865... Training loss: 1.6183... 0.0599 sec/batch
Epoch: 2/20... Training Step: 6866... Training loss: 1.5754... 0.0678 sec/batch
Epoch: 2/20... Training Step: 6867... Training loss: 1.6402... 0.0660 sec/batch
Epoch: 2/20... Training Step: 6868... Training loss: 1.7889... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6869... Training loss: 1.7253... 0.0747 sec/batch
Epoch: 2/20... Training Step: 6870... Training loss: 1.5820... 0.0652 sec/batch
Epoch: 2/20... Training Step: 6871... Training loss: 1.5525... 0.0758 sec/batch
Epoch: 2/20... Training Step: 6872... Training loss: 1.7255... 0.0735 sec/batch
Epoch: 2/20... Training Step: 6873... Training loss: 1.6141... 0.0732 sec/batch
Epoch: 2/20... Training Step: 6874... Training loss: 1.7289... 0.0565 sec/batch
Epoch: 2/20... Training Step: 6875... Training loss: 1.7520... 0.0621 sec/batch
Epoch: 2/20... Training Step: 6876... Training loss: 1.7305... 0.0627 sec/batch
Epoch: 2/20... Training Step: 6877... Training loss: 1.7024... 0.0677 sec/batch
Epoch: 2/20... Training Step: 6878... Training loss: 1.7160... 0.0810 sec/batch
Epoch: 2/20... Training Step: 6879... Training loss: 1.5991... 0.0612 sec/batch
Epoch: 2/20... Training Step: 6880... Training loss: 1.7677... 0.0677 sec/batch
Epoch: 2/20... Training Step: 6881... Training loss: 1.5799... 0.0576 sec/batch
Epoch: 2/20... Training Step: 6882... Training loss: 1.8225... 0.0649 sec/batch
Epoch: 2/20... Training Step: 6883... Training loss: 1.7229... 0.0668 sec/batch
Epoch: 2/20... Training Step: 6884... Training loss: 1.7318... 0.0816 sec/batch
Epoch: 2/20... Training Step: 6885... Training loss: 1.8958... 0.0735 sec/batch
Epoch: 2/20... Training Step: 6886... Training loss: 1.7281... 0.0722 sec/batch
Epoch: 2/20... Training Step: 6887... Training loss: 1.7668... 0.0771 sec/batch
Epoch: 2/20... Training Step: 6888... Training loss: 1.7227... 0.0639 sec/batch
Epoch: 2/20... Training Step: 6889... Training loss: 1.7553... 0.0636 sec/batch
Epoch: 2/20... Training Step: 6890... Training loss: 1.8080... 0.0616 sec/batch
Epoch: 2/20... Training Step: 6891... Training loss: 1.8143... 0.0702 sec/batch
Epoch: 2/20... Training Step: 6892... Training loss: 1.7021... 0.0824 sec/batch
Epoch: 2/20... Training Step: 6893... Training loss: 1.6257... 0.0720 sec/batch
Epoch: 2/20... Training Step: 6894... Training loss: 1.8403... 0.0757 sec/batch
Epoch: 2/20... Training Step: 6895... Training loss: 1.6634... 0.0665 sec/batch
Epoch: 2/20... Training Step: 6896... Training loss: 1.6225... 0.0747 sec/batch
Epoch: 2/20... Training Step: 6897... Training loss: 1.7102... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6898... Training loss: 1.7505... 0.0804 sec/batch
Epoch: 2/20... Training Step: 6899... Training loss: 1.5473... 0.0687 sec/batch
Epoch: 2/20... Training Step: 6900... Training loss: 1.6151... 0.0759 sec/batch
Epoch: 2/20... Training Step: 6901... Training loss: 1.6058... 0.0644 sec/batch
Epoch: 2/20... Training Step: 6902... Training loss: 1.5596... 0.0694 sec/batch
Epoch: 2/20... Training Step: 6903... Training loss: 1.6880... 0.0714 sec/batch
Epoch: 2/20... Training Step: 6904... Training loss: 1.6958... 0.0616 sec/batch
Epoch: 2/20... Training Step: 6905... Training loss: 1.6726... 0.0600 sec/batch
Epoch: 2/20... Training Step: 6906... Training loss: 1.7260... 0.0681 sec/batch
Epoch: 2/20... Training Step: 6907... Training loss: 1.7860... 0.0683 sec/batch
Epoch: 2/20... Training Step: 6908... Training loss: 1.6338... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6909... Training loss: 1.6952... 0.0714 sec/batch
Epoch: 2/20... Training Step: 6910... Training loss: 1.7546... 0.0718 sec/batch
Epoch: 2/20... Training Step: 6911... Training loss: 1.8290... 0.0781 sec/batch
Epoch: 2/20... Training Step: 6912... Training loss: 1.9034... 0.0737 sec/batch
Epoch: 2/20... Training Step: 6913... Training loss: 1.7511... 0.0856 sec/batch
Epoch: 2/20... Training Step: 6914... Training loss: 1.6381... 0.0653 sec/batch
Epoch: 2/20... Training Step: 6915... Training loss: 1.6504... 0.0897 sec/batch
Epoch: 2/20... Training Step: 6916... Training loss: 1.7426... 0.0766 sec/batch
Epoch: 2/20... Training Step: 6917... Training loss: 1.7331... 0.0666 sec/batch
Epoch: 2/20... Training Step: 6918... Training loss: 1.6055... 0.0673 sec/batch
Epoch: 2/20... Training Step: 6919... Training loss: 1.5977... 0.0715 sec/batch
Epoch: 2/20... Training Step: 6920... Training loss: 1.5441... 0.0876 sec/batch
Epoch: 2/20... Training Step: 6921... Training loss: 1.6385... 0.0726 sec/batch
Epoch: 2/20... Training Step: 6922... Training loss: 1.6484... 0.0678 sec/batch
Epoch: 2/20... Training Step: 6923... Training loss: 1.6183... 0.0576 sec/batch
Epoch: 2/20... Training Step: 6924... Training loss: 1.6589... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6925... Training loss: 1.5952... 0.0691 sec/batch
Epoch: 2/20... Training Step: 6926... Training loss: 1.6387... 0.0775 sec/batch
Epoch: 2/20... Training Step: 6927... Training loss: 1.6457... 0.0751 sec/batch
Epoch: 2/20... Training Step: 6928... Training loss: 1.6169... 0.0800 sec/batch
Epoch: 2/20... Training Step: 6929... Training loss: 1.6998... 0.0783 sec/batch
Epoch: 2/20... Training Step: 6930... Training loss: 1.7494... 0.0741 sec/batch
Epoch: 2/20... Training Step: 6931... Training loss: 1.6996... 0.0663 sec/batch
Epoch: 2/20... Training Step: 6932... Training loss: 1.7977... 0.0701 sec/batch
Epoch: 2/20... Training Step: 6933... Training loss: 1.7286... 0.0602 sec/batch
Epoch: 2/20... Training Step: 6934... Training loss: 1.6162... 0.0809 sec/batch
Epoch: 2/20... Training Step: 6935... Training loss: 1.6546... 0.0577 sec/batch
Epoch: 2/20... Training Step: 6936... Training loss: 1.6577... 0.0623 sec/batch
Epoch: 2/20... Training Step: 6937... Training loss: 1.6664... 0.0865 sec/batch
Epoch: 2/20... Training Step: 6938... Training loss: 1.7898... 0.0876 sec/batch
Epoch: 2/20... Training Step: 6939... Training loss: 1.6182... 0.0713 sec/batch
Epoch: 2/20... Training Step: 6940... Training loss: 1.6560... 0.0749 sec/batch
Epoch: 2/20... Training Step: 6941... Training loss: 1.8470... 0.0609 sec/batch
Epoch: 2/20... Training Step: 6942... Training loss: 1.6475... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6943... Training loss: 1.6941... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6944... Training loss: 1.7986... 0.0717 sec/batch
Epoch: 2/20... Training Step: 6945... Training loss: 1.7699... 0.0622 sec/batch
Epoch: 2/20... Training Step: 6946... Training loss: 1.6908... 0.0593 sec/batch
Epoch: 2/20... Training Step: 6947... Training loss: 1.6410... 0.0553 sec/batch
Epoch: 2/20... Training Step: 6948... Training loss: 1.7770... 0.0728 sec/batch
Epoch: 2/20... Training Step: 6949... Training loss: 1.7742... 0.0659 sec/batch
Epoch: 2/20... Training Step: 6950... Training loss: 1.8734... 0.0656 sec/batch
Epoch: 2/20... Training Step: 6951... Training loss: 1.8967... 0.0634 sec/batch
Epoch: 2/20... Training Step: 6952... Training loss: 1.8423... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6953... Training loss: 1.5855... 0.0753 sec/batch
Epoch: 2/20... Training Step: 6954... Training loss: 1.6076... 0.0731 sec/batch
Epoch: 2/20... Training Step: 6955... Training loss: 1.7342... 0.0760 sec/batch
Epoch: 2/20... Training Step: 6956... Training loss: 1.8548... 0.0740 sec/batch
Epoch: 2/20... Training Step: 6957... Training loss: 1.7099... 0.0692 sec/batch
Epoch: 2/20... Training Step: 6958... Training loss: 1.7816... 0.0675 sec/batch
Epoch: 2/20... Training Step: 6959... Training loss: 1.7612... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6960... Training loss: 1.7130... 0.0757 sec/batch
Epoch: 2/20... Training Step: 6961... Training loss: 1.8319... 0.0618 sec/batch
Epoch: 2/20... Training Step: 6962... Training loss: 1.7207... 0.0593 sec/batch
Epoch: 2/20... Training Step: 6963... Training loss: 1.5620... 0.0765 sec/batch
Epoch: 2/20... Training Step: 6964... Training loss: 1.7786... 0.0711 sec/batch
Epoch: 2/20... Training Step: 6965... Training loss: 1.7496... 0.0614 sec/batch
Epoch: 2/20... Training Step: 6966... Training loss: 1.8627... 0.0561 sec/batch
Epoch: 2/20... Training Step: 6967... Training loss: 1.8756... 0.0932 sec/batch
Epoch: 2/20... Training Step: 6968... Training loss: 1.7670... 0.0738 sec/batch
Epoch: 2/20... Training Step: 6969... Training loss: 1.6302... 0.0707 sec/batch
Epoch: 2/20... Training Step: 6970... Training loss: 1.6969... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6971... Training loss: 1.6676... 0.0712 sec/batch
Epoch: 2/20... Training Step: 6972... Training loss: 1.5373... 0.0889 sec/batch
Epoch: 2/20... Training Step: 6973... Training loss: 1.6963... 0.0619 sec/batch
Epoch: 2/20... Training Step: 6974... Training loss: 1.6601... 0.0606 sec/batch
Epoch: 2/20... Training Step: 6975... Training loss: 1.8248... 0.0777 sec/batch
Epoch: 2/20... Training Step: 6976... Training loss: 1.6762... 0.0831 sec/batch
Epoch: 2/20... Training Step: 6977... Training loss: 1.6849... 0.0795 sec/batch
Epoch: 2/20... Training Step: 6978... Training loss: 1.7690... 0.0753 sec/batch
Epoch: 2/20... Training Step: 6979... Training loss: 1.6591... 0.0742 sec/batch
Epoch: 2/20... Training Step: 6980... Training loss: 1.8302... 0.0861 sec/batch
Epoch: 2/20... Training Step: 6981... Training loss: 1.6500... 0.0872 sec/batch
Epoch: 2/20... Training Step: 6982... Training loss: 1.8707... 0.0650 sec/batch
Epoch: 2/20... Training Step: 6983... Training loss: 1.6767... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6984... Training loss: 1.7104... 0.0709 sec/batch
Epoch: 2/20... Training Step: 6985... Training loss: 1.8316... 0.0619 sec/batch
Epoch: 2/20... Training Step: 6986... Training loss: 1.6844... 0.0630 sec/batch
Epoch: 2/20... Training Step: 6987... Training loss: 1.7120... 0.0695 sec/batch
Epoch: 2/20... Training Step: 6988... Training loss: 1.7864... 0.0716 sec/batch
Epoch: 2/20... Training Step: 6989... Training loss: 1.6227... 0.0672 sec/batch
Epoch: 2/20... Training Step: 6990... Training loss: 1.8715... 0.0642 sec/batch
Epoch: 2/20... Training Step: 6991... Training loss: 1.7642... 0.0626 sec/batch
Epoch: 2/20... Training Step: 6992... Training loss: 1.7454... 0.0703 sec/batch
Epoch: 2/20... Training Step: 6993... Training loss: 1.7287... 0.0666 sec/batch
Epoch: 2/20... Training Step: 6994... Training loss: 1.6853... 0.0664 sec/batch
Epoch: 2/20... Training Step: 6995... Training loss: 1.7135... 0.0685 sec/batch
Epoch: 2/20... Training Step: 6996... Training loss: 1.6573... 0.0778 sec/batch
Epoch: 2/20... Training Step: 6997... Training loss: 1.7455... 0.0748 sec/batch
Epoch: 2/20... Training Step: 6998... Training loss: 1.8241... 0.0657 sec/batch
Epoch: 2/20... Training Step: 6999... Training loss: 1.9324... 0.0617 sec/batch
Epoch: 2/20... Training Step: 7000... Training loss: 1.6422... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7001... Training loss: 1.7723... 0.0845 sec/batch
Epoch: 2/20... Training Step: 7002... Training loss: 1.8254... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7003... Training loss: 1.7609... 0.0842 sec/batch
Epoch: 2/20... Training Step: 7004... Training loss: 1.7582... 0.0953 sec/batch
Epoch: 2/20... Training Step: 7005... Training loss: 1.4989... 0.0744 sec/batch
Epoch: 2/20... Training Step: 7006... Training loss: 1.8736... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7007... Training loss: 1.7884... 0.0644 sec/batch
Epoch: 2/20... Training Step: 7008... Training loss: 1.8950... 0.0722 sec/batch
Epoch: 2/20... Training Step: 7009... Training loss: 1.6687... 0.0907 sec/batch
Epoch: 2/20... Training Step: 7010... Training loss: 1.8322... 0.0786 sec/batch
Epoch: 2/20... Training Step: 7011... Training loss: 1.6740... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7012... Training loss: 1.7559... 0.0882 sec/batch
Epoch: 2/20... Training Step: 7013... Training loss: 1.8422... 0.0778 sec/batch
Epoch: 2/20... Training Step: 7014... Training loss: 1.7437... 0.0774 sec/batch
Epoch: 2/20... Training Step: 7015... Training loss: 1.7738... 0.0675 sec/batch
Epoch: 2/20... Training Step: 7016... Training loss: 1.7671... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7017... Training loss: 1.6114... 0.0811 sec/batch
Epoch: 2/20... Training Step: 7018... Training loss: 1.9181... 0.0690 sec/batch
Epoch: 2/20... Training Step: 7019... Training loss: 1.7870... 0.0906 sec/batch
Epoch: 2/20... Training Step: 7020... Training loss: 1.8773... 0.0631 sec/batch
Epoch: 2/20... Training Step: 7021... Training loss: 1.8161... 0.0656 sec/batch
Epoch: 2/20... Training Step: 7022... Training loss: 1.8147... 0.0700 sec/batch
Epoch: 2/20... Training Step: 7023... Training loss: 1.9149... 0.0833 sec/batch
Epoch: 2/20... Training Step: 7024... Training loss: 1.7301... 0.0861 sec/batch
Epoch: 2/20... Training Step: 7025... Training loss: 1.6435... 0.0733 sec/batch
Epoch: 2/20... Training Step: 7026... Training loss: 1.6622... 0.0688 sec/batch
Epoch: 2/20... Training Step: 7027... Training loss: 1.8893... 0.0682 sec/batch
Epoch: 2/20... Training Step: 7028... Training loss: 1.7002... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7029... Training loss: 1.9293... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7030... Training loss: 1.6992... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7031... Training loss: 1.8129... 0.0566 sec/batch
Epoch: 2/20... Training Step: 7032... Training loss: 1.7693... 0.0824 sec/batch
Epoch: 2/20... Training Step: 7033... Training loss: 1.8232... 0.0738 sec/batch
Epoch: 2/20... Training Step: 7034... Training loss: 1.6180... 0.0828 sec/batch
Epoch: 2/20... Training Step: 7035... Training loss: 1.6617... 0.0823 sec/batch
Epoch: 2/20... Training Step: 7036... Training loss: 1.6927... 0.0734 sec/batch
Epoch: 2/20... Training Step: 7037... Training loss: 1.8754... 0.0783 sec/batch
Epoch: 2/20... Training Step: 7038... Training loss: 1.7361... 0.0734 sec/batch
Epoch: 2/20... Training Step: 7039... Training loss: 1.7505... 0.0823 sec/batch
Epoch: 2/20... Training Step: 7040... Training loss: 1.6253... 0.0853 sec/batch
Epoch: 2/20... Training Step: 7041... Training loss: 1.6677... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7042... Training loss: 1.7112... 0.0620 sec/batch
Epoch: 2/20... Training Step: 7043... Training loss: 1.7843... 0.0694 sec/batch
Epoch: 2/20... Training Step: 7044... Training loss: 1.5353... 0.0592 sec/batch
Epoch: 2/20... Training Step: 7045... Training loss: 1.7648... 0.0707 sec/batch
Epoch: 2/20... Training Step: 7046... Training loss: 1.7362... 0.0750 sec/batch
Epoch: 2/20... Training Step: 7047... Training loss: 1.7173... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7048... Training loss: 1.6600... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7049... Training loss: 1.6336... 0.0703 sec/batch
Epoch: 2/20... Training Step: 7050... Training loss: 1.7052... 0.0654 sec/batch
Epoch: 2/20... Training Step: 7051... Training loss: 1.7103... 0.0861 sec/batch
Epoch: 2/20... Training Step: 7052... Training loss: 1.6235... 0.0805 sec/batch
Epoch: 2/20... Training Step: 7053... Training loss: 1.7728... 0.0673 sec/batch
Epoch: 2/20... Training Step: 7054... Training loss: 1.6592... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7055... Training loss: 1.7701... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7056... Training loss: 1.6488... 0.0614 sec/batch
Epoch: 2/20... Training Step: 7057... Training loss: 1.7004... 0.0639 sec/batch
Epoch: 2/20... Training Step: 7058... Training loss: 1.7544... 0.0805 sec/batch
Epoch: 2/20... Training Step: 7059... Training loss: 1.7592... 0.0781 sec/batch
Epoch: 2/20... Training Step: 7060... Training loss: 1.7425... 0.0791 sec/batch
Epoch: 2/20... Training Step: 7061... Training loss: 1.6151... 0.0701 sec/batch
Epoch: 2/20... Training Step: 7062... Training loss: 1.6548... 0.0556 sec/batch
Epoch: 2/20... Training Step: 7063... Training loss: 1.7883... 0.0620 sec/batch
Epoch: 2/20... Training Step: 7064... Training loss: 1.8309... 0.0634 sec/batch
Epoch: 2/20... Training Step: 7065... Training loss: 1.7940... 0.0716 sec/batch
Epoch: 2/20... Training Step: 7066... Training loss: 1.7336... 0.0758 sec/batch
Epoch: 2/20... Training Step: 7067... Training loss: 1.8153... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7068... Training loss: 1.7069... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7069... Training loss: 1.5940... 0.0615 sec/batch
Epoch: 2/20... Training Step: 7070... Training loss: 1.6293... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7071... Training loss: 1.8359... 0.0779 sec/batch
Epoch: 2/20... Training Step: 7072... Training loss: 1.7216... 0.0766 sec/batch
Epoch: 2/20... Training Step: 7073... Training loss: 1.5735... 0.0666 sec/batch
Epoch: 2/20... Training Step: 7074... Training loss: 1.7112... 0.0777 sec/batch
Epoch: 2/20... Training Step: 7075... Training loss: 1.6763... 0.0574 sec/batch
Epoch: 2/20... Training Step: 7076... Training loss: 1.6250... 0.0783 sec/batch
Epoch: 2/20... Training Step: 7077... Training loss: 1.8069... 0.0751 sec/batch
Epoch: 2/20... Training Step: 7078... Training loss: 1.6319... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7079... Training loss: 1.5430... 0.0802 sec/batch
Epoch: 2/20... Training Step: 7080... Training loss: 1.6637... 0.0694 sec/batch
Epoch: 2/20... Training Step: 7081... Training loss: 1.7001... 0.0706 sec/batch
Epoch: 2/20... Training Step: 7082... Training loss: 1.6536... 0.0730 sec/batch
Epoch: 2/20... Training Step: 7083... Training loss: 1.6819... 0.0722 sec/batch
Epoch: 2/20... Training Step: 7084... Training loss: 1.6708... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7085... Training loss: 1.6998... 0.0581 sec/batch
Epoch: 2/20... Training Step: 7086... Training loss: 1.6128... 0.0777 sec/batch
Epoch: 2/20... Training Step: 7087... Training loss: 1.5786... 0.0728 sec/batch
Epoch: 2/20... Training Step: 7088... Training loss: 1.7017... 0.0862 sec/batch
Epoch: 2/20... Training Step: 7089... Training loss: 1.6977... 0.0606 sec/batch
Epoch: 2/20... Training Step: 7090... Training loss: 1.6350... 0.0642 sec/batch
Epoch: 2/20... Training Step: 7091... Training loss: 1.6926... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7092... Training loss: 1.6600... 0.0662 sec/batch
Epoch: 2/20... Training Step: 7093... Training loss: 1.7799... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7094... Training loss: 1.6960... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7095... Training loss: 1.6940... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7096... Training loss: 1.7694... 0.0693 sec/batch
Epoch: 2/20... Training Step: 7097... Training loss: 1.7344... 0.0714 sec/batch
Epoch: 2/20... Training Step: 7098... Training loss: 1.6470... 0.0628 sec/batch
Epoch: 2/20... Training Step: 7099... Training loss: 1.5647... 0.0637 sec/batch
Epoch: 2/20... Training Step: 7100... Training loss: 1.6702... 0.0633 sec/batch
Epoch: 2/20... Training Step: 7101... Training loss: 1.7103... 0.0615 sec/batch
Epoch: 2/20... Training Step: 7102... Training loss: 1.6730... 0.0635 sec/batch
Epoch: 2/20... Training Step: 7103... Training loss: 1.8551... 0.0663 sec/batch
Epoch: 2/20... Training Step: 7104... Training loss: 1.7119... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7105... Training loss: 1.8087... 0.0634 sec/batch
Epoch: 2/20... Training Step: 7106... Training loss: 1.7776... 0.0737 sec/batch
Epoch: 2/20... Training Step: 7107... Training loss: 1.5553... 0.0647 sec/batch
Epoch: 2/20... Training Step: 7108... Training loss: 1.7139... 0.0595 sec/batch
Epoch: 2/20... Training Step: 7109... Training loss: 1.7516... 0.0635 sec/batch
Epoch: 2/20... Training Step: 7110... Training loss: 1.8342... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7111... Training loss: 1.7799... 0.0631 sec/batch
Epoch: 2/20... Training Step: 7112... Training loss: 1.6444... 0.0809 sec/batch
Epoch: 2/20... Training Step: 7113... Training loss: 1.7253... 0.0693 sec/batch
Epoch: 2/20... Training Step: 7114... Training loss: 1.5968... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7115... Training loss: 1.7804... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7116... Training loss: 1.6944... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7117... Training loss: 1.6740... 0.0861 sec/batch
Epoch: 2/20... Training Step: 7118... Training loss: 1.7276... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7119... Training loss: 1.6767... 0.0782 sec/batch
Epoch: 2/20... Training Step: 7120... Training loss: 1.6564... 0.0820 sec/batch
Epoch: 2/20... Training Step: 7121... Training loss: 1.7341... 0.0875 sec/batch
Epoch: 2/20... Training Step: 7122... Training loss: 1.7353... 0.0678 sec/batch
Epoch: 2/20... Training Step: 7123... Training loss: 1.7058... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7124... Training loss: 1.6208... 0.0581 sec/batch
Epoch: 2/20... Training Step: 7125... Training loss: 1.7632... 0.0910 sec/batch
Epoch: 2/20... Training Step: 7126... Training loss: 1.8041... 0.0820 sec/batch
Epoch: 2/20... Training Step: 7127... Training loss: 1.6115... 0.0693 sec/batch
Epoch: 2/20... Training Step: 7128... Training loss: 1.6832... 0.0658 sec/batch
Epoch: 2/20... Training Step: 7129... Training loss: 1.6659... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7130... Training loss: 1.6621... 0.0764 sec/batch
Epoch: 2/20... Training Step: 7131... Training loss: 1.8077... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7132... Training loss: 1.7458... 0.0682 sec/batch
Epoch: 2/20... Training Step: 7133... Training loss: 1.8479... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7134... Training loss: 1.6980... 0.0836 sec/batch
Epoch: 2/20... Training Step: 7135... Training loss: 1.5981... 0.0682 sec/batch
Epoch: 2/20... Training Step: 7136... Training loss: 1.6165... 0.0814 sec/batch
Epoch: 2/20... Training Step: 7137... Training loss: 1.6450... 0.0781 sec/batch
Epoch: 2/20... Training Step: 7138... Training loss: 1.6743... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7139... Training loss: 1.6306... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7140... Training loss: 1.7554... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7141... Training loss: 1.7658... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7142... Training loss: 1.7059... 0.0682 sec/batch
Epoch: 2/20... Training Step: 7143... Training loss: 1.7570... 0.0646 sec/batch
Epoch: 2/20... Training Step: 7144... Training loss: 1.5942... 0.0649 sec/batch
Epoch: 2/20... Training Step: 7145... Training loss: 1.6997... 0.0662 sec/batch
Epoch: 2/20... Training Step: 7146... Training loss: 1.7603... 0.0715 sec/batch
Epoch: 2/20... Training Step: 7147... Training loss: 1.8227... 0.0587 sec/batch
Epoch: 2/20... Training Step: 7148... Training loss: 1.6254... 0.0744 sec/batch
Epoch: 2/20... Training Step: 7149... Training loss: 1.7487... 0.0733 sec/batch
Epoch: 2/20... Training Step: 7150... Training loss: 1.5607... 0.0668 sec/batch
Epoch: 2/20... Training Step: 7151... Training loss: 1.6792... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7152... Training loss: 1.6310... 0.0820 sec/batch
Epoch: 2/20... Training Step: 7153... Training loss: 1.7052... 0.0702 sec/batch
Epoch: 2/20... Training Step: 7154... Training loss: 1.7042... 0.0750 sec/batch
Epoch: 2/20... Training Step: 7155... Training loss: 1.8153... 0.0837 sec/batch
Epoch: 2/20... Training Step: 7156... Training loss: 1.8030... 0.0766 sec/batch
Epoch: 2/20... Training Step: 7157... Training loss: 1.5623... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7158... Training loss: 1.8021... 0.0799 sec/batch
Epoch: 2/20... Training Step: 7159... Training loss: 1.6486... 0.0843 sec/batch
Epoch: 2/20... Training Step: 7160... Training loss: 1.6814... 0.0759 sec/batch
Epoch: 2/20... Training Step: 7161... Training loss: 1.5919... 0.0816 sec/batch
Epoch: 2/20... Training Step: 7162... Training loss: 1.5758... 0.0719 sec/batch
Epoch: 2/20... Training Step: 7163... Training loss: 1.7534... 0.0745 sec/batch
Epoch: 2/20... Training Step: 7164... Training loss: 1.7265... 0.0742 sec/batch
Epoch: 2/20... Training Step: 7165... Training loss: 1.6798... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7166... Training loss: 1.7419... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7167... Training loss: 1.7160... 0.0633 sec/batch
Epoch: 2/20... Training Step: 7168... Training loss: 1.7713... 0.0736 sec/batch
Epoch: 2/20... Training Step: 7169... Training loss: 1.7032... 0.0615 sec/batch
Epoch: 2/20... Training Step: 7170... Training loss: 1.7782... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7171... Training loss: 1.7811... 0.0709 sec/batch
Epoch: 2/20... Training Step: 7172... Training loss: 1.6602... 0.0643 sec/batch
Epoch: 2/20... Training Step: 7173... Training loss: 1.6547... 0.0724 sec/batch
Epoch: 2/20... Training Step: 7174... Training loss: 1.8094... 0.0632 sec/batch
Epoch: 2/20... Training Step: 7175... Training loss: 1.7149... 0.0824 sec/batch
Epoch: 2/20... Training Step: 7176... Training loss: 1.7372... 0.0726 sec/batch
Epoch: 2/20... Training Step: 7177... Training loss: 1.7097... 0.0746 sec/batch
Epoch: 2/20... Training Step: 7178... Training loss: 1.7789... 0.0643 sec/batch
Epoch: 2/20... Training Step: 7179... Training loss: 1.6465... 0.0666 sec/batch
Epoch: 2/20... Training Step: 7180... Training loss: 1.7034... 0.0688 sec/batch
Epoch: 2/20... Training Step: 7181... Training loss: 1.9042... 0.0634 sec/batch
Epoch: 2/20... Training Step: 7182... Training loss: 1.8323... 0.0597 sec/batch
Epoch: 2/20... Training Step: 7183... Training loss: 1.8335... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7184... Training loss: 1.7806... 0.0731 sec/batch
Epoch: 2/20... Training Step: 7185... Training loss: 1.7353... 0.0629 sec/batch
Epoch: 2/20... Training Step: 7186... Training loss: 1.7933... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7187... Training loss: 1.6549... 0.0738 sec/batch
Epoch: 2/20... Training Step: 7188... Training loss: 1.5640... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7189... Training loss: 1.6297... 0.0846 sec/batch
Epoch: 2/20... Training Step: 7190... Training loss: 1.7322... 0.0755 sec/batch
Epoch: 2/20... Training Step: 7191... Training loss: 1.5790... 0.0634 sec/batch
Epoch: 2/20... Training Step: 7192... Training loss: 1.7862... 0.0790 sec/batch
Epoch: 2/20... Training Step: 7193... Training loss: 1.6678... 0.0672 sec/batch
Epoch: 2/20... Training Step: 7194... Training loss: 1.6484... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7195... Training loss: 1.6495... 0.0613 sec/batch
Epoch: 2/20... Training Step: 7196... Training loss: 1.6010... 0.0895 sec/batch
Epoch: 2/20... Training Step: 7197... Training loss: 1.6984... 0.0703 sec/batch
Epoch: 2/20... Training Step: 7198... Training loss: 1.7223... 0.0716 sec/batch
Epoch: 2/20... Training Step: 7199... Training loss: 1.6570... 0.0652 sec/batch
Epoch: 2/20... Training Step: 7200... Training loss: 1.8335... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7201... Training loss: 1.7741... 0.0779 sec/batch
Epoch: 2/20... Training Step: 7202... Training loss: 1.6239... 0.0775 sec/batch
Epoch: 2/20... Training Step: 7203... Training loss: 1.6684... 0.0660 sec/batch
Epoch: 2/20... Training Step: 7204... Training loss: 1.6097... 0.0618 sec/batch
Epoch: 2/20... Training Step: 7205... Training loss: 1.7079... 0.0632 sec/batch
Epoch: 2/20... Training Step: 7206... Training loss: 1.7592... 0.0756 sec/batch
Epoch: 2/20... Training Step: 7207... Training loss: 1.6755... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7208... Training loss: 1.6329... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7209... Training loss: 1.6736... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7210... Training loss: 1.6752... 0.0640 sec/batch
Epoch: 2/20... Training Step: 7211... Training loss: 1.7495... 0.0543 sec/batch
Epoch: 2/20... Training Step: 7212... Training loss: 1.6298... 0.0747 sec/batch
Epoch: 2/20... Training Step: 7213... Training loss: 1.7317... 0.0726 sec/batch
Epoch: 2/20... Training Step: 7214... Training loss: 1.7710... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7215... Training loss: 1.7615... 0.0601 sec/batch
Epoch: 2/20... Training Step: 7216... Training loss: 1.7913... 0.0633 sec/batch
Epoch: 2/20... Training Step: 7217... Training loss: 1.7523... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7218... Training loss: 1.8038... 0.0791 sec/batch
Epoch: 2/20... Training Step: 7219... Training loss: 1.8763... 0.0863 sec/batch
Epoch: 2/20... Training Step: 7220... Training loss: 1.7213... 0.0714 sec/batch
Epoch: 2/20... Training Step: 7221... Training loss: 1.8450... 0.0760 sec/batch
Epoch: 2/20... Training Step: 7222... Training loss: 1.8359... 0.0728 sec/batch
Epoch: 2/20... Training Step: 7223... Training loss: 1.9539... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7224... Training loss: 1.6994... 0.0774 sec/batch
Epoch: 2/20... Training Step: 7225... Training loss: 1.6525... 0.0848 sec/batch
Epoch: 2/20... Training Step: 7226... Training loss: 1.7667... 0.0632 sec/batch
Epoch: 2/20... Training Step: 7227... Training loss: 1.5877... 0.0751 sec/batch
Epoch: 2/20... Training Step: 7228... Training loss: 1.6638... 0.0670 sec/batch
Epoch: 2/20... Training Step: 7229... Training loss: 1.8925... 0.0657 sec/batch
Epoch: 2/20... Training Step: 7230... Training loss: 1.6786... 0.0765 sec/batch
Epoch: 2/20... Training Step: 7231... Training loss: 1.7397... 0.0801 sec/batch
Epoch: 2/20... Training Step: 7232... Training loss: 1.6563... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7233... Training loss: 1.6609... 0.0705 sec/batch
Epoch: 2/20... Training Step: 7234... Training loss: 1.5602... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7235... Training loss: 1.6469... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7236... Training loss: 1.7218... 0.0741 sec/batch
Epoch: 2/20... Training Step: 7237... Training loss: 1.6223... 0.0619 sec/batch
Epoch: 2/20... Training Step: 7238... Training loss: 1.6934... 0.0701 sec/batch
Epoch: 2/20... Training Step: 7239... Training loss: 1.5920... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7240... Training loss: 1.8307... 0.0749 sec/batch
Epoch: 2/20... Training Step: 7241... Training loss: 1.7333... 0.0597 sec/batch
Epoch: 2/20... Training Step: 7242... Training loss: 1.6599... 0.0589 sec/batch
Epoch: 2/20... Training Step: 7243... Training loss: 1.8004... 0.0626 sec/batch
Epoch: 2/20... Training Step: 7244... Training loss: 1.6619... 0.0775 sec/batch
Epoch: 2/20... Training Step: 7245... Training loss: 1.6581... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7246... Training loss: 1.7229... 0.0726 sec/batch
Epoch: 2/20... Training Step: 7247... Training loss: 1.7296... 0.0743 sec/batch
Epoch: 2/20... Training Step: 7248... Training loss: 1.7213... 0.0761 sec/batch
Epoch: 2/20... Training Step: 7249... Training loss: 1.7096... 0.0882 sec/batch
Epoch: 2/20... Training Step: 7250... Training loss: 1.6312... 0.0698 sec/batch
Epoch: 2/20... Training Step: 7251... Training loss: 1.6507... 0.0775 sec/batch
Epoch: 2/20... Training Step: 7252... Training loss: 1.6524... 0.0733 sec/batch
Epoch: 2/20... Training Step: 7253... Training loss: 1.6890... 0.0641 sec/batch
Epoch: 2/20... Training Step: 7254... Training loss: 1.6742... 0.0724 sec/batch
Epoch: 2/20... Training Step: 7255... Training loss: 1.6149... 0.0710 sec/batch
Epoch: 2/20... Training Step: 7256... Training loss: 1.6441... 0.0667 sec/batch
Epoch: 2/20... Training Step: 7257... Training loss: 1.6609... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7258... Training loss: 1.6810... 0.0747 sec/batch
Epoch: 2/20... Training Step: 7259... Training loss: 1.8047... 0.0620 sec/batch
Epoch: 2/20... Training Step: 7260... Training loss: 1.8411... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7261... Training loss: 1.8689... 0.0697 sec/batch
Epoch: 2/20... Training Step: 7262... Training loss: 1.7985... 0.0877 sec/batch
Epoch: 2/20... Training Step: 7263... Training loss: 1.6555... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7264... Training loss: 1.8118... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7265... Training loss: 1.7399... 0.0713 sec/batch
Epoch: 2/20... Training Step: 7266... Training loss: 1.7739... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7267... Training loss: 1.8257... 0.0711 sec/batch
Epoch: 2/20... Training Step: 7268... Training loss: 1.7985... 0.0675 sec/batch
Epoch: 2/20... Training Step: 7269... Training loss: 1.7333... 0.0759 sec/batch
Epoch: 2/20... Training Step: 7270... Training loss: 1.8063... 0.0609 sec/batch
Epoch: 2/20... Training Step: 7271... Training loss: 1.7655... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7272... Training loss: 1.7848... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7273... Training loss: 1.7529... 0.0713 sec/batch
Epoch: 2/20... Training Step: 7274... Training loss: 1.8070... 0.0718 sec/batch
Epoch: 2/20... Training Step: 7275... Training loss: 1.8244... 0.0880 sec/batch
Epoch: 2/20... Training Step: 7276... Training loss: 1.7921... 0.0990 sec/batch
Epoch: 2/20... Training Step: 7277... Training loss: 1.6209... 0.1279 sec/batch
Epoch: 2/20... Training Step: 7278... Training loss: 1.7112... 0.1151 sec/batch
Epoch: 2/20... Training Step: 7279... Training loss: 1.6587... 0.1208 sec/batch
Epoch: 2/20... Training Step: 7280... Training loss: 1.8066... 0.0991 sec/batch
Epoch: 2/20... Training Step: 7281... Training loss: 1.6983... 0.0952 sec/batch
Epoch: 2/20... Training Step: 7282... Training loss: 1.6741... 0.1042 sec/batch
Epoch: 2/20... Training Step: 7283... Training loss: 1.6596... 0.1012 sec/batch
Epoch: 2/20... Training Step: 7284... Training loss: 1.7817... 0.1121 sec/batch
Epoch: 2/20... Training Step: 7285... Training loss: 1.7333... 0.0977 sec/batch
Epoch: 2/20... Training Step: 7286... Training loss: 1.8099... 0.0995 sec/batch
Epoch: 2/20... Training Step: 7287... Training loss: 1.7594... 0.1162 sec/batch
Epoch: 2/20... Training Step: 7288... Training loss: 1.6962... 0.1267 sec/batch
Epoch: 2/20... Training Step: 7289... Training loss: 1.7684... 0.1014 sec/batch
Epoch: 2/20... Training Step: 7290... Training loss: 1.7180... 0.1158 sec/batch
Epoch: 2/20... Training Step: 7291... Training loss: 1.7203... 0.1269 sec/batch
Epoch: 2/20... Training Step: 7292... Training loss: 1.6199... 0.1194 sec/batch
Epoch: 2/20... Training Step: 7293... Training loss: 1.7652... 0.1174 sec/batch
Epoch: 2/20... Training Step: 7294... Training loss: 1.5972... 0.1153 sec/batch
Epoch: 2/20... Training Step: 7295... Training loss: 1.8371... 0.1362 sec/batch
Epoch: 2/20... Training Step: 7296... Training loss: 1.8244... 0.1052 sec/batch
Epoch: 2/20... Training Step: 7297... Training loss: 1.7764... 0.1197 sec/batch
Epoch: 2/20... Training Step: 7298... Training loss: 1.7014... 0.1201 sec/batch
Epoch: 2/20... Training Step: 7299... Training loss: 1.9012... 0.1038 sec/batch
Epoch: 2/20... Training Step: 7300... Training loss: 1.8296... 0.1006 sec/batch
Epoch: 2/20... Training Step: 7301... Training loss: 1.6185... 0.0949 sec/batch
Epoch: 2/20... Training Step: 7302... Training loss: 1.8214... 0.0887 sec/batch
Epoch: 2/20... Training Step: 7303... Training loss: 1.7151... 0.1211 sec/batch
Epoch: 2/20... Training Step: 7304... Training loss: 1.7033... 0.1073 sec/batch
Epoch: 2/20... Training Step: 7305... Training loss: 1.7823... 0.1227 sec/batch
Epoch: 2/20... Training Step: 7306... Training loss: 1.8779... 0.1194 sec/batch
Epoch: 2/20... Training Step: 7307... Training loss: 1.7133... 0.1156 sec/batch
Epoch: 2/20... Training Step: 7308... Training loss: 1.7855... 0.1332 sec/batch
Epoch: 2/20... Training Step: 7309... Training loss: 1.6188... 0.1039 sec/batch
Epoch: 2/20... Training Step: 7310... Training loss: 1.6895... 0.0956 sec/batch
Epoch: 2/20... Training Step: 7311... Training loss: 1.8020... 0.0888 sec/batch
Epoch: 2/20... Training Step: 7312... Training loss: 1.7563... 0.0774 sec/batch
Epoch: 2/20... Training Step: 7313... Training loss: 1.7327... 0.0755 sec/batch
Epoch: 2/20... Training Step: 7314... Training loss: 1.7111... 0.0711 sec/batch
Epoch: 2/20... Training Step: 7315... Training loss: 1.7546... 0.0876 sec/batch
Epoch: 2/20... Training Step: 7316... Training loss: 1.8744... 0.0698 sec/batch
Epoch: 2/20... Training Step: 7317... Training loss: 1.8429... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7318... Training loss: 1.8030... 0.0896 sec/batch
Epoch: 2/20... Training Step: 7319... Training loss: 1.8699... 0.0964 sec/batch
Epoch: 2/20... Training Step: 7320... Training loss: 1.6835... 0.0880 sec/batch
Epoch: 2/20... Training Step: 7321... Training loss: 1.9166... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7322... Training loss: 1.6186... 0.0825 sec/batch
Epoch: 2/20... Training Step: 7323... Training loss: 1.6476... 0.0835 sec/batch
Epoch: 2/20... Training Step: 7324... Training loss: 1.5948... 0.0912 sec/batch
Epoch: 2/20... Training Step: 7325... Training loss: 1.6851... 0.0868 sec/batch
Epoch: 2/20... Training Step: 7326... Training loss: 1.6465... 0.0848 sec/batch
Epoch: 2/20... Training Step: 7327... Training loss: 1.7995... 0.0901 sec/batch
Epoch: 2/20... Training Step: 7328... Training loss: 1.8177... 0.0903 sec/batch
Epoch: 2/20... Training Step: 7329... Training loss: 1.7403... 0.0783 sec/batch
Epoch: 2/20... Training Step: 7330... Training loss: 1.6962... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7331... Training loss: 1.6773... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7332... Training loss: 1.6065... 0.0857 sec/batch
Epoch: 2/20... Training Step: 7333... Training loss: 1.7074... 0.0802 sec/batch
Epoch: 2/20... Training Step: 7334... Training loss: 1.5988... 0.0776 sec/batch
Epoch: 2/20... Training Step: 7335... Training loss: 1.7222... 0.0865 sec/batch
Epoch: 2/20... Training Step: 7336... Training loss: 1.7120... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7337... Training loss: 1.7310... 0.0703 sec/batch
Epoch: 2/20... Training Step: 7338... Training loss: 1.8516... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7339... Training loss: 1.7300... 0.0694 sec/batch
Epoch: 2/20... Training Step: 7340... Training loss: 1.6532... 0.0813 sec/batch
Epoch: 2/20... Training Step: 7341... Training loss: 1.7596... 0.0776 sec/batch
Epoch: 2/20... Training Step: 7342... Training loss: 1.8440... 0.0846 sec/batch
Epoch: 2/20... Training Step: 7343... Training loss: 1.6795... 0.0859 sec/batch
Epoch: 2/20... Training Step: 7344... Training loss: 1.6902... 0.0830 sec/batch
Epoch: 2/20... Training Step: 7345... Training loss: 1.7394... 0.0892 sec/batch
Epoch: 2/20... Training Step: 7346... Training loss: 1.7179... 0.0832 sec/batch
Epoch: 2/20... Training Step: 7347... Training loss: 1.8223... 0.0888 sec/batch
Epoch: 2/20... Training Step: 7348... Training loss: 1.5674... 0.0829 sec/batch
Epoch: 2/20... Training Step: 7349... Training loss: 1.5176... 0.0758 sec/batch
Epoch: 2/20... Training Step: 7350... Training loss: 1.5946... 0.0804 sec/batch
Epoch: 2/20... Training Step: 7351... Training loss: 1.6221... 0.0719 sec/batch
Epoch: 2/20... Training Step: 7352... Training loss: 1.6586... 0.0771 sec/batch
Epoch: 2/20... Training Step: 7353... Training loss: 1.6200... 0.0728 sec/batch
Epoch: 2/20... Training Step: 7354... Training loss: 1.7287... 0.0923 sec/batch
Epoch: 2/20... Training Step: 7355... Training loss: 1.7104... 0.0857 sec/batch
Epoch: 2/20... Training Step: 7356... Training loss: 1.5823... 0.0856 sec/batch
Epoch: 2/20... Training Step: 7357... Training loss: 1.7502... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7358... Training loss: 1.7958... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7359... Training loss: 1.6953... 0.0934 sec/batch
Epoch: 2/20... Training Step: 7360... Training loss: 1.7264... 0.0621 sec/batch
Epoch: 2/20... Training Step: 7361... Training loss: 1.7635... 0.0844 sec/batch
Epoch: 2/20... Training Step: 7362... Training loss: 1.6957... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7363... Training loss: 1.6475... 0.0697 sec/batch
Epoch: 2/20... Training Step: 7364... Training loss: 1.7049... 0.0694 sec/batch
Epoch: 2/20... Training Step: 7365... Training loss: 1.5731... 0.0650 sec/batch
Epoch: 2/20... Training Step: 7366... Training loss: 1.8362... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7367... Training loss: 1.7140... 0.0754 sec/batch
Epoch: 2/20... Training Step: 7368... Training loss: 1.7589... 0.0751 sec/batch
Epoch: 2/20... Training Step: 7369... Training loss: 1.7960... 0.0725 sec/batch
Epoch: 2/20... Training Step: 7370... Training loss: 1.7758... 0.0754 sec/batch
Epoch: 2/20... Training Step: 7371... Training loss: 1.6569... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7372... Training loss: 1.7546... 0.0780 sec/batch
Epoch: 2/20... Training Step: 7373... Training loss: 1.5980... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7374... Training loss: 1.6501... 0.0677 sec/batch
Epoch: 2/20... Training Step: 7375... Training loss: 1.6891... 0.0702 sec/batch
Epoch: 2/20... Training Step: 7376... Training loss: 1.6541... 0.0742 sec/batch
Epoch: 2/20... Training Step: 7377... Training loss: 1.4789... 0.0696 sec/batch
Epoch: 2/20... Training Step: 7378... Training loss: 1.6417... 0.0849 sec/batch
Epoch: 2/20... Training Step: 7379... Training loss: 1.6886... 0.0753 sec/batch
Epoch: 2/20... Training Step: 7380... Training loss: 1.7760... 0.0876 sec/batch
Epoch: 2/20... Training Step: 7381... Training loss: 1.6685... 0.0852 sec/batch
Epoch: 2/20... Training Step: 7382... Training loss: 1.5770... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7383... Training loss: 1.5887... 0.0822 sec/batch
Epoch: 2/20... Training Step: 7384... Training loss: 1.6088... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7385... Training loss: 1.7065... 0.0631 sec/batch
Epoch: 2/20... Training Step: 7386... Training loss: 1.6428... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7387... Training loss: 1.8388... 0.0815 sec/batch
Epoch: 2/20... Training Step: 7388... Training loss: 1.8462... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7389... Training loss: 1.7307... 0.0716 sec/batch
Epoch: 2/20... Training Step: 7390... Training loss: 1.8090... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7391... Training loss: 1.7398... 0.0712 sec/batch
Epoch: 2/20... Training Step: 7392... Training loss: 1.7186... 0.0635 sec/batch
Epoch: 2/20... Training Step: 7393... Training loss: 1.7998... 0.0635 sec/batch
Epoch: 2/20... Training Step: 7394... Training loss: 1.7432... 0.0744 sec/batch
Epoch: 2/20... Training Step: 7395... Training loss: 1.8076... 0.0711 sec/batch
Epoch: 2/20... Training Step: 7396... Training loss: 1.6564... 0.0831 sec/batch
Epoch: 2/20... Training Step: 7397... Training loss: 1.6499... 0.0675 sec/batch
Epoch: 2/20... Training Step: 7398... Training loss: 1.8836... 0.0880 sec/batch
Epoch: 2/20... Training Step: 7399... Training loss: 1.7469... 0.0706 sec/batch
Epoch: 2/20... Training Step: 7400... Training loss: 1.6163... 0.0636 sec/batch
Epoch: 2/20... Training Step: 7401... Training loss: 1.7277... 0.0668 sec/batch
Epoch: 2/20... Training Step: 7402... Training loss: 1.8124... 0.0651 sec/batch
Epoch: 2/20... Training Step: 7403... Training loss: 1.6789... 0.0896 sec/batch
Epoch: 2/20... Training Step: 7404... Training loss: 1.7110... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7405... Training loss: 1.6585... 0.0775 sec/batch
Epoch: 2/20... Training Step: 7406... Training loss: 1.7249... 0.0721 sec/batch
Epoch: 2/20... Training Step: 7407... Training loss: 1.7044... 0.0760 sec/batch
Epoch: 2/20... Training Step: 7408... Training loss: 1.7448... 0.0703 sec/batch
Epoch: 2/20... Training Step: 7409... Training loss: 1.8440... 0.0788 sec/batch
Epoch: 2/20... Training Step: 7410... Training loss: 1.7663... 0.0821 sec/batch
Epoch: 2/20... Training Step: 7411... Training loss: 1.7136... 0.0794 sec/batch
Epoch: 2/20... Training Step: 7412... Training loss: 1.6767... 0.0810 sec/batch
Epoch: 2/20... Training Step: 7413... Training loss: 1.7032... 0.0575 sec/batch
Epoch: 2/20... Training Step: 7414... Training loss: 1.6946... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7415... Training loss: 1.7213... 0.0700 sec/batch
Epoch: 2/20... Training Step: 7416... Training loss: 1.7550... 0.0752 sec/batch
Epoch: 2/20... Training Step: 7417... Training loss: 1.7215... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7418... Training loss: 1.6301... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7419... Training loss: 1.6523... 0.0654 sec/batch
Epoch: 2/20... Training Step: 7420... Training loss: 1.5912... 0.0587 sec/batch
Epoch: 2/20... Training Step: 7421... Training loss: 1.7967... 0.0680 sec/batch
Epoch: 2/20... Training Step: 7422... Training loss: 1.7708... 0.0560 sec/batch
Epoch: 2/20... Training Step: 7423... Training loss: 1.7420... 0.0544 sec/batch
Epoch: 2/20... Training Step: 7424... Training loss: 1.5667... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7425... Training loss: 1.6321... 0.0745 sec/batch
Epoch: 2/20... Training Step: 7426... Training loss: 1.8159... 0.0804 sec/batch
Epoch: 2/20... Training Step: 7427... Training loss: 1.6441... 0.0623 sec/batch
Epoch: 2/20... Training Step: 7428... Training loss: 1.6378... 0.0669 sec/batch
Epoch: 2/20... Training Step: 7429... Training loss: 1.6479... 0.0619 sec/batch
Epoch: 2/20... Training Step: 7430... Training loss: 1.6620... 0.0615 sec/batch
Epoch: 2/20... Training Step: 7431... Training loss: 1.7253... 0.0643 sec/batch
Epoch: 2/20... Training Step: 7432... Training loss: 1.6822... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7433... Training loss: 1.8568... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7434... Training loss: 1.6513... 0.0680 sec/batch
Epoch: 2/20... Training Step: 7435... Training loss: 1.5966... 0.0603 sec/batch
Epoch: 2/20... Training Step: 7436... Training loss: 1.5369... 0.0587 sec/batch
Epoch: 2/20... Training Step: 7437... Training loss: 1.5106... 0.0598 sec/batch
Epoch: 2/20... Training Step: 7438... Training loss: 1.6272... 0.0752 sec/batch
Epoch: 2/20... Training Step: 7439... Training loss: 1.8491... 0.0728 sec/batch
Epoch: 2/20... Training Step: 7440... Training loss: 1.7271... 0.0585 sec/batch
Epoch: 2/20... Training Step: 7441... Training loss: 1.7265... 0.0670 sec/batch
Epoch: 2/20... Training Step: 7442... Training loss: 1.6177... 0.0696 sec/batch
Epoch: 2/20... Training Step: 7443... Training loss: 1.8198... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7444... Training loss: 1.7145... 0.0692 sec/batch
Epoch: 2/20... Training Step: 7445... Training loss: 1.8192... 0.0646 sec/batch
Epoch: 2/20... Training Step: 7446... Training loss: 1.5861... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7447... Training loss: 1.8203... 0.0626 sec/batch
Epoch: 2/20... Training Step: 7448... Training loss: 1.7277... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7449... Training loss: 1.7738... 0.0669 sec/batch
Epoch: 2/20... Training Step: 7450... Training loss: 1.5261... 0.0819 sec/batch
Epoch: 2/20... Training Step: 7451... Training loss: 1.7335... 0.0747 sec/batch
Epoch: 2/20... Training Step: 7452... Training loss: 1.7303... 0.0754 sec/batch
Epoch: 2/20... Training Step: 7453... Training loss: 1.8485... 0.0618 sec/batch
Epoch: 2/20... Training Step: 7454... Training loss: 1.8405... 0.0774 sec/batch
Epoch: 2/20... Training Step: 7455... Training loss: 1.7426... 0.0706 sec/batch
Epoch: 2/20... Training Step: 7456... Training loss: 1.9599... 0.0612 sec/batch
Epoch: 2/20... Training Step: 7457... Training loss: 1.7403... 0.0581 sec/batch
Epoch: 2/20... Training Step: 7458... Training loss: 1.8876... 0.0707 sec/batch
Epoch: 2/20... Training Step: 7459... Training loss: 1.6823... 0.0715 sec/batch
Epoch: 2/20... Training Step: 7460... Training loss: 1.8907... 0.0680 sec/batch
Epoch: 2/20... Training Step: 7461... Training loss: 1.8559... 0.0718 sec/batch
Epoch: 2/20... Training Step: 7462... Training loss: 1.7046... 0.0625 sec/batch
Epoch: 2/20... Training Step: 7463... Training loss: 1.6537... 0.0711 sec/batch
Epoch: 2/20... Training Step: 7464... Training loss: 1.7829... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7465... Training loss: 1.8097... 0.0802 sec/batch
Epoch: 2/20... Training Step: 7466... Training loss: 1.7010... 0.0775 sec/batch
Epoch: 2/20... Training Step: 7467... Training loss: 1.7227... 0.0763 sec/batch
Epoch: 2/20... Training Step: 7468... Training loss: 1.7176... 0.0618 sec/batch
Epoch: 2/20... Training Step: 7469... Training loss: 1.7667... 0.0797 sec/batch
Epoch: 2/20... Training Step: 7470... Training loss: 1.7932... 0.0814 sec/batch
Epoch: 2/20... Training Step: 7471... Training loss: 1.6421... 0.0672 sec/batch
Epoch: 2/20... Training Step: 7472... Training loss: 1.7146... 0.0725 sec/batch
Epoch: 2/20... Training Step: 7473... Training loss: 1.6120... 0.0728 sec/batch
Epoch: 2/20... Training Step: 7474... Training loss: 1.6942... 0.0856 sec/batch
Epoch: 2/20... Training Step: 7475... Training loss: 1.7237... 0.0631 sec/batch
Epoch: 2/20... Training Step: 7476... Training loss: 1.6890... 0.0611 sec/batch
Epoch: 2/20... Training Step: 7477... Training loss: 1.6975... 0.0877 sec/batch
Epoch: 2/20... Training Step: 7478... Training loss: 1.7037... 0.0754 sec/batch
Epoch: 2/20... Training Step: 7479... Training loss: 1.6177... 0.0785 sec/batch
Epoch: 2/20... Training Step: 7480... Training loss: 1.6696... 0.0589 sec/batch
Epoch: 2/20... Training Step: 7481... Training loss: 1.7936... 0.0631 sec/batch
Epoch: 2/20... Training Step: 7482... Training loss: 1.6516... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7483... Training loss: 1.7035... 0.0660 sec/batch
Epoch: 2/20... Training Step: 7484... Training loss: 1.6021... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7485... Training loss: 1.5984... 0.0793 sec/batch
Epoch: 2/20... Training Step: 7486... Training loss: 1.7121... 0.0881 sec/batch
Epoch: 2/20... Training Step: 7487... Training loss: 1.6057... 0.0812 sec/batch
Epoch: 2/20... Training Step: 7488... Training loss: 1.7476... 0.0785 sec/batch
Epoch: 2/20... Training Step: 7489... Training loss: 1.8403... 0.0709 sec/batch
Epoch: 2/20... Training Step: 7490... Training loss: 1.6875... 0.0601 sec/batch
Epoch: 2/20... Training Step: 7491... Training loss: 1.7818... 0.0901 sec/batch
Epoch: 2/20... Training Step: 7492... Training loss: 1.8720... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7493... Training loss: 1.7346... 0.0627 sec/batch
Epoch: 2/20... Training Step: 7494... Training loss: 1.6793... 0.0761 sec/batch
Epoch: 2/20... Training Step: 7495... Training loss: 1.4161... 0.0644 sec/batch
Epoch: 2/20... Training Step: 7496... Training loss: 1.7229... 0.0712 sec/batch
Epoch: 2/20... Training Step: 7497... Training loss: 1.6332... 0.0705 sec/batch
Epoch: 2/20... Training Step: 7498... Training loss: 1.7143... 0.0879 sec/batch
Epoch: 2/20... Training Step: 7499... Training loss: 1.6460... 0.0729 sec/batch
Epoch: 2/20... Training Step: 7500... Training loss: 1.6296... 0.0626 sec/batch
Epoch: 2/20... Training Step: 7501... Training loss: 1.5645... 0.0916 sec/batch
Epoch: 2/20... Training Step: 7502... Training loss: 1.7543... 0.0804 sec/batch
Epoch: 2/20... Training Step: 7503... Training loss: 1.7178... 0.0749 sec/batch
Epoch: 2/20... Training Step: 7504... Training loss: 1.7350... 0.0842 sec/batch
Epoch: 2/20... Training Step: 7505... Training loss: 1.8394... 0.0826 sec/batch
Epoch: 2/20... Training Step: 7506... Training loss: 1.6152... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7507... Training loss: 1.7153... 0.0799 sec/batch
Epoch: 2/20... Training Step: 7508... Training loss: 1.7668... 0.0862 sec/batch
Epoch: 2/20... Training Step: 7509... Training loss: 1.7406... 0.0738 sec/batch
Epoch: 2/20... Training Step: 7510... Training loss: 1.6640... 0.0682 sec/batch
Epoch: 2/20... Training Step: 7511... Training loss: 1.8358... 0.0800 sec/batch
Epoch: 2/20... Training Step: 7512... Training loss: 1.7198... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7513... Training loss: 1.5628... 0.0800 sec/batch
Epoch: 2/20... Training Step: 7514... Training loss: 1.6357... 0.0829 sec/batch
Epoch: 2/20... Training Step: 7515... Training loss: 1.7003... 0.0814 sec/batch
Epoch: 2/20... Training Step: 7516... Training loss: 1.5699... 0.0901 sec/batch
Epoch: 2/20... Training Step: 7517... Training loss: 1.6746... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7518... Training loss: 1.5870... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7519... Training loss: 1.6506... 0.0654 sec/batch
Epoch: 2/20... Training Step: 7520... Training loss: 1.7887... 0.0648 sec/batch
Epoch: 2/20... Training Step: 7521... Training loss: 1.6765... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7522... Training loss: 1.7312... 0.0657 sec/batch
Epoch: 2/20... Training Step: 7523... Training loss: 1.8220... 0.0609 sec/batch
Epoch: 2/20... Training Step: 7524... Training loss: 1.7654... 0.0966 sec/batch
Epoch: 2/20... Training Step: 7525... Training loss: 1.5577... 0.0758 sec/batch
Epoch: 2/20... Training Step: 7526... Training loss: 1.7839... 0.0822 sec/batch
Epoch: 2/20... Training Step: 7527... Training loss: 1.6648... 0.0712 sec/batch
Epoch: 2/20... Training Step: 7528... Training loss: 1.5548... 0.0786 sec/batch
Epoch: 2/20... Training Step: 7529... Training loss: 1.7291... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7530... Training loss: 1.7159... 0.0633 sec/batch
Epoch: 2/20... Training Step: 7531... Training loss: 1.6731... 0.0663 sec/batch
Epoch: 2/20... Training Step: 7532... Training loss: 1.6427... 0.0719 sec/batch
Epoch: 2/20... Training Step: 7533... Training loss: 1.7230... 0.0766 sec/batch
Epoch: 2/20... Training Step: 7534... Training loss: 1.6887... 0.0692 sec/batch
Epoch: 2/20... Training Step: 7535... Training loss: 1.7391... 0.0621 sec/batch
Epoch: 2/20... Training Step: 7536... Training loss: 1.7542... 0.0668 sec/batch
Epoch: 2/20... Training Step: 7537... Training loss: 1.7903... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7538... Training loss: 1.7285... 0.0624 sec/batch
Epoch: 2/20... Training Step: 7539... Training loss: 1.7141... 0.0737 sec/batch
Epoch: 2/20... Training Step: 7540... Training loss: 1.5196... 0.0751 sec/batch
Epoch: 2/20... Training Step: 7541... Training loss: 1.6927... 0.0679 sec/batch
Epoch: 2/20... Training Step: 7542... Training loss: 1.9161... 0.0833 sec/batch
Epoch: 2/20... Training Step: 7543... Training loss: 1.7344... 0.0648 sec/batch
Epoch: 2/20... Training Step: 7544... Training loss: 1.6963... 0.0624 sec/batch
Epoch: 2/20... Training Step: 7545... Training loss: 1.5174... 0.0758 sec/batch
Epoch: 2/20... Training Step: 7546... Training loss: 1.6208... 0.0891 sec/batch
Epoch: 2/20... Training Step: 7547... Training loss: 1.6423... 0.0799 sec/batch
Epoch: 2/20... Training Step: 7548... Training loss: 1.6388... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7549... Training loss: 1.6253... 0.0735 sec/batch
Epoch: 2/20... Training Step: 7550... Training loss: 1.5974... 0.0846 sec/batch
Epoch: 2/20... Training Step: 7551... Training loss: 1.6850... 0.0833 sec/batch
Epoch: 2/20... Training Step: 7552... Training loss: 1.6749... 0.0593 sec/batch
Epoch: 2/20... Training Step: 7553... Training loss: 1.6108... 0.0618 sec/batch
Epoch: 2/20... Training Step: 7554... Training loss: 1.8919... 0.0777 sec/batch
Epoch: 2/20... Training Step: 7555... Training loss: 1.7320... 0.0770 sec/batch
Epoch: 2/20... Training Step: 7556... Training loss: 1.8170... 0.0677 sec/batch
Epoch: 2/20... Training Step: 7557... Training loss: 1.7924... 0.0734 sec/batch
Epoch: 2/20... Training Step: 7558... Training loss: 1.6817... 0.0652 sec/batch
Epoch: 2/20... Training Step: 7559... Training loss: 1.6294... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7560... Training loss: 1.6535... 0.0734 sec/batch
Epoch: 2/20... Training Step: 7561... Training loss: 1.6440... 0.0757 sec/batch
Epoch: 2/20... Training Step: 7562... Training loss: 1.6012... 0.0666 sec/batch
Epoch: 2/20... Training Step: 7563... Training loss: 1.6987... 0.0642 sec/batch
Epoch: 2/20... Training Step: 7564... Training loss: 2.4337... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7565... Training loss: 1.8949... 0.0577 sec/batch
Epoch: 2/20... Training Step: 7566... Training loss: 1.7488... 0.0590 sec/batch
Epoch: 2/20... Training Step: 7567... Training loss: 1.8608... 0.0786 sec/batch
Epoch: 2/20... Training Step: 7568... Training loss: 1.7819... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7569... Training loss: 1.8203... 0.0608 sec/batch
Epoch: 2/20... Training Step: 7570... Training loss: 1.7333... 0.0756 sec/batch
Epoch: 2/20... Training Step: 7571... Training loss: 1.7962... 0.0715 sec/batch
Epoch: 2/20... Training Step: 7572... Training loss: 1.7102... 0.0814 sec/batch
Epoch: 2/20... Training Step: 7573... Training loss: 1.8115... 0.0653 sec/batch
Epoch: 2/20... Training Step: 7574... Training loss: 1.7795... 0.0626 sec/batch
Epoch: 2/20... Training Step: 7575... Training loss: 1.7789... 0.0877 sec/batch
Epoch: 2/20... Training Step: 7576... Training loss: 1.6401... 0.0834 sec/batch
Epoch: 2/20... Training Step: 7577... Training loss: 1.7522... 0.0771 sec/batch
Epoch: 2/20... Training Step: 7578... Training loss: 1.7071... 0.0793 sec/batch
Epoch: 2/20... Training Step: 7579... Training loss: 1.8479... 0.0684 sec/batch
Epoch: 2/20... Training Step: 7580... Training loss: 1.8205... 0.0640 sec/batch
Epoch: 2/20... Training Step: 7581... Training loss: 1.7488... 0.0962 sec/batch
Epoch: 2/20... Training Step: 7582... Training loss: 1.6827... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7583... Training loss: 1.7806... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7584... Training loss: 1.6682... 0.0746 sec/batch
Epoch: 2/20... Training Step: 7585... Training loss: 1.7511... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7586... Training loss: 1.7613... 0.0692 sec/batch
Epoch: 2/20... Training Step: 7587... Training loss: 1.8002... 0.0831 sec/batch
Epoch: 2/20... Training Step: 7588... Training loss: 1.7810... 0.0748 sec/batch
Epoch: 2/20... Training Step: 7589... Training loss: 1.7391... 0.0716 sec/batch
Epoch: 2/20... Training Step: 7590... Training loss: 1.6887... 0.0709 sec/batch
Epoch: 2/20... Training Step: 7591... Training loss: 1.6496... 0.0640 sec/batch
Epoch: 2/20... Training Step: 7592... Training loss: 1.7632... 0.0697 sec/batch
Epoch: 2/20... Training Step: 7593... Training loss: 1.7051... 0.0667 sec/batch
Epoch: 2/20... Training Step: 7594... Training loss: 1.8818... 0.0647 sec/batch
Epoch: 2/20... Training Step: 7595... Training loss: 1.6854... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7596... Training loss: 1.6974... 0.0725 sec/batch
Epoch: 2/20... Training Step: 7597... Training loss: 1.9641... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7598... Training loss: 1.6768... 0.1010 sec/batch
Epoch: 2/20... Training Step: 7599... Training loss: 1.8204... 0.0686 sec/batch
Epoch: 2/20... Training Step: 7600... Training loss: 1.7061... 0.0576 sec/batch
Epoch: 2/20... Training Step: 7601... Training loss: 1.8217... 0.0729 sec/batch
Epoch: 2/20... Training Step: 7602... Training loss: 1.8747... 0.0784 sec/batch
Epoch: 2/20... Training Step: 7603... Training loss: 1.8672... 0.0662 sec/batch
Epoch: 2/20... Training Step: 7604... Training loss: 1.8147... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7605... Training loss: 1.7472... 0.0670 sec/batch
Epoch: 2/20... Training Step: 7606... Training loss: 1.8210... 0.0596 sec/batch
Epoch: 2/20... Training Step: 7607... Training loss: 1.6680... 0.0676 sec/batch
Epoch: 2/20... Training Step: 7608... Training loss: 1.8010... 0.0666 sec/batch
Epoch: 2/20... Training Step: 7609... Training loss: 1.7759... 0.0621 sec/batch
Epoch: 2/20... Training Step: 7610... Training loss: 1.6640... 0.0720 sec/batch
Epoch: 2/20... Training Step: 7611... Training loss: 1.8235... 0.0708 sec/batch
Epoch: 2/20... Training Step: 7612... Training loss: 1.7895... 0.0840 sec/batch
Epoch: 2/20... Training Step: 7613... Training loss: 1.9451... 0.0788 sec/batch
Epoch: 2/20... Training Step: 7614... Training loss: 1.8544... 0.0774 sec/batch
Epoch: 2/20... Training Step: 7615... Training loss: 1.7732... 0.0755 sec/batch
Epoch: 2/20... Training Step: 7616... Training loss: 1.5888... 0.0842 sec/batch
Epoch: 2/20... Training Step: 7617... Training loss: 1.7370... 0.0831 sec/batch
Epoch: 2/20... Training Step: 7618... Training loss: 1.7758... 0.0857 sec/batch
Epoch: 2/20... Training Step: 7619... Training loss: 1.6667... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7620... Training loss: 1.9163... 0.0770 sec/batch
Epoch: 2/20... Training Step: 7621... Training loss: 1.8219... 0.0805 sec/batch
Epoch: 2/20... Training Step: 7622... Training loss: 1.6759... 0.0771 sec/batch
Epoch: 2/20... Training Step: 7623... Training loss: 1.7247... 0.0742 sec/batch
Epoch: 2/20... Training Step: 7624... Training loss: 1.7459... 0.0918 sec/batch
Epoch: 2/20... Training Step: 7625... Training loss: 1.9162... 0.0793 sec/batch
Epoch: 2/20... Training Step: 7626... Training loss: 2.0122... 0.0758 sec/batch
Epoch: 2/20... Training Step: 7627... Training loss: 1.7148... 0.0805 sec/batch
Epoch: 2/20... Training Step: 7628... Training loss: 1.8025... 0.0766 sec/batch
Epoch: 2/20... Training Step: 7629... Training loss: 1.7229... 0.0886 sec/batch
Epoch: 2/20... Training Step: 7630... Training loss: 1.7736... 0.0653 sec/batch
Epoch: 2/20... Training Step: 7631... Training loss: 1.7947... 0.0708 sec/batch
Epoch: 2/20... Training Step: 7632... Training loss: 1.7276... 0.0614 sec/batch
Epoch: 2/20... Training Step: 7633... Training loss: 1.7554... 0.0679 sec/batch
Epoch: 2/20... Training Step: 7634... Training loss: 1.9191... 0.0802 sec/batch
Epoch: 2/20... Training Step: 7635... Training loss: 1.7069... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7636... Training loss: 1.8189... 0.0688 sec/batch
Epoch: 2/20... Training Step: 7637... Training loss: 1.6292... 0.0642 sec/batch
Epoch: 2/20... Training Step: 7638... Training loss: 1.8408... 0.0662 sec/batch
Epoch: 2/20... Training Step: 7639... Training loss: 1.8231... 0.0714 sec/batch
Epoch: 2/20... Training Step: 7640... Training loss: 1.7871... 0.0611 sec/batch
Epoch: 2/20... Training Step: 7641... Training loss: 1.6855... 0.0589 sec/batch
Epoch: 2/20... Training Step: 7642... Training loss: 1.8051... 0.0582 sec/batch
Epoch: 2/20... Training Step: 7643... Training loss: 1.7232... 0.0625 sec/batch
Epoch: 2/20... Training Step: 7644... Training loss: 1.7620... 0.0847 sec/batch
Epoch: 2/20... Training Step: 7645... Training loss: 1.6307... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7646... Training loss: 1.9207... 0.0730 sec/batch
Epoch: 2/20... Training Step: 7647... Training loss: 1.7879... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7648... Training loss: 1.7417... 0.0586 sec/batch
Epoch: 2/20... Training Step: 7649... Training loss: 1.8229... 0.0857 sec/batch
Epoch: 2/20... Training Step: 7650... Training loss: 1.8837... 0.0600 sec/batch
Epoch: 2/20... Training Step: 7651... Training loss: 1.7785... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7652... Training loss: 1.8093... 0.0622 sec/batch
Epoch: 2/20... Training Step: 7653... Training loss: 1.6690... 0.0703 sec/batch
Epoch: 2/20... Training Step: 7654... Training loss: 1.6490... 0.0742 sec/batch
Epoch: 2/20... Training Step: 7655... Training loss: 1.6962... 0.0660 sec/batch
Epoch: 2/20... Training Step: 7656... Training loss: 1.8251... 0.0616 sec/batch
Epoch: 2/20... Training Step: 7657... Training loss: 1.7985... 0.0678 sec/batch
Epoch: 2/20... Training Step: 7658... Training loss: 1.7759... 0.0573 sec/batch
Epoch: 2/20... Training Step: 7659... Training loss: 1.7155... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7660... Training loss: 1.7270... 0.0781 sec/batch
Epoch: 2/20... Training Step: 7661... Training loss: 1.8702... 0.0741 sec/batch
Epoch: 2/20... Training Step: 7662... Training loss: 1.7736... 0.0726 sec/batch
Epoch: 2/20... Training Step: 7663... Training loss: 1.9304... 0.0609 sec/batch
Epoch: 2/20... Training Step: 7664... Training loss: 1.8165... 0.0788 sec/batch
Epoch: 2/20... Training Step: 7665... Training loss: 1.8072... 0.0721 sec/batch
Epoch: 2/20... Training Step: 7666... Training loss: 1.7990... 0.0658 sec/batch
Epoch: 2/20... Training Step: 7667... Training loss: 1.7407... 0.0736 sec/batch
Epoch: 2/20... Training Step: 7668... Training loss: 1.7883... 0.0755 sec/batch
Epoch: 2/20... Training Step: 7669... Training loss: 1.7438... 0.0726 sec/batch
Epoch: 2/20... Training Step: 7670... Training loss: 1.6673... 0.0736 sec/batch
Epoch: 2/20... Training Step: 7671... Training loss: 1.8867... 0.0748 sec/batch
Epoch: 2/20... Training Step: 7672... Training loss: 1.8176... 0.0747 sec/batch
Epoch: 2/20... Training Step: 7673... Training loss: 1.6726... 0.0587 sec/batch
Epoch: 2/20... Training Step: 7674... Training loss: 1.9052... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7675... Training loss: 1.7141... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7676... Training loss: 1.6714... 0.0679 sec/batch
Epoch: 2/20... Training Step: 7677... Training loss: 1.6824... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7678... Training loss: 1.8161... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7679... Training loss: 1.7581... 0.0757 sec/batch
Epoch: 2/20... Training Step: 7680... Training loss: 1.7163... 0.0741 sec/batch
Epoch: 2/20... Training Step: 7681... Training loss: 1.7330... 0.0772 sec/batch
Epoch: 2/20... Training Step: 7682... Training loss: 1.8771... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7683... Training loss: 1.7270... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7684... Training loss: 1.8104... 0.0913 sec/batch
Epoch: 2/20... Training Step: 7685... Training loss: 2.0817... 0.0624 sec/batch
Epoch: 2/20... Training Step: 7686... Training loss: 1.7239... 0.0675 sec/batch
Epoch: 2/20... Training Step: 7687... Training loss: 1.7558... 0.0707 sec/batch
Epoch: 2/20... Training Step: 7688... Training loss: 1.6978... 0.0652 sec/batch
Epoch: 2/20... Training Step: 7689... Training loss: 1.7436... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7690... Training loss: 1.6822... 0.0644 sec/batch
Epoch: 2/20... Training Step: 7691... Training loss: 1.6113... 0.0941 sec/batch
Epoch: 2/20... Training Step: 7692... Training loss: 1.5928... 0.0644 sec/batch
Epoch: 2/20... Training Step: 7693... Training loss: 1.9071... 0.0706 sec/batch
Epoch: 2/20... Training Step: 7694... Training loss: 1.8916... 0.0735 sec/batch
Epoch: 2/20... Training Step: 7695... Training loss: 1.7793... 0.0729 sec/batch
Epoch: 2/20... Training Step: 7696... Training loss: 1.8454... 0.0721 sec/batch
Epoch: 2/20... Training Step: 7697... Training loss: 1.8877... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7698... Training loss: 1.8728... 0.0680 sec/batch
Epoch: 2/20... Training Step: 7699... Training loss: 1.8709... 0.0641 sec/batch
Epoch: 2/20... Training Step: 7700... Training loss: 1.8423... 0.0684 sec/batch
Epoch: 2/20... Training Step: 7701... Training loss: 1.7223... 0.0843 sec/batch
Epoch: 2/20... Training Step: 7702... Training loss: 1.8555... 0.0737 sec/batch
Epoch: 2/20... Training Step: 7703... Training loss: 1.6556... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7704... Training loss: 1.6761... 0.0818 sec/batch
Epoch: 2/20... Training Step: 7705... Training loss: 1.8388... 0.0757 sec/batch
Epoch: 2/20... Training Step: 7706... Training loss: 1.7453... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7707... Training loss: 1.6730... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7708... Training loss: 1.9761... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7709... Training loss: 1.9403... 0.0810 sec/batch
Epoch: 2/20... Training Step: 7710... Training loss: 1.7426... 0.0809 sec/batch
Epoch: 2/20... Training Step: 7711... Training loss: 1.8581... 0.0776 sec/batch
Epoch: 2/20... Training Step: 7712... Training loss: 1.8945... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7713... Training loss: 1.8145... 0.0793 sec/batch
Epoch: 2/20... Training Step: 7714... Training loss: 1.9040... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7715... Training loss: 1.6617... 0.0810 sec/batch
Epoch: 2/20... Training Step: 7716... Training loss: 1.9665... 0.0769 sec/batch
Epoch: 2/20... Training Step: 7717... Training loss: 1.7780... 0.0867 sec/batch
Epoch: 2/20... Training Step: 7718... Training loss: 2.0213... 0.0655 sec/batch
Epoch: 2/20... Training Step: 7719... Training loss: 1.7026... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7720... Training loss: 1.8133... 0.0567 sec/batch
Epoch: 2/20... Training Step: 7721... Training loss: 1.6706... 0.0838 sec/batch
Epoch: 2/20... Training Step: 7722... Training loss: 1.9809... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7723... Training loss: 1.7382... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7724... Training loss: 1.7766... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7725... Training loss: 1.7501... 0.0767 sec/batch
Epoch: 2/20... Training Step: 7726... Training loss: 1.8895... 0.0797 sec/batch
Epoch: 2/20... Training Step: 7727... Training loss: 1.7669... 0.0875 sec/batch
Epoch: 2/20... Training Step: 7728... Training loss: 1.9781... 0.0762 sec/batch
Epoch: 2/20... Training Step: 7729... Training loss: 1.8101... 0.0729 sec/batch
Epoch: 2/20... Training Step: 7730... Training loss: 1.6661... 0.0685 sec/batch
Epoch: 2/20... Training Step: 7731... Training loss: 1.7066... 0.0798 sec/batch
Epoch: 2/20... Training Step: 7732... Training loss: 1.7033... 0.0760 sec/batch
Epoch: 2/20... Training Step: 7733... Training loss: 1.6975... 0.0784 sec/batch
Epoch: 2/20... Training Step: 7734... Training loss: 1.7704... 0.0686 sec/batch
Epoch: 2/20... Training Step: 7735... Training loss: 1.6758... 0.0810 sec/batch
Epoch: 2/20... Training Step: 7736... Training loss: 1.7670... 0.0959 sec/batch
Epoch: 2/20... Training Step: 7737... Training loss: 1.6557... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7738... Training loss: 1.8149... 0.0725 sec/batch
Epoch: 2/20... Training Step: 7739... Training loss: 1.9441... 0.0744 sec/batch
Epoch: 2/20... Training Step: 7740... Training loss: 1.7126... 0.0720 sec/batch
Epoch: 2/20... Training Step: 7741... Training loss: 1.6963... 0.0650 sec/batch
Epoch: 2/20... Training Step: 7742... Training loss: 1.6603... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7743... Training loss: 1.8967... 0.0660 sec/batch
Epoch: 2/20... Training Step: 7744... Training loss: 1.7681... 0.0742 sec/batch
Epoch: 2/20... Training Step: 7745... Training loss: 1.7879... 0.0841 sec/batch
Epoch: 2/20... Training Step: 7746... Training loss: 1.7323... 0.0765 sec/batch
Epoch: 2/20... Training Step: 7747... Training loss: 1.7658... 0.0628 sec/batch
Epoch: 2/20... Training Step: 7748... Training loss: 1.7749... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7749... Training loss: 1.6977... 0.0797 sec/batch
Epoch: 2/20... Training Step: 7750... Training loss: 1.9240... 0.0580 sec/batch
Epoch: 2/20... Training Step: 7751... Training loss: 1.7169... 0.0735 sec/batch
Epoch: 2/20... Training Step: 7752... Training loss: 1.6363... 0.0687 sec/batch
Epoch: 2/20... Training Step: 7753... Training loss: 1.6724... 0.0859 sec/batch
Epoch: 2/20... Training Step: 7754... Training loss: 1.7635... 0.0841 sec/batch
Epoch: 2/20... Training Step: 7755... Training loss: 1.6668... 0.0809 sec/batch
Epoch: 2/20... Training Step: 7756... Training loss: 1.6238... 0.0628 sec/batch
Epoch: 2/20... Training Step: 7757... Training loss: 1.7521... 0.0749 sec/batch
Epoch: 2/20... Training Step: 7758... Training loss: 1.7093... 0.0697 sec/batch
Epoch: 2/20... Training Step: 7759... Training loss: 1.6041... 0.0630 sec/batch
Epoch: 2/20... Training Step: 7760... Training loss: 1.7027... 0.0821 sec/batch
Epoch: 2/20... Training Step: 7761... Training loss: 1.5986... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7762... Training loss: 1.7394... 0.0730 sec/batch
Epoch: 2/20... Training Step: 7763... Training loss: 1.7844... 0.0751 sec/batch
Epoch: 2/20... Training Step: 7764... Training loss: 1.9342... 0.0660 sec/batch
Epoch: 2/20... Training Step: 7765... Training loss: 1.9023... 0.0761 sec/batch
Epoch: 2/20... Training Step: 7766... Training loss: 1.9094... 0.0841 sec/batch
Epoch: 2/20... Training Step: 7767... Training loss: 1.8879... 0.0802 sec/batch
Epoch: 2/20... Training Step: 7768... Training loss: 2.0209... 0.0792 sec/batch
Epoch: 2/20... Training Step: 7769... Training loss: 1.8023... 0.0730 sec/batch
Epoch: 2/20... Training Step: 7770... Training loss: 1.6814... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7771... Training loss: 1.8454... 0.0783 sec/batch
Epoch: 2/20... Training Step: 7772... Training loss: 1.6514... 0.0650 sec/batch
Epoch: 2/20... Training Step: 7773... Training loss: 1.6680... 0.0634 sec/batch
Epoch: 2/20... Training Step: 7774... Training loss: 1.6569... 0.0640 sec/batch
Epoch: 2/20... Training Step: 7775... Training loss: 1.7506... 0.0587 sec/batch
Epoch: 2/20... Training Step: 7776... Training loss: 1.6002... 0.0665 sec/batch
Epoch: 2/20... Training Step: 7777... Training loss: 1.8694... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7778... Training loss: 2.0890... 0.0791 sec/batch
Epoch: 2/20... Training Step: 7779... Training loss: 1.7611... 0.0691 sec/batch
Epoch: 2/20... Training Step: 7780... Training loss: 1.8173... 0.0788 sec/batch
Epoch: 2/20... Training Step: 7781... Training loss: 1.6074... 0.0800 sec/batch
Epoch: 2/20... Training Step: 7782... Training loss: 1.8623... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7783... Training loss: 1.5352... 0.0791 sec/batch
Epoch: 2/20... Training Step: 7784... Training loss: 1.6428... 0.0678 sec/batch
Epoch: 2/20... Training Step: 7785... Training loss: 1.8795... 0.0713 sec/batch
Epoch: 2/20... Training Step: 7786... Training loss: 2.1928... 0.0686 sec/batch
Epoch: 2/20... Training Step: 7787... Training loss: 2.2704... 0.0661 sec/batch
Epoch: 2/20... Training Step: 7788... Training loss: 2.1930... 0.0552 sec/batch
Epoch: 2/20... Training Step: 7789... Training loss: 2.0791... 0.0610 sec/batch
Epoch: 2/20... Training Step: 7790... Training loss: 2.0782... 0.0717 sec/batch
Epoch: 2/20... Training Step: 7791... Training loss: 1.9874... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7792... Training loss: 1.9755... 0.0627 sec/batch
Epoch: 2/20... Training Step: 7793... Training loss: 2.0755... 0.0825 sec/batch
Epoch: 2/20... Training Step: 7794... Training loss: 1.8753... 0.0664 sec/batch
Epoch: 2/20... Training Step: 7795... Training loss: 1.7258... 0.0550 sec/batch
Epoch: 2/20... Training Step: 7796... Training loss: 1.7305... 0.0821 sec/batch
Epoch: 2/20... Training Step: 7797... Training loss: 1.6861... 0.0628 sec/batch
Epoch: 2/20... Training Step: 7798... Training loss: 1.8373... 0.0697 sec/batch
Epoch: 2/20... Training Step: 7799... Training loss: 1.7150... 0.0795 sec/batch
Epoch: 2/20... Training Step: 7800... Training loss: 1.8537... 0.0899 sec/batch
Epoch: 2/20... Training Step: 7801... Training loss: 1.6791... 0.0710 sec/batch
Epoch: 2/20... Training Step: 7802... Training loss: 1.5617... 0.0659 sec/batch
Epoch: 2/20... Training Step: 7803... Training loss: 1.6651... 0.0721 sec/batch
Epoch: 2/20... Training Step: 7804... Training loss: 1.5748... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7805... Training loss: 1.8078... 0.0776 sec/batch
Epoch: 2/20... Training Step: 7806... Training loss: 1.6479... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7807... Training loss: 1.8405... 0.0849 sec/batch
Epoch: 2/20... Training Step: 7808... Training loss: 1.7586... 0.0779 sec/batch
Epoch: 2/20... Training Step: 7809... Training loss: 1.8486... 0.0730 sec/batch
Epoch: 2/20... Training Step: 7810... Training loss: 1.8663... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7811... Training loss: 1.8097... 0.0791 sec/batch
Epoch: 2/20... Training Step: 7812... Training loss: 1.8943... 0.0848 sec/batch
Epoch: 2/20... Training Step: 7813... Training loss: 1.9735... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7814... Training loss: 1.9160... 0.0721 sec/batch
Epoch: 2/20... Training Step: 7815... Training loss: 1.8441... 0.0808 sec/batch
Epoch: 2/20... Training Step: 7816... Training loss: 1.7563... 0.0864 sec/batch
Epoch: 2/20... Training Step: 7817... Training loss: 1.7183... 0.0807 sec/batch
Epoch: 2/20... Training Step: 7818... Training loss: 1.8801... 0.0772 sec/batch
Epoch: 2/20... Training Step: 7819... Training loss: 1.6601... 0.0805 sec/batch
Epoch: 2/20... Training Step: 7820... Training loss: 1.6547... 0.0782 sec/batch
Epoch: 2/20... Training Step: 7821... Training loss: 1.8809... 0.0772 sec/batch
Epoch: 2/20... Training Step: 7822... Training loss: 1.8480... 0.0992 sec/batch
Epoch: 2/20... Training Step: 7823... Training loss: 1.9702... 0.0892 sec/batch
Epoch: 2/20... Training Step: 7824... Training loss: 1.6675... 0.0816 sec/batch
Epoch: 2/20... Training Step: 7825... Training loss: 1.9194... 0.0875 sec/batch
Epoch: 2/20... Training Step: 7826... Training loss: 1.6389... 0.0772 sec/batch
Epoch: 2/20... Training Step: 7827... Training loss: 1.8291... 0.0629 sec/batch
Epoch: 2/20... Training Step: 7828... Training loss: 1.6381... 0.0553 sec/batch
Epoch: 2/20... Training Step: 7829... Training loss: 1.8516... 0.0671 sec/batch
Epoch: 2/20... Training Step: 7830... Training loss: 1.7576... 0.0681 sec/batch
Epoch: 2/20... Training Step: 7831... Training loss: 1.8835... 0.0670 sec/batch
Epoch: 2/20... Training Step: 7832... Training loss: 1.6569... 0.0822 sec/batch
Epoch: 2/20... Training Step: 7833... Training loss: 1.7423... 0.0864 sec/batch
Epoch: 2/20... Training Step: 7834... Training loss: 1.7634... 0.0790 sec/batch
Epoch: 2/20... Training Step: 7835... Training loss: 1.7888... 0.0659 sec/batch
Epoch: 2/20... Training Step: 7836... Training loss: 1.7668... 0.0747 sec/batch
Epoch: 2/20... Training Step: 7837... Training loss: 1.7934... 0.0628 sec/batch
Epoch: 2/20... Training Step: 7838... Training loss: 1.6805... 0.0695 sec/batch
Epoch: 2/20... Training Step: 7839... Training loss: 1.7707... 0.0674 sec/batch
Epoch: 2/20... Training Step: 7840... Training loss: 1.9117... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7841... Training loss: 1.7100... 0.0746 sec/batch
Epoch: 2/20... Training Step: 7842... Training loss: 1.7909... 0.0692 sec/batch
Epoch: 2/20... Training Step: 7843... Training loss: 1.8042... 0.0718 sec/batch
Epoch: 2/20... Training Step: 7844... Training loss: 1.6466... 0.0700 sec/batch
Epoch: 2/20... Training Step: 7845... Training loss: 1.6817... 0.0739 sec/batch
Epoch: 2/20... Training Step: 7846... Training loss: 1.7905... 0.0711 sec/batch
Epoch: 2/20... Training Step: 7847... Training loss: 1.8249... 0.0840 sec/batch
Epoch: 2/20... Training Step: 7848... Training loss: 1.7051... 0.0806 sec/batch
Epoch: 2/20... Training Step: 7849... Training loss: 1.7628... 0.0653 sec/batch
Epoch: 2/20... Training Step: 7850... Training loss: 1.7479... 0.1225 sec/batch
Epoch: 2/20... Training Step: 7851... Training loss: 1.8043... 0.1000 sec/batch
Epoch: 2/20... Training Step: 7852... Training loss: 1.8716... 0.1265 sec/batch
Epoch: 2/20... Training Step: 7853... Training loss: 1.7718... 0.0950 sec/batch
Epoch: 2/20... Training Step: 7854... Training loss: 1.7186... 0.1130 sec/batch
Epoch: 2/20... Training Step: 7855... Training loss: 1.7940... 0.1192 sec/batch
Epoch: 2/20... Training Step: 7856... Training loss: 1.6629... 0.1176 sec/batch
Epoch: 2/20... Training Step: 7857... Training loss: 1.8512... 0.1245 sec/batch
Epoch: 2/20... Training Step: 7858... Training loss: 1.9739... 0.1004 sec/batch
Epoch: 2/20... Training Step: 7859... Training loss: 1.7288... 0.0936 sec/batch
Epoch: 2/20... Training Step: 7860... Training loss: 1.8328... 0.1149 sec/batch
Epoch: 2/20... Training Step: 7861... Training loss: 1.8363... 0.0945 sec/batch
Epoch: 2/20... Training Step: 7862... Training loss: 1.7760... 0.1172 sec/batch
Epoch: 2/20... Training Step: 7863... Training loss: 1.7477... 0.1419 sec/batch
Epoch: 2/20... Training Step: 7864... Training loss: 1.8416... 0.1119 sec/batch
Epoch: 2/20... Training Step: 7865... Training loss: 1.7688... 0.1520 sec/batch
Epoch: 2/20... Training Step: 7866... Training loss: 1.9831... 0.1303 sec/batch
Epoch: 2/20... Training Step: 7867... Training loss: 1.9624... 0.1081 sec/batch
Epoch: 2/20... Training Step: 7868... Training loss: 1.8457... 0.1453 sec/batch
Epoch: 2/20... Training Step: 7869... Training loss: 1.8299... 0.1222 sec/batch
Epoch: 2/20... Training Step: 7870... Training loss: 1.8790... 0.1355 sec/batch
Epoch: 2/20... Training Step: 7871... Training loss: 1.7618... 0.1037 sec/batch
Epoch: 2/20... Training Step: 7872... Training loss: 2.0177... 0.1128 sec/batch
Epoch: 2/20... Training Step: 7873... Training loss: 1.9051... 0.1022 sec/batch
Epoch: 2/20... Training Step: 7874... Training loss: 1.8733... 0.1222 sec/batch
Epoch: 2/20... Training Step: 7875... Training loss: 2.0530... 0.1126 sec/batch
Epoch: 2/20... Training Step: 7876... Training loss: 1.9764... 0.1105 sec/batch
Epoch: 2/20... Training Step: 7877... Training loss: 1.8882... 0.1019 sec/batch
Epoch: 2/20... Training Step: 7878... Training loss: 1.7745... 0.0988 sec/batch
Epoch: 2/20... Training Step: 7879... Training loss: 2.0018... 0.0964 sec/batch
Epoch: 2/20... Training Step: 7880... Training loss: 1.9359... 0.1255 sec/batch
Epoch: 2/20... Training Step: 7881... Training loss: 1.9527... 0.1252 sec/batch
Epoch: 2/20... Training Step: 7882... Training loss: 1.7591... 0.1321 sec/batch
Epoch: 2/20... Training Step: 7883... Training loss: 1.7655... 0.1196 sec/batch
Epoch: 2/20... Training Step: 7884... Training loss: 1.8030... 0.0996 sec/batch
Epoch: 2/20... Training Step: 7885... Training loss: 1.7377... 0.0817 sec/batch
Epoch: 2/20... Training Step: 7886... Training loss: 1.7672... 0.0750 sec/batch
Epoch: 2/20... Training Step: 7887... Training loss: 1.8012... 0.0921 sec/batch
Epoch: 2/20... Training Step: 7888... Training loss: 1.7278... 0.0956 sec/batch
Epoch: 2/20... Training Step: 7889... Training loss: 1.7215... 0.0786 sec/batch
Epoch: 2/20... Training Step: 7890... Training loss: 2.0355... 0.0868 sec/batch
Epoch: 2/20... Training Step: 7891... Training loss: 1.6862... 0.0731 sec/batch
Epoch: 2/20... Training Step: 7892... Training loss: 1.9142... 0.0727 sec/batch
Epoch: 2/20... Training Step: 7893... Training loss: 1.7328... 0.0678 sec/batch
Epoch: 2/20... Training Step: 7894... Training loss: 1.8646... 0.0780 sec/batch
Epoch: 2/20... Training Step: 7895... Training loss: 1.8435... 0.0704 sec/batch
Epoch: 2/20... Training Step: 7896... Training loss: 1.7639... 0.0826 sec/batch
Epoch: 2/20... Training Step: 7897... Training loss: 1.8252... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7898... Training loss: 1.7562... 0.0853 sec/batch
Epoch: 2/20... Training Step: 7899... Training loss: 1.7583... 0.0863 sec/batch
Epoch: 2/20... Training Step: 7900... Training loss: 1.8238... 0.0732 sec/batch
Epoch: 2/20... Training Step: 7901... Training loss: 1.7781... 0.0955 sec/batch
Epoch: 2/20... Training Step: 7902... Training loss: 1.9723... 0.0715 sec/batch
Epoch: 2/20... Training Step: 7903... Training loss: 1.8780... 0.0729 sec/batch
Epoch: 2/20... Training Step: 7904... Training loss: 1.6879... 0.0949 sec/batch
Epoch: 2/20... Training Step: 7905... Training loss: 1.7130... 0.0820 sec/batch
Epoch: 2/20... Training Step: 7906... Training loss: 1.6948... 0.0953 sec/batch
Epoch: 2/20... Training Step: 7907... Training loss: 1.8456... 0.0870 sec/batch
Epoch: 2/20... Training Step: 7908... Training loss: 1.7099... 0.0645 sec/batch
Epoch: 2/20... Training Step: 7909... Training loss: 1.9614... 0.1072 sec/batch
Epoch: 2/20... Training Step: 7910... Training loss: 1.8536... 0.0819 sec/batch
Epoch: 2/20... Training Step: 7911... Training loss: 1.8051... 0.0651 sec/batch
Epoch: 2/20... Training Step: 7912... Training loss: 1.8633... 0.0683 sec/batch
Epoch: 2/20... Training Step: 7913... Training loss: 1.6723... 0.0764 sec/batch
Epoch: 2/20... Training Step: 7914... Training loss: 1.6778... 0.1125 sec/batch
Epoch: 2/20... Training Step: 7915... Training loss: 1.7413... 0.0699 sec/batch
Epoch: 2/20... Training Step: 7916... Training loss: 2.0112... 0.0978 sec/batch
Epoch: 2/20... Training Step: 7917... Training loss: 1.9062... 0.0911 sec/batch
Epoch: 2/20... Training Step: 7918... Training loss: 1.8015... 0.0895 sec/batch
Epoch: 2/20... Training Step: 7919... Training loss: 1.8079... 0.0793 sec/batch
Epoch: 2/20... Training Step: 7920... Training loss: 1.6743... 0.0773 sec/batch
Epoch: 2/20... Training Step: 7921... Training loss: 1.7635... 0.0921 sec/batch
Epoch: 2/20... Training Step: 7922... Training loss: 1.6406... 0.0905 sec/batch
Epoch: 2/20... Training Step: 7923... Training loss: 1.6515... 0.0923 sec/batch
Epoch: 2/20... Training Step: 7924... Training loss: 1.8355... 0.0784 sec/batch
Epoch: 2/20... Training Step: 7925... Training loss: 1.6210... 0.0740 sec/batch
Epoch: 2/20... Training Step: 7926... Training loss: 1.9441... 0.0705 sec/batch
Epoch: 2/20... Training Step: 7927... Training loss: 1.7906... 0.0772 sec/batch
Epoch: 2/20... Training Step: 7928... Training loss: 1.8253... 0.1031 sec/batch
Epoch: 2/20... Training Step: 7929... Training loss: 1.6268... 0.0956 sec/batch
Epoch: 2/20... Training Step: 7930... Training loss: 1.7776... 0.0723 sec/batch
Epoch: 2/20... Training Step: 7931... Training loss: 1.7176... 0.0794 sec/batch
Epoch: 2/20... Training Step: 7932... Training loss: 1.8620... 0.0702 sec/batch
Epoch: 2/20... Training Step: 7933... Training loss: 1.7984... 0.0725 sec/batch
Epoch: 2/20... Training Step: 7934... Training loss: 1.7209... 0.0847 sec/batch
Epoch: 2/20... Training Step: 7935... Training loss: 1.7497... 0.0831 sec/batch
Epoch: 2/20... Training Step: 7936... Training loss: 1.8105... 0.0830 sec/batch
Epoch: 2/20... Training Step: 7937... Training loss: 1.6861... 0.0599 sec/batch
Epoch: 2/20... Training Step: 7938... Training loss: 1.7355... 0.0684 sec/batch
Epoch: 2/20... Training Step: 7939... Training loss: 1.8614... 0.0993 sec/batch
Epoch: 2/20... Training Step: 7940... Training loss: 1.8532... 0.1072 sec/batch
Epoch: 3/20... Training Step: 7941... Training loss: 1.6841... 0.0688 sec/batch
Epoch: 3/20... Training Step: 7942... Training loss: 1.6769... 0.0902 sec/batch
Epoch: 3/20... Training Step: 7943... Training loss: 1.6760... 0.0794 sec/batch
Epoch: 3/20... Training Step: 7944... Training loss: 1.6607... 0.0859 sec/batch
Epoch: 3/20... Training Step: 7945... Training loss: 1.7729... 0.0837 sec/batch
Epoch: 3/20... Training Step: 7946... Training loss: 1.6642... 0.0663 sec/batch
Epoch: 3/20... Training Step: 7947... Training loss: 1.5348... 0.0866 sec/batch
Epoch: 3/20... Training Step: 7948... Training loss: 1.7074... 0.0792 sec/batch
Epoch: 3/20... Training Step: 7949... Training loss: 1.7864... 0.0791 sec/batch
Epoch: 3/20... Training Step: 7950... Training loss: 1.7763... 0.0837 sec/batch
Epoch: 3/20... Training Step: 7951... Training loss: 1.6626... 0.0878 sec/batch
Epoch: 3/20... Training Step: 7952... Training loss: 1.5575... 0.0872 sec/batch
Epoch: 3/20... Training Step: 7953... Training loss: 1.6699... 0.0903 sec/batch
Epoch: 3/20... Training Step: 7954... Training loss: 1.6512... 0.0854 sec/batch
Epoch: 3/20... Training Step: 7955... Training loss: 1.5565... 0.0717 sec/batch
Epoch: 3/20... Training Step: 7956... Training loss: 1.6448... 0.0696 sec/batch
Epoch: 3/20... Training Step: 7957... Training loss: 1.6546... 0.0711 sec/batch
Epoch: 3/20... Training Step: 7958... Training loss: 1.6459... 0.0756 sec/batch
Epoch: 3/20... Training Step: 7959... Training loss: 1.7377... 0.0727 sec/batch
Epoch: 3/20... Training Step: 7960... Training loss: 1.9110... 0.0656 sec/batch
Epoch: 3/20... Training Step: 7961... Training loss: 1.7091... 0.0716 sec/batch
Epoch: 3/20... Training Step: 7962... Training loss: 1.6614... 0.0621 sec/batch
Epoch: 3/20... Training Step: 7963... Training loss: 1.6917... 0.0641 sec/batch
Epoch: 3/20... Training Step: 7964... Training loss: 1.6269... 0.0783 sec/batch
Epoch: 3/20... Training Step: 7965... Training loss: 1.6609... 0.0722 sec/batch
Epoch: 3/20... Training Step: 7966... Training loss: 1.7558... 0.0747 sec/batch
Epoch: 3/20... Training Step: 7967... Training loss: 1.8558... 0.0798 sec/batch
Epoch: 3/20... Training Step: 7968... Training loss: 1.9101... 0.0674 sec/batch
Epoch: 3/20... Training Step: 7969... Training loss: 1.8848... 0.0619 sec/batch
Epoch: 3/20... Training Step: 7970... Training loss: 1.6606... 0.0735 sec/batch
Epoch: 3/20... Training Step: 7971... Training loss: 1.7986... 0.0611 sec/batch
Epoch: 3/20... Training Step: 7972... Training loss: 1.8445... 0.0735 sec/batch
Epoch: 3/20... Training Step: 7973... Training loss: 1.7279... 0.0678 sec/batch
Epoch: 3/20... Training Step: 7974... Training loss: 1.7607... 0.0660 sec/batch
Epoch: 3/20... Training Step: 7975... Training loss: 1.6939... 0.0804 sec/batch
Epoch: 3/20... Training Step: 7976... Training loss: 1.5775... 0.0690 sec/batch
Epoch: 3/20... Training Step: 7977... Training loss: 1.7061... 0.0760 sec/batch
Epoch: 3/20... Training Step: 7978... Training loss: 1.7535... 0.0644 sec/batch
Epoch: 3/20... Training Step: 7979... Training loss: 1.6974... 0.0657 sec/batch
Epoch: 3/20... Training Step: 7980... Training loss: 1.7238... 0.0694 sec/batch
Epoch: 3/20... Training Step: 7981... Training loss: 1.7562... 0.0675 sec/batch
Epoch: 3/20... Training Step: 7982... Training loss: 1.7288... 0.0839 sec/batch
Epoch: 3/20... Training Step: 7983... Training loss: 1.7051... 0.0648 sec/batch
Epoch: 3/20... Training Step: 7984... Training loss: 1.6263... 0.0614 sec/batch
Epoch: 3/20... Training Step: 7985... Training loss: 1.7302... 0.0717 sec/batch
Epoch: 3/20... Training Step: 7986... Training loss: 1.8396... 0.0711 sec/batch
Epoch: 3/20... Training Step: 7987... Training loss: 1.6758... 0.0831 sec/batch
Epoch: 3/20... Training Step: 7988... Training loss: 1.7113... 0.0702 sec/batch
Epoch: 3/20... Training Step: 7989... Training loss: 1.7693... 0.0743 sec/batch
Epoch: 3/20... Training Step: 7990... Training loss: 1.8540... 0.0809 sec/batch
Epoch: 3/20... Training Step: 7991... Training loss: 1.7923... 0.0620 sec/batch
Epoch: 3/20... Training Step: 7992... Training loss: 1.7455... 0.0655 sec/batch
Epoch: 3/20... Training Step: 7993... Training loss: 1.8471... 0.0681 sec/batch
Epoch: 3/20... Training Step: 7994... Training loss: 1.8639... 0.0731 sec/batch
Epoch: 3/20... Training Step: 7995... Training loss: 1.8358... 0.0633 sec/batch
Epoch: 3/20... Training Step: 7996... Training loss: 1.6737... 0.0689 sec/batch
Epoch: 3/20... Training Step: 7997... Training loss: 1.7798... 0.0680 sec/batch
Epoch: 3/20... Training Step: 7998... Training loss: 1.6924... 0.0612 sec/batch
Epoch: 3/20... Training Step: 7999... Training loss: 1.7127... 0.0669 sec/batch
Epoch: 3/20... Training Step: 8000... Training loss: 1.7071... 0.0638 sec/batch
Epoch: 3/20... Training Step: 8001... Training loss: 1.9938... 0.0707 sec/batch
Epoch: 3/20... Training Step: 8002... Training loss: 1.8441... 0.0656 sec/batch
Epoch: 3/20... Training Step: 8003... Training loss: 1.7694... 0.0702 sec/batch
Epoch: 3/20... Training Step: 8004... Training loss: 1.5309... 0.0687 sec/batch
Epoch: 3/20... Training Step: 8005... Training loss: 1.8743... 0.0781 sec/batch
Epoch: 3/20... Training Step: 8006... Training loss: 1.7797... 0.0730 sec/batch
Epoch: 3/20... Training Step: 8007... Training loss: 1.7624... 0.0686 sec/batch
Epoch: 3/20... Training Step: 8008... Training loss: 1.9412... 0.0533 sec/batch
Epoch: 3/20... Training Step: 8009... Training loss: 1.9714... 0.0589 sec/batch
Epoch: 3/20... Training Step: 8010... Training loss: 1.7668... 0.0691 sec/batch
Epoch: 3/20... Training Step: 8011... Training loss: 1.7928... 0.0690 sec/batch
Epoch: 3/20... Training Step: 8012... Training loss: 1.7284... 0.0663 sec/batch
Epoch: 3/20... Training Step: 8013... Training loss: 1.6858... 0.0656 sec/batch
Epoch: 3/20... Training Step: 8014... Training loss: 1.7873... 0.0637 sec/batch
Epoch: 3/20... Training Step: 8015... Training loss: 1.6634... 0.0649 sec/batch
Epoch: 3/20... Training Step: 8016... Training loss: 1.6933... 0.0730 sec/batch
Epoch: 3/20... Training Step: 8017... Training loss: 1.6314... 0.0757 sec/batch
Epoch: 3/20... Training Step: 8018... Training loss: 1.6931... 0.0635 sec/batch
Epoch: 3/20... Training Step: 8019... Training loss: 1.6574... 0.0773 sec/batch
Epoch: 3/20... Training Step: 8020... Training loss: 1.6928... 0.0683 sec/batch
Epoch: 3/20... Training Step: 8021... Training loss: 1.6264... 0.0745 sec/batch
Epoch: 3/20... Training Step: 8022... Training loss: 1.6661... 0.0746 sec/batch
Epoch: 3/20... Training Step: 8023... Training loss: 1.6418... 0.0757 sec/batch
Epoch: 3/20... Training Step: 8024... Training loss: 1.7127... 0.0691 sec/batch
Epoch: 3/20... Training Step: 8025... Training loss: 1.5298... 0.0639 sec/batch
Epoch: 3/20... Training Step: 8026... Training loss: 1.6692... 0.0598 sec/batch
Epoch: 3/20... Training Step: 8027... Training loss: 1.7540... 0.0860 sec/batch
Epoch: 3/20... Training Step: 8028... Training loss: 1.5888... 0.0717 sec/batch
Epoch: 3/20... Training Step: 8029... Training loss: 1.6694... 0.0685 sec/batch
Epoch: 3/20... Training Step: 8030... Training loss: 1.7523... 0.0687 sec/batch
Epoch: 3/20... Training Step: 8031... Training loss: 1.6800... 0.0588 sec/batch
Epoch: 3/20... Training Step: 8032... Training loss: 1.6707... 0.0652 sec/batch
Epoch: 3/20... Training Step: 8033... Training loss: 1.7222... 0.0618 sec/batch
Epoch: 3/20... Training Step: 8034... Training loss: 1.7531... 0.0733 sec/batch
Epoch: 3/20... Training Step: 8035... Training loss: 1.6206... 0.0743 sec/batch
Epoch: 3/20... Training Step: 8036... Training loss: 1.7110... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8037... Training loss: 1.7370... 0.0763 sec/batch
Epoch: 3/20... Training Step: 8038... Training loss: 1.7731... 0.0684 sec/batch
Epoch: 3/20... Training Step: 8039... Training loss: 1.6514... 0.0547 sec/batch
Epoch: 3/20... Training Step: 8040... Training loss: 1.6789... 0.0743 sec/batch
Epoch: 3/20... Training Step: 8041... Training loss: 1.8649... 0.0731 sec/batch
Epoch: 3/20... Training Step: 8042... Training loss: 1.6832... 0.0689 sec/batch
Epoch: 3/20... Training Step: 8043... Training loss: 1.7758... 0.0772 sec/batch
Epoch: 3/20... Training Step: 8044... Training loss: 1.5480... 0.0629 sec/batch
Epoch: 3/20... Training Step: 8045... Training loss: 1.7951... 0.0744 sec/batch
Epoch: 3/20... Training Step: 8046... Training loss: 1.7030... 0.0867 sec/batch
Epoch: 3/20... Training Step: 8047... Training loss: 1.6040... 0.0818 sec/batch
Epoch: 3/20... Training Step: 8048... Training loss: 1.6755... 0.0766 sec/batch
Epoch: 3/20... Training Step: 8049... Training loss: 1.6070... 0.0812 sec/batch
Epoch: 3/20... Training Step: 8050... Training loss: 1.7148... 0.0640 sec/batch
Epoch: 3/20... Training Step: 8051... Training loss: 1.5707... 0.0723 sec/batch
Epoch: 3/20... Training Step: 8052... Training loss: 1.6644... 0.0721 sec/batch
Epoch: 3/20... Training Step: 8053... Training loss: 1.6390... 0.0824 sec/batch
Epoch: 3/20... Training Step: 8054... Training loss: 1.6986... 0.0967 sec/batch
Epoch: 3/20... Training Step: 8055... Training loss: 1.6950... 0.0713 sec/batch
Epoch: 3/20... Training Step: 8056... Training loss: 1.7591... 0.0743 sec/batch
Epoch: 3/20... Training Step: 8057... Training loss: 1.6419... 0.0876 sec/batch
Epoch: 3/20... Training Step: 8058... Training loss: 1.5876... 0.0789 sec/batch
Epoch: 3/20... Training Step: 8059... Training loss: 1.6449... 0.0817 sec/batch
Epoch: 3/20... Training Step: 8060... Training loss: 1.6640... 0.0719 sec/batch
Epoch: 3/20... Training Step: 8061... Training loss: 1.5975... 0.0628 sec/batch
Epoch: 3/20... Training Step: 8062... Training loss: 1.5627... 0.0737 sec/batch
Epoch: 3/20... Training Step: 8063... Training loss: 1.7003... 0.0817 sec/batch
Epoch: 3/20... Training Step: 8064... Training loss: 1.6353... 0.0769 sec/batch
Epoch: 3/20... Training Step: 8065... Training loss: 1.7170... 0.0765 sec/batch
Epoch: 3/20... Training Step: 8066... Training loss: 1.7386... 0.0685 sec/batch
Epoch: 3/20... Training Step: 8067... Training loss: 1.7356... 0.0605 sec/batch
Epoch: 3/20... Training Step: 8068... Training loss: 1.6434... 0.0652 sec/batch
Epoch: 3/20... Training Step: 8069... Training loss: 1.8268... 0.0660 sec/batch
Epoch: 3/20... Training Step: 8070... Training loss: 1.6387... 0.0863 sec/batch
Epoch: 3/20... Training Step: 8071... Training loss: 1.7066... 0.0771 sec/batch
Epoch: 3/20... Training Step: 8072... Training loss: 1.6812... 0.0752 sec/batch
Epoch: 3/20... Training Step: 8073... Training loss: 1.7481... 0.0635 sec/batch
Epoch: 3/20... Training Step: 8074... Training loss: 1.6380... 0.0621 sec/batch
Epoch: 3/20... Training Step: 8075... Training loss: 1.7153... 0.0770 sec/batch
Epoch: 3/20... Training Step: 8076... Training loss: 1.6479... 0.0742 sec/batch
Epoch: 3/20... Training Step: 8077... Training loss: 1.8052... 0.0583 sec/batch
Epoch: 3/20... Training Step: 8078... Training loss: 1.6546... 0.0671 sec/batch
Epoch: 3/20... Training Step: 8079... Training loss: 1.7452... 0.0783 sec/batch
Epoch: 3/20... Training Step: 8080... Training loss: 1.6425... 0.0785 sec/batch
Epoch: 3/20... Training Step: 8081... Training loss: 1.7833... 0.0667 sec/batch
Epoch: 3/20... Training Step: 8082... Training loss: 1.6466... 0.0849 sec/batch
Epoch: 3/20... Training Step: 8083... Training loss: 1.7138... 0.0770 sec/batch
Epoch: 3/20... Training Step: 8084... Training loss: 1.7718... 0.0804 sec/batch
Epoch: 3/20... Training Step: 8085... Training loss: 1.6416... 0.0749 sec/batch
Epoch: 3/20... Training Step: 8086... Training loss: 1.6885... 0.0786 sec/batch
Epoch: 3/20... Training Step: 8087... Training loss: 1.7848... 0.0733 sec/batch
Epoch: 3/20... Training Step: 8088... Training loss: 1.5735... 0.0645 sec/batch
Epoch: 3/20... Training Step: 8089... Training loss: 1.5839... 0.0660 sec/batch
Epoch: 3/20... Training Step: 8090... Training loss: 1.5334... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8091... Training loss: 1.7548... 0.0687 sec/batch
Epoch: 3/20... Training Step: 8092... Training loss: 1.6542... 0.0756 sec/batch
Epoch: 3/20... Training Step: 8093... Training loss: 1.6458... 0.0595 sec/batch
Epoch: 3/20... Training Step: 8094... Training loss: 1.6703... 0.0822 sec/batch
Epoch: 3/20... Training Step: 8095... Training loss: 1.8757... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8096... Training loss: 1.6936... 0.0741 sec/batch
Epoch: 3/20... Training Step: 8097... Training loss: 1.5429... 0.0886 sec/batch
Epoch: 3/20... Training Step: 8098... Training loss: 1.8379... 0.0624 sec/batch
Epoch: 3/20... Training Step: 8099... Training loss: 1.9481... 0.0553 sec/batch
Epoch: 3/20... Training Step: 8100... Training loss: 1.8838... 0.0684 sec/batch
Epoch: 3/20... Training Step: 8101... Training loss: 1.7739... 0.0628 sec/batch
Epoch: 3/20... Training Step: 8102... Training loss: 1.7209... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8103... Training loss: 1.7121... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8104... Training loss: 1.7452... 0.0719 sec/batch
Epoch: 3/20... Training Step: 8105... Training loss: 1.8069... 0.0872 sec/batch
Epoch: 3/20... Training Step: 8106... Training loss: 1.7208... 0.0656 sec/batch
Epoch: 3/20... Training Step: 8107... Training loss: 1.7100... 0.0638 sec/batch
Epoch: 3/20... Training Step: 8108... Training loss: 1.8178... 0.0878 sec/batch
Epoch: 3/20... Training Step: 8109... Training loss: 1.7198... 0.0554 sec/batch
Epoch: 3/20... Training Step: 8110... Training loss: 1.8234... 0.0852 sec/batch
Epoch: 3/20... Training Step: 8111... Training loss: 1.7669... 0.0694 sec/batch
Epoch: 3/20... Training Step: 8112... Training loss: 1.7486... 0.0841 sec/batch
Epoch: 3/20... Training Step: 8113... Training loss: 1.7454... 0.0653 sec/batch
Epoch: 3/20... Training Step: 8114... Training loss: 1.7116... 0.0732 sec/batch
Epoch: 3/20... Training Step: 8115... Training loss: 1.7229... 0.0671 sec/batch
Epoch: 3/20... Training Step: 8116... Training loss: 1.6579... 0.0539 sec/batch
Epoch: 3/20... Training Step: 8117... Training loss: 1.6783... 0.0713 sec/batch
Epoch: 3/20... Training Step: 8118... Training loss: 1.6391... 0.0789 sec/batch
Epoch: 3/20... Training Step: 8119... Training loss: 1.6812... 0.0703 sec/batch
Epoch: 3/20... Training Step: 8120... Training loss: 1.6094... 0.0654 sec/batch
Epoch: 3/20... Training Step: 8121... Training loss: 1.6992... 0.0671 sec/batch
Epoch: 3/20... Training Step: 8122... Training loss: 1.8219... 0.0762 sec/batch
Epoch: 3/20... Training Step: 8123... Training loss: 1.6747... 0.0738 sec/batch
Epoch: 3/20... Training Step: 8124... Training loss: 1.6356... 0.0739 sec/batch
Epoch: 3/20... Training Step: 8125... Training loss: 1.6432... 0.0765 sec/batch
Epoch: 3/20... Training Step: 8126... Training loss: 1.6716... 0.0791 sec/batch
Epoch: 3/20... Training Step: 8127... Training loss: 1.5648... 0.0679 sec/batch
Epoch: 3/20... Training Step: 8128... Training loss: 1.6534... 0.0686 sec/batch
Epoch: 3/20... Training Step: 8129... Training loss: 1.7052... 0.0733 sec/batch
Epoch: 3/20... Training Step: 8130... Training loss: 1.7943... 0.0722 sec/batch
Epoch: 3/20... Training Step: 8131... Training loss: 1.6827... 0.0726 sec/batch
Epoch: 3/20... Training Step: 8132... Training loss: 1.6802... 0.0677 sec/batch
Epoch: 3/20... Training Step: 8133... Training loss: 1.7533... 0.0639 sec/batch
Epoch: 3/20... Training Step: 8134... Training loss: 1.8859... 0.0852 sec/batch
Epoch: 3/20... Training Step: 8135... Training loss: 1.7109... 0.0782 sec/batch
Epoch: 3/20... Training Step: 8136... Training loss: 1.6397... 0.0651 sec/batch
Epoch: 3/20... Training Step: 8137... Training loss: 1.6177... 0.0652 sec/batch
Epoch: 3/20... Training Step: 8138... Training loss: 1.7907... 0.0787 sec/batch
Epoch: 3/20... Training Step: 8139... Training loss: 1.7172... 0.0752 sec/batch
Epoch: 3/20... Training Step: 8140... Training loss: 1.7278... 0.0670 sec/batch
Epoch: 3/20... Training Step: 8141... Training loss: 1.5668... 0.0785 sec/batch
Epoch: 3/20... Training Step: 8142... Training loss: 1.7794... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8143... Training loss: 1.6837... 0.0606 sec/batch
Epoch: 3/20... Training Step: 8144... Training loss: 1.7483... 0.0631 sec/batch
Epoch: 3/20... Training Step: 8145... Training loss: 1.7708... 0.0664 sec/batch
Epoch: 3/20... Training Step: 8146... Training loss: 1.7369... 0.0634 sec/batch
Epoch: 3/20... Training Step: 8147... Training loss: 1.7327... 0.0669 sec/batch
Epoch: 3/20... Training Step: 8148... Training loss: 1.6819... 0.0677 sec/batch
Epoch: 3/20... Training Step: 8149... Training loss: 1.7739... 0.0624 sec/batch
Epoch: 3/20... Training Step: 8150... Training loss: 1.6620... 0.0591 sec/batch
Epoch: 3/20... Training Step: 8151... Training loss: 1.5851... 0.0769 sec/batch
Epoch: 3/20... Training Step: 8152... Training loss: 1.7392... 0.0684 sec/batch
Epoch: 3/20... Training Step: 8153... Training loss: 1.7277... 0.0824 sec/batch
Epoch: 3/20... Training Step: 8154... Training loss: 1.7639... 0.0659 sec/batch
Epoch: 3/20... Training Step: 8155... Training loss: 1.6721... 0.0997 sec/batch
Epoch: 3/20... Training Step: 8156... Training loss: 1.7619... 0.0681 sec/batch
Epoch: 3/20... Training Step: 8157... Training loss: 1.8154... 0.0737 sec/batch
Epoch: 3/20... Training Step: 8158... Training loss: 1.7747... 0.0674 sec/batch
Epoch: 3/20... Training Step: 8159... Training loss: 1.7495... 0.0738 sec/batch
Epoch: 3/20... Training Step: 8160... Training loss: 1.6833... 0.0805 sec/batch
Epoch: 3/20... Training Step: 8161... Training loss: 1.7325... 0.0801 sec/batch
Epoch: 3/20... Training Step: 8162... Training loss: 1.7659... 0.0796 sec/batch
Epoch: 3/20... Training Step: 8163... Training loss: 1.8033... 0.0596 sec/batch
Epoch: 3/20... Training Step: 8164... Training loss: 1.8235... 0.0842 sec/batch
Epoch: 3/20... Training Step: 8165... Training loss: 1.7354... 0.0787 sec/batch
Epoch: 3/20... Training Step: 8166... Training loss: 1.7901... 0.0785 sec/batch
Epoch: 3/20... Training Step: 8167... Training loss: 1.8455... 0.0792 sec/batch
Epoch: 3/20... Training Step: 8168... Training loss: 1.8469... 0.0801 sec/batch
Epoch: 3/20... Training Step: 8169... Training loss: 1.7001... 0.0818 sec/batch
Epoch: 3/20... Training Step: 8170... Training loss: 1.6721... 0.0669 sec/batch
Epoch: 3/20... Training Step: 8171... Training loss: 1.7258... 0.0746 sec/batch
Epoch: 3/20... Training Step: 8172... Training loss: 1.6234... 0.0801 sec/batch
Epoch: 3/20... Training Step: 8173... Training loss: 1.7179... 0.0785 sec/batch
Epoch: 3/20... Training Step: 8174... Training loss: 1.7652... 0.0913 sec/batch
Epoch: 3/20... Training Step: 8175... Training loss: 1.7834... 0.0722 sec/batch
Epoch: 3/20... Training Step: 8176... Training loss: 1.8511... 0.0617 sec/batch
Epoch: 3/20... Training Step: 8177... Training loss: 1.8481... 0.0721 sec/batch
Epoch: 3/20... Training Step: 8178... Training loss: 1.8299... 0.0766 sec/batch
Epoch: 3/20... Training Step: 8179... Training loss: 1.8173... 0.0758 sec/batch
Epoch: 3/20... Training Step: 8180... Training loss: 1.7304... 0.0675 sec/batch
Epoch: 3/20... Training Step: 8181... Training loss: 1.8134... 0.0663 sec/batch
Epoch: 3/20... Training Step: 8182... Training loss: 1.8331... 0.0657 sec/batch
Epoch: 3/20... Training Step: 8183... Training loss: 1.7338... 0.0888 sec/batch
Epoch: 3/20... Training Step: 8184... Training loss: 1.8639... 0.0843 sec/batch
Epoch: 3/20... Training Step: 8185... Training loss: 1.8147... 0.0769 sec/batch
Epoch: 3/20... Training Step: 8186... Training loss: 1.8588... 0.0840 sec/batch
Epoch: 3/20... Training Step: 8187... Training loss: 1.7751... 0.0859 sec/batch
Epoch: 3/20... Training Step: 8188... Training loss: 1.7902... 0.0836 sec/batch
Epoch: 3/20... Training Step: 8189... Training loss: 1.5449... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8190... Training loss: 1.7028... 0.0803 sec/batch
Epoch: 3/20... Training Step: 8191... Training loss: 1.7441... 0.0724 sec/batch
Epoch: 3/20... Training Step: 8192... Training loss: 1.8241... 0.0798 sec/batch
Epoch: 3/20... Training Step: 8193... Training loss: 1.6915... 0.0855 sec/batch
Epoch: 3/20... Training Step: 8194... Training loss: 1.6946... 0.0810 sec/batch
Epoch: 3/20... Training Step: 8195... Training loss: 1.8179... 0.0762 sec/batch
Epoch: 3/20... Training Step: 8196... Training loss: 1.6966... 0.0873 sec/batch
Epoch: 3/20... Training Step: 8197... Training loss: 1.7099... 0.0737 sec/batch
Epoch: 3/20... Training Step: 8198... Training loss: 1.6723... 0.0895 sec/batch
Epoch: 3/20... Training Step: 8199... Training loss: 1.6267... 0.0767 sec/batch
Epoch: 3/20... Training Step: 8200... Training loss: 1.6871... 0.0896 sec/batch
Epoch: 3/20... Training Step: 8201... Training loss: 1.7306... 0.0834 sec/batch
Epoch: 3/20... Training Step: 8202... Training loss: 1.6799... 0.0652 sec/batch
Epoch: 3/20... Training Step: 8203... Training loss: 1.7749... 0.0696 sec/batch
Epoch: 3/20... Training Step: 8204... Training loss: 1.7602... 0.0595 sec/batch
Epoch: 3/20... Training Step: 8205... Training loss: 1.5734... 0.0741 sec/batch
Epoch: 3/20... Training Step: 8206... Training loss: 1.6525... 0.0607 sec/batch
Epoch: 3/20... Training Step: 8207... Training loss: 1.5784... 0.0868 sec/batch
Epoch: 3/20... Training Step: 8208... Training loss: 1.7413... 0.0662 sec/batch
Epoch: 3/20... Training Step: 8209... Training loss: 1.7343... 0.0600 sec/batch
Epoch: 3/20... Training Step: 8210... Training loss: 1.7908... 0.0808 sec/batch
Epoch: 3/20... Training Step: 8211... Training loss: 1.7395... 0.0674 sec/batch
Epoch: 3/20... Training Step: 8212... Training loss: 1.6751... 0.0751 sec/batch
Epoch: 3/20... Training Step: 8213... Training loss: 1.8460... 0.0695 sec/batch
Epoch: 3/20... Training Step: 8214... Training loss: 1.7969... 0.0622 sec/batch
Epoch: 3/20... Training Step: 8215... Training loss: 1.8034... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8216... Training loss: 1.6274... 0.0689 sec/batch
Epoch: 3/20... Training Step: 8217... Training loss: 1.6350... 0.0642 sec/batch
Epoch: 3/20... Training Step: 8218... Training loss: 1.7671... 0.0645 sec/batch
Epoch: 3/20... Training Step: 8219... Training loss: 1.6870... 0.0626 sec/batch
Epoch: 3/20... Training Step: 8220... Training loss: 1.7223... 0.0643 sec/batch
Epoch: 3/20... Training Step: 8221... Training loss: 1.6587... 0.0655 sec/batch
Epoch: 3/20... Training Step: 8222... Training loss: 1.7101... 0.0698 sec/batch
Epoch: 3/20... Training Step: 8223... Training loss: 1.5659... 0.0740 sec/batch
Epoch: 3/20... Training Step: 8224... Training loss: 1.8787... 0.0688 sec/batch
Epoch: 3/20... Training Step: 8225... Training loss: 1.6816... 0.0605 sec/batch
Epoch: 3/20... Training Step: 8226... Training loss: 1.5840... 0.0758 sec/batch
Epoch: 3/20... Training Step: 8227... Training loss: 1.5730... 0.0743 sec/batch
Epoch: 3/20... Training Step: 8228... Training loss: 1.7343... 0.0668 sec/batch
Epoch: 3/20... Training Step: 8229... Training loss: 1.6831... 0.0622 sec/batch
Epoch: 3/20... Training Step: 8230... Training loss: 1.6873... 0.0651 sec/batch
Epoch: 3/20... Training Step: 8231... Training loss: 1.7236... 0.0612 sec/batch
Epoch: 3/20... Training Step: 8232... Training loss: 1.6206... 0.0770 sec/batch
Epoch: 3/20... Training Step: 8233... Training loss: 1.5931... 0.0710 sec/batch
Epoch: 3/20... Training Step: 8234... Training loss: 1.5818... 0.0770 sec/batch
Epoch: 3/20... Training Step: 8235... Training loss: 1.8127... 0.0838 sec/batch
Epoch: 3/20... Training Step: 8236... Training loss: 1.7703... 0.0763 sec/batch
Epoch: 3/20... Training Step: 8237... Training loss: 1.6139... 0.0783 sec/batch
Epoch: 3/20... Training Step: 8238... Training loss: 1.7499... 0.0808 sec/batch
Epoch: 3/20... Training Step: 8239... Training loss: 1.6036... 0.0715 sec/batch
Epoch: 3/20... Training Step: 8240... Training loss: 1.6103... 0.0607 sec/batch
Epoch: 3/20... Training Step: 8241... Training loss: 1.6883... 0.0625 sec/batch
Epoch: 3/20... Training Step: 8242... Training loss: 1.6595... 0.0675 sec/batch
Epoch: 3/20... Training Step: 8243... Training loss: 1.6941... 0.0698 sec/batch
Epoch: 3/20... Training Step: 8244... Training loss: 1.6854... 0.0927 sec/batch
Epoch: 3/20... Training Step: 8245... Training loss: 1.8094... 0.0819 sec/batch
Epoch: 3/20... Training Step: 8246... Training loss: 1.5798... 0.0685 sec/batch
Epoch: 3/20... Training Step: 8247... Training loss: 1.9052... 0.0700 sec/batch
Epoch: 3/20... Training Step: 8248... Training loss: 1.7579... 0.0807 sec/batch
Epoch: 3/20... Training Step: 8249... Training loss: 1.7444... 0.0815 sec/batch
Epoch: 3/20... Training Step: 8250... Training loss: 1.8041... 0.0892 sec/batch
Epoch: 3/20... Training Step: 8251... Training loss: 1.6840... 0.0822 sec/batch
Epoch: 3/20... Training Step: 8252... Training loss: 1.6659... 0.0573 sec/batch
Epoch: 3/20... Training Step: 8253... Training loss: 1.7250... 0.0780 sec/batch
Epoch: 3/20... Training Step: 8254... Training loss: 1.7724... 0.0699 sec/batch
Epoch: 3/20... Training Step: 8255... Training loss: 1.7761... 0.0696 sec/batch
Epoch: 3/20... Training Step: 8256... Training loss: 1.8347... 0.0776 sec/batch
Epoch: 3/20... Training Step: 8257... Training loss: 1.6958... 0.0974 sec/batch
Epoch: 3/20... Training Step: 8258... Training loss: 1.6624... 0.0784 sec/batch
Epoch: 3/20... Training Step: 8259... Training loss: 1.7812... 0.0744 sec/batch
Epoch: 3/20... Training Step: 8260... Training loss: 1.7760... 0.0743 sec/batch
Epoch: 3/20... Training Step: 8261... Training loss: 1.6713... 0.0758 sec/batch
Epoch: 3/20... Training Step: 8262... Training loss: 1.4917... 0.0732 sec/batch
Epoch: 3/20... Training Step: 8263... Training loss: 1.7927... 0.0841 sec/batch
Epoch: 3/20... Training Step: 8264... Training loss: 1.7495... 0.0639 sec/batch
Epoch: 3/20... Training Step: 8265... Training loss: 1.7542... 0.0744 sec/batch
Epoch: 3/20... Training Step: 8266... Training loss: 1.8295... 0.0683 sec/batch
Epoch: 3/20... Training Step: 8267... Training loss: 1.8577... 0.0713 sec/batch
Epoch: 3/20... Training Step: 8268... Training loss: 1.7826... 0.0773 sec/batch
Epoch: 3/20... Training Step: 8269... Training loss: 1.7533... 0.1001 sec/batch
Epoch: 3/20... Training Step: 8270... Training loss: 1.7155... 0.0769 sec/batch
Epoch: 3/20... Training Step: 8271... Training loss: 1.8172... 0.0866 sec/batch
Epoch: 3/20... Training Step: 8272... Training loss: 1.8477... 0.0753 sec/batch
Epoch: 3/20... Training Step: 8273... Training loss: 1.7279... 0.0766 sec/batch
Epoch: 3/20... Training Step: 8274... Training loss: 1.7978... 0.0878 sec/batch
Epoch: 3/20... Training Step: 8275... Training loss: 1.8080... 0.0807 sec/batch
Epoch: 3/20... Training Step: 8276... Training loss: 1.8249... 0.0683 sec/batch
Epoch: 3/20... Training Step: 8277... Training loss: 1.6142... 0.0893 sec/batch
Epoch: 3/20... Training Step: 8278... Training loss: 1.6126... 0.0814 sec/batch
Epoch: 3/20... Training Step: 8279... Training loss: 1.7258... 0.0838 sec/batch
Epoch: 3/20... Training Step: 8280... Training loss: 1.7202... 0.0825 sec/batch
Epoch: 3/20... Training Step: 8281... Training loss: 1.9682... 0.0769 sec/batch
Epoch: 3/20... Training Step: 8282... Training loss: 2.0245... 0.0647 sec/batch
Epoch: 3/20... Training Step: 8283... Training loss: 1.8815... 0.0729 sec/batch
Epoch: 3/20... Training Step: 8284... Training loss: 1.6941... 0.0864 sec/batch
Epoch: 3/20... Training Step: 8285... Training loss: 1.7642... 0.0801 sec/batch
Epoch: 3/20... Training Step: 8286... Training loss: 1.8104... 0.0768 sec/batch
Epoch: 3/20... Training Step: 8287... Training loss: 1.9052... 0.0964 sec/batch
Epoch: 3/20... Training Step: 8288... Training loss: 1.7232... 0.0763 sec/batch
Epoch: 3/20... Training Step: 8289... Training loss: 1.7638... 0.0751 sec/batch
Epoch: 3/20... Training Step: 8290... Training loss: 1.7395... 0.0668 sec/batch
Epoch: 3/20... Training Step: 8291... Training loss: 1.7124... 0.0957 sec/batch
Epoch: 3/20... Training Step: 8292... Training loss: 1.6202... 0.0813 sec/batch
Epoch: 3/20... Training Step: 8293... Training loss: 1.7636... 0.0882 sec/batch
Epoch: 3/20... Training Step: 8294... Training loss: 1.7781... 0.0699 sec/batch
Epoch: 3/20... Training Step: 8295... Training loss: 1.7690... 0.0700 sec/batch
Epoch: 3/20... Training Step: 8296... Training loss: 1.7374... 0.0870 sec/batch
Epoch: 3/20... Training Step: 8297... Training loss: 1.8299... 0.0665 sec/batch
Epoch: 3/20... Training Step: 8298... Training loss: 1.8024... 0.0794 sec/batch
Epoch: 3/20... Training Step: 8299... Training loss: 1.5957... 0.0709 sec/batch
Epoch: 3/20... Training Step: 8300... Training loss: 1.8210... 0.0712 sec/batch
Epoch: 3/20... Training Step: 8301... Training loss: 1.7704... 0.0825 sec/batch
Epoch: 3/20... Training Step: 8302... Training loss: 1.7294... 0.0771 sec/batch
Epoch: 3/20... Training Step: 8303... Training loss: 1.8111... 0.0687 sec/batch
Epoch: 3/20... Training Step: 8304... Training loss: 1.5810... 0.0636 sec/batch
Epoch: 3/20... Training Step: 8305... Training loss: 1.7655... 0.0905 sec/batch
Epoch: 3/20... Training Step: 8306... Training loss: 1.7539... 0.0803 sec/batch
Epoch: 3/20... Training Step: 8307... Training loss: 1.8026... 0.1093 sec/batch
Epoch: 3/20... Training Step: 8308... Training loss: 1.7449... 0.0867 sec/batch
Epoch: 3/20... Training Step: 8309... Training loss: 1.7515... 0.0703 sec/batch
Epoch: 3/20... Training Step: 8310... Training loss: 1.8137... 0.0915 sec/batch
Epoch: 3/20... Training Step: 8311... Training loss: 1.6835... 0.0599 sec/batch
Epoch: 3/20... Training Step: 8312... Training loss: 1.7234... 0.0663 sec/batch
Epoch: 3/20... Training Step: 8313... Training loss: 1.7684... 0.0640 sec/batch
Epoch: 3/20... Training Step: 8314... Training loss: 1.5882... 0.0954 sec/batch
Epoch: 3/20... Training Step: 8315... Training loss: 1.8261... 0.0977 sec/batch
Epoch: 3/20... Training Step: 8316... Training loss: 1.6894... 0.0689 sec/batch
Epoch: 3/20... Training Step: 8317... Training loss: 1.8129... 0.0653 sec/batch
Epoch: 3/20... Training Step: 8318... Training loss: 1.7367... 0.1013 sec/batch
Epoch: 3/20... Training Step: 8319... Training loss: 1.7430... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8320... Training loss: 1.7099... 0.0472 sec/batch
Epoch: 3/20... Training Step: 8321... Training loss: 1.7693... 0.0410 sec/batch
Epoch: 3/20... Training Step: 8322... Training loss: 1.7748... 0.0528 sec/batch
Epoch: 3/20... Training Step: 8323... Training loss: 1.9170... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8324... Training loss: 1.9210... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8325... Training loss: 1.8671... 0.0580 sec/batch
Epoch: 3/20... Training Step: 8326... Training loss: 1.7516... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8327... Training loss: 1.7321... 0.0414 sec/batch
Epoch: 3/20... Training Step: 8328... Training loss: 1.7668... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8329... Training loss: 1.6166... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8330... Training loss: 1.6302... 0.0451 sec/batch
Epoch: 3/20... Training Step: 8331... Training loss: 1.6695... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8332... Training loss: 1.8558... 0.0578 sec/batch
Epoch: 3/20... Training Step: 8333... Training loss: 1.6751... 0.0554 sec/batch
Epoch: 3/20... Training Step: 8334... Training loss: 1.6511... 0.0554 sec/batch
Epoch: 3/20... Training Step: 8335... Training loss: 1.7757... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8336... Training loss: 1.7369... 0.0407 sec/batch
Epoch: 3/20... Training Step: 8337... Training loss: 1.7530... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8338... Training loss: 1.7417... 0.0529 sec/batch
Epoch: 3/20... Training Step: 8339... Training loss: 1.7341... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8340... Training loss: 1.6847... 0.0471 sec/batch
Epoch: 3/20... Training Step: 8341... Training loss: 1.6260... 0.0581 sec/batch
Epoch: 3/20... Training Step: 8342... Training loss: 1.7047... 0.0510 sec/batch
Epoch: 3/20... Training Step: 8343... Training loss: 1.7408... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8344... Training loss: 1.6684... 0.0581 sec/batch
Epoch: 3/20... Training Step: 8345... Training loss: 1.7514... 0.0569 sec/batch
Epoch: 3/20... Training Step: 8346... Training loss: 1.7050... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8347... Training loss: 1.6989... 0.0422 sec/batch
Epoch: 3/20... Training Step: 8348... Training loss: 1.8795... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8349... Training loss: 1.6661... 0.0529 sec/batch
Epoch: 3/20... Training Step: 8350... Training loss: 1.7040... 0.0474 sec/batch
Epoch: 3/20... Training Step: 8351... Training loss: 1.7007... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8352... Training loss: 1.7170... 0.0606 sec/batch
Epoch: 3/20... Training Step: 8353... Training loss: 1.9296... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8354... Training loss: 1.6859... 0.0633 sec/batch
Epoch: 3/20... Training Step: 8355... Training loss: 1.7123... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8356... Training loss: 1.7484... 0.0542 sec/batch
Epoch: 3/20... Training Step: 8357... Training loss: 1.8195... 0.0768 sec/batch
Epoch: 3/20... Training Step: 8358... Training loss: 1.6248... 0.0592 sec/batch
Epoch: 3/20... Training Step: 8359... Training loss: 1.6752... 0.0842 sec/batch
Epoch: 3/20... Training Step: 8360... Training loss: 1.7004... 0.0801 sec/batch
Epoch: 3/20... Training Step: 8361... Training loss: 1.6687... 0.0589 sec/batch
Epoch: 3/20... Training Step: 8362... Training loss: 1.7785... 0.0572 sec/batch
Epoch: 3/20... Training Step: 8363... Training loss: 1.6245... 0.0436 sec/batch
Epoch: 3/20... Training Step: 8364... Training loss: 1.6076... 0.0637 sec/batch
Epoch: 3/20... Training Step: 8365... Training loss: 1.7102... 0.0444 sec/batch
Epoch: 3/20... Training Step: 8366... Training loss: 1.7465... 0.0518 sec/batch
Epoch: 3/20... Training Step: 8367... Training loss: 1.7689... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8368... Training loss: 1.7845... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8369... Training loss: 1.7351... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8370... Training loss: 1.6659... 0.0472 sec/batch
Epoch: 3/20... Training Step: 8371... Training loss: 1.6042... 0.0486 sec/batch
Epoch: 3/20... Training Step: 8372... Training loss: 1.7050... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8373... Training loss: 1.6348... 0.0536 sec/batch
Epoch: 3/20... Training Step: 8374... Training loss: 1.6508... 0.0496 sec/batch
Epoch: 3/20... Training Step: 8375... Training loss: 1.5844... 0.0561 sec/batch
Epoch: 3/20... Training Step: 8376... Training loss: 1.7396... 0.0463 sec/batch
Epoch: 3/20... Training Step: 8377... Training loss: 1.7324... 0.0436 sec/batch
Epoch: 3/20... Training Step: 8378... Training loss: 1.6607... 0.0538 sec/batch
Epoch: 3/20... Training Step: 8379... Training loss: 1.6675... 0.0514 sec/batch
Epoch: 3/20... Training Step: 8380... Training loss: 1.7570... 0.0498 sec/batch
Epoch: 3/20... Training Step: 8381... Training loss: 1.9088... 0.0503 sec/batch
Epoch: 3/20... Training Step: 8382... Training loss: 1.5965... 0.0504 sec/batch
Epoch: 3/20... Training Step: 8383... Training loss: 1.7284... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8384... Training loss: 1.6379... 0.0551 sec/batch
Epoch: 3/20... Training Step: 8385... Training loss: 1.8017... 0.0631 sec/batch
Epoch: 3/20... Training Step: 8386... Training loss: 1.8044... 0.0600 sec/batch
Epoch: 3/20... Training Step: 8387... Training loss: 1.8307... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8388... Training loss: 1.5919... 0.0507 sec/batch
Epoch: 3/20... Training Step: 8389... Training loss: 1.5947... 0.0558 sec/batch
Epoch: 3/20... Training Step: 8390... Training loss: 1.7886... 0.0463 sec/batch
Epoch: 3/20... Training Step: 8391... Training loss: 1.7317... 0.0518 sec/batch
Epoch: 3/20... Training Step: 8392... Training loss: 1.7024... 0.0474 sec/batch
Epoch: 3/20... Training Step: 8393... Training loss: 1.5829... 0.0639 sec/batch
Epoch: 3/20... Training Step: 8394... Training loss: 1.7003... 0.0502 sec/batch
Epoch: 3/20... Training Step: 8395... Training loss: 1.6632... 0.0509 sec/batch
Epoch: 3/20... Training Step: 8396... Training loss: 1.6442... 0.0553 sec/batch
Epoch: 3/20... Training Step: 8397... Training loss: 1.7029... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8398... Training loss: 1.7705... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8399... Training loss: 1.8733... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8400... Training loss: 1.8827... 0.0501 sec/batch
Epoch: 3/20... Training Step: 8401... Training loss: 1.7304... 0.0446 sec/batch
Epoch: 3/20... Training Step: 8402... Training loss: 1.7778... 0.0425 sec/batch
Epoch: 3/20... Training Step: 8403... Training loss: 1.7592... 0.0515 sec/batch
Epoch: 3/20... Training Step: 8404... Training loss: 1.6730... 0.0504 sec/batch
Epoch: 3/20... Training Step: 8405... Training loss: 1.7213... 0.0504 sec/batch
Epoch: 3/20... Training Step: 8406... Training loss: 1.7481... 0.0767 sec/batch
Epoch: 3/20... Training Step: 8407... Training loss: 1.7242... 0.0624 sec/batch
Epoch: 3/20... Training Step: 8408... Training loss: 1.7091... 0.0864 sec/batch
Epoch: 3/20... Training Step: 8409... Training loss: 1.6065... 0.0807 sec/batch
Epoch: 3/20... Training Step: 8410... Training loss: 1.7665... 0.0668 sec/batch
Epoch: 3/20... Training Step: 8411... Training loss: 1.6480... 0.0583 sec/batch
Epoch: 3/20... Training Step: 8412... Training loss: 1.7310... 0.0534 sec/batch
Epoch: 3/20... Training Step: 8413... Training loss: 1.7292... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8414... Training loss: 1.7692... 0.0571 sec/batch
Epoch: 3/20... Training Step: 8415... Training loss: 1.7257... 0.0444 sec/batch
Epoch: 3/20... Training Step: 8416... Training loss: 1.5741... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8417... Training loss: 1.6456... 0.0537 sec/batch
Epoch: 3/20... Training Step: 8418... Training loss: 1.6340... 0.0604 sec/batch
Epoch: 3/20... Training Step: 8419... Training loss: 1.6768... 0.0581 sec/batch
Epoch: 3/20... Training Step: 8420... Training loss: 1.7847... 0.0551 sec/batch
Epoch: 3/20... Training Step: 8421... Training loss: 1.6647... 0.0502 sec/batch
Epoch: 3/20... Training Step: 8422... Training loss: 1.6987... 0.0583 sec/batch
Epoch: 3/20... Training Step: 8423... Training loss: 1.7789... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8424... Training loss: 1.6908... 0.0493 sec/batch
Epoch: 3/20... Training Step: 8425... Training loss: 1.7626... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8426... Training loss: 1.8289... 0.0478 sec/batch
Epoch: 3/20... Training Step: 8427... Training loss: 1.8809... 0.0532 sec/batch
Epoch: 3/20... Training Step: 8428... Training loss: 1.7120... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8429... Training loss: 1.7562... 0.0560 sec/batch
Epoch: 3/20... Training Step: 8430... Training loss: 1.7654... 0.0648 sec/batch
Epoch: 3/20... Training Step: 8431... Training loss: 1.6665... 0.0555 sec/batch
Epoch: 3/20... Training Step: 8432... Training loss: 1.7787... 0.0492 sec/batch
Epoch: 3/20... Training Step: 8433... Training loss: 1.8271... 0.0576 sec/batch
Epoch: 3/20... Training Step: 8434... Training loss: 1.8339... 0.0518 sec/batch
Epoch: 3/20... Training Step: 8435... Training loss: 1.5967... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8436... Training loss: 1.6400... 0.0747 sec/batch
Epoch: 3/20... Training Step: 8437... Training loss: 1.5688... 0.0629 sec/batch
Epoch: 3/20... Training Step: 8438... Training loss: 1.7366... 0.0579 sec/batch
Epoch: 3/20... Training Step: 8439... Training loss: 1.6121... 0.0577 sec/batch
Epoch: 3/20... Training Step: 8440... Training loss: 1.8983... 0.0455 sec/batch
Epoch: 3/20... Training Step: 8441... Training loss: 1.6033... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8442... Training loss: 1.7525... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8443... Training loss: 1.7388... 0.0505 sec/batch
Epoch: 3/20... Training Step: 8444... Training loss: 1.6943... 0.0541 sec/batch
Epoch: 3/20... Training Step: 8445... Training loss: 1.7032... 0.0508 sec/batch
Epoch: 3/20... Training Step: 8446... Training loss: 1.7367... 0.0452 sec/batch
Epoch: 3/20... Training Step: 8447... Training loss: 1.7534... 0.0669 sec/batch
Epoch: 3/20... Training Step: 8448... Training loss: 1.8848... 0.0449 sec/batch
Epoch: 3/20... Training Step: 8449... Training loss: 1.8619... 0.0418 sec/batch
Epoch: 3/20... Training Step: 8450... Training loss: 1.7997... 0.0416 sec/batch
Epoch: 3/20... Training Step: 8451... Training loss: 1.8129... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8452... Training loss: 1.9066... 0.0407 sec/batch
Epoch: 3/20... Training Step: 8453... Training loss: 1.7047... 0.0535 sec/batch
Epoch: 3/20... Training Step: 8454... Training loss: 1.7322... 0.0483 sec/batch
Epoch: 3/20... Training Step: 8455... Training loss: 1.8202... 0.0441 sec/batch
Epoch: 3/20... Training Step: 8456... Training loss: 1.8289... 0.0407 sec/batch
Epoch: 3/20... Training Step: 8457... Training loss: 1.8067... 0.0441 sec/batch
Epoch: 3/20... Training Step: 8458... Training loss: 1.7344... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8459... Training loss: 1.7225... 0.0453 sec/batch
Epoch: 3/20... Training Step: 8460... Training loss: 1.6227... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8461... Training loss: 1.6954... 0.0452 sec/batch
Epoch: 3/20... Training Step: 8462... Training loss: 1.7554... 0.0471 sec/batch
Epoch: 3/20... Training Step: 8463... Training loss: 1.7220... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8464... Training loss: 1.6932... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8465... Training loss: 1.7472... 0.0441 sec/batch
Epoch: 3/20... Training Step: 8466... Training loss: 1.6475... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8467... Training loss: 1.6783... 0.0436 sec/batch
Epoch: 3/20... Training Step: 8468... Training loss: 1.6881... 0.0479 sec/batch
Epoch: 3/20... Training Step: 8469... Training loss: 1.7809... 0.0567 sec/batch
Epoch: 3/20... Training Step: 8470... Training loss: 1.7587... 0.0557 sec/batch
Epoch: 3/20... Training Step: 8471... Training loss: 1.8691... 0.0540 sec/batch
Epoch: 3/20... Training Step: 8472... Training loss: 1.6977... 0.0577 sec/batch
Epoch: 3/20... Training Step: 8473... Training loss: 1.6215... 0.0455 sec/batch
Epoch: 3/20... Training Step: 8474... Training loss: 1.6663... 0.0411 sec/batch
Epoch: 3/20... Training Step: 8475... Training loss: 1.7126... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8476... Training loss: 1.6745... 0.0430 sec/batch
Epoch: 3/20... Training Step: 8477... Training loss: 1.5353... 0.0428 sec/batch
Epoch: 3/20... Training Step: 8478... Training loss: 1.6272... 0.0495 sec/batch
Epoch: 3/20... Training Step: 8479... Training loss: 1.5607... 0.0442 sec/batch
Epoch: 3/20... Training Step: 8480... Training loss: 1.6313... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8481... Training loss: 1.6137... 0.0501 sec/batch
Epoch: 3/20... Training Step: 8482... Training loss: 1.6572... 0.0568 sec/batch
Epoch: 3/20... Training Step: 8483... Training loss: 1.7441... 0.0687 sec/batch
Epoch: 3/20... Training Step: 8484... Training loss: 1.6577... 0.0693 sec/batch
Epoch: 3/20... Training Step: 8485... Training loss: 1.6855... 0.0546 sec/batch
Epoch: 3/20... Training Step: 8486... Training loss: 1.6461... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8487... Training loss: 1.7116... 0.0948 sec/batch
Epoch: 3/20... Training Step: 8488... Training loss: 1.6205... 0.0754 sec/batch
Epoch: 3/20... Training Step: 8489... Training loss: 1.6642... 0.0640 sec/batch
Epoch: 3/20... Training Step: 8490... Training loss: 1.5400... 0.0579 sec/batch
Epoch: 3/20... Training Step: 8491... Training loss: 1.5853... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8492... Training loss: 1.7223... 0.0604 sec/batch
Epoch: 3/20... Training Step: 8493... Training loss: 1.5808... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8494... Training loss: 1.6452... 0.0548 sec/batch
Epoch: 3/20... Training Step: 8495... Training loss: 1.7407... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8496... Training loss: 1.7297... 0.0450 sec/batch
Epoch: 3/20... Training Step: 8497... Training loss: 1.5969... 0.0624 sec/batch
Epoch: 3/20... Training Step: 8498... Training loss: 1.6106... 0.0612 sec/batch
Epoch: 3/20... Training Step: 8499... Training loss: 1.7167... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8500... Training loss: 1.7758... 0.0575 sec/batch
Epoch: 3/20... Training Step: 8501... Training loss: 1.6817... 0.0521 sec/batch
Epoch: 3/20... Training Step: 8502... Training loss: 1.7760... 0.0465 sec/batch
Epoch: 3/20... Training Step: 8503... Training loss: 1.6554... 0.0647 sec/batch
Epoch: 3/20... Training Step: 8504... Training loss: 1.6035... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8505... Training loss: 1.5657... 0.0437 sec/batch
Epoch: 3/20... Training Step: 8506... Training loss: 1.7146... 0.0582 sec/batch
Epoch: 3/20... Training Step: 8507... Training loss: 1.6286... 0.0491 sec/batch
Epoch: 3/20... Training Step: 8508... Training loss: 1.7505... 0.0615 sec/batch
Epoch: 3/20... Training Step: 8509... Training loss: 1.7387... 0.0505 sec/batch
Epoch: 3/20... Training Step: 8510... Training loss: 1.7310... 0.0592 sec/batch
Epoch: 3/20... Training Step: 8511... Training loss: 1.7185... 0.0476 sec/batch
Epoch: 3/20... Training Step: 8512... Training loss: 1.6733... 0.0578 sec/batch
Epoch: 3/20... Training Step: 8513... Training loss: 1.5597... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8514... Training loss: 1.5991... 0.0533 sec/batch
Epoch: 3/20... Training Step: 8515... Training loss: 1.6208... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8516... Training loss: 1.6664... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8517... Training loss: 1.7673... 0.0498 sec/batch
Epoch: 3/20... Training Step: 8518... Training loss: 1.7789... 0.0529 sec/batch
Epoch: 3/20... Training Step: 8519... Training loss: 1.7148... 0.0606 sec/batch
Epoch: 3/20... Training Step: 8520... Training loss: 1.8467... 0.0615 sec/batch
Epoch: 3/20... Training Step: 8521... Training loss: 1.7882... 0.0546 sec/batch
Epoch: 3/20... Training Step: 8522... Training loss: 1.6364... 0.0551 sec/batch
Epoch: 3/20... Training Step: 8523... Training loss: 1.6694... 0.0608 sec/batch
Epoch: 3/20... Training Step: 8524... Training loss: 1.7260... 0.0635 sec/batch
Epoch: 3/20... Training Step: 8525... Training loss: 1.5548... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8526... Training loss: 1.8281... 0.0627 sec/batch
Epoch: 3/20... Training Step: 8527... Training loss: 1.5592... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8528... Training loss: 1.6939... 0.0574 sec/batch
Epoch: 3/20... Training Step: 8529... Training loss: 2.0315... 0.0470 sec/batch
Epoch: 3/20... Training Step: 8530... Training loss: 1.8327... 0.0557 sec/batch
Epoch: 3/20... Training Step: 8531... Training loss: 1.8227... 0.0440 sec/batch
Epoch: 3/20... Training Step: 8532... Training loss: 1.6518... 0.0607 sec/batch
Epoch: 3/20... Training Step: 8533... Training loss: 1.6492... 0.0530 sec/batch
Epoch: 3/20... Training Step: 8534... Training loss: 1.8508... 0.0557 sec/batch
Epoch: 3/20... Training Step: 8535... Training loss: 1.8712... 0.0588 sec/batch
Epoch: 3/20... Training Step: 8536... Training loss: 1.8211... 0.0563 sec/batch
Epoch: 3/20... Training Step: 8537... Training loss: 1.5564... 0.0621 sec/batch
Epoch: 3/20... Training Step: 8538... Training loss: 1.7752... 0.0657 sec/batch
Epoch: 3/20... Training Step: 8539... Training loss: 1.7438... 0.0681 sec/batch
Epoch: 3/20... Training Step: 8540... Training loss: 1.5833... 0.0476 sec/batch
Epoch: 3/20... Training Step: 8541... Training loss: 1.8905... 0.0597 sec/batch
Epoch: 3/20... Training Step: 8542... Training loss: 1.6171... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8543... Training loss: 1.7132... 0.0452 sec/batch
Epoch: 3/20... Training Step: 8544... Training loss: 1.6231... 0.0594 sec/batch
Epoch: 3/20... Training Step: 8545... Training loss: 1.7068... 0.0506 sec/batch
Epoch: 3/20... Training Step: 8546... Training loss: 1.7051... 0.0485 sec/batch
Epoch: 3/20... Training Step: 8547... Training loss: 1.6130... 0.0607 sec/batch
Epoch: 3/20... Training Step: 8548... Training loss: 1.7887... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8549... Training loss: 1.7947... 0.0447 sec/batch
Epoch: 3/20... Training Step: 8550... Training loss: 1.7151... 0.0464 sec/batch
Epoch: 3/20... Training Step: 8551... Training loss: 1.7991... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8552... Training loss: 1.6509... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8553... Training loss: 1.6738... 0.0433 sec/batch
Epoch: 3/20... Training Step: 8554... Training loss: 1.8167... 0.0489 sec/batch
Epoch: 3/20... Training Step: 8555... Training loss: 1.6612... 0.0607 sec/batch
Epoch: 3/20... Training Step: 8556... Training loss: 1.7471... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8557... Training loss: 1.6128... 0.0533 sec/batch
Epoch: 3/20... Training Step: 8558... Training loss: 1.6580... 0.0641 sec/batch
Epoch: 3/20... Training Step: 8559... Training loss: 1.6272... 0.0571 sec/batch
Epoch: 3/20... Training Step: 8560... Training loss: 1.5580... 0.0496 sec/batch
Epoch: 3/20... Training Step: 8561... Training loss: 1.6324... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8562... Training loss: 1.6693... 0.0452 sec/batch
Epoch: 3/20... Training Step: 8563... Training loss: 1.7389... 0.0572 sec/batch
Epoch: 3/20... Training Step: 8564... Training loss: 1.7438... 0.0619 sec/batch
Epoch: 3/20... Training Step: 8565... Training loss: 1.5780... 0.0538 sec/batch
Epoch: 3/20... Training Step: 8566... Training loss: 1.6969... 0.0600 sec/batch
Epoch: 3/20... Training Step: 8567... Training loss: 1.6624... 0.0473 sec/batch
Epoch: 3/20... Training Step: 8568... Training loss: 1.5863... 0.0593 sec/batch
Epoch: 3/20... Training Step: 8569... Training loss: 1.6937... 0.0524 sec/batch
Epoch: 3/20... Training Step: 8570... Training loss: 1.7986... 0.0513 sec/batch
Epoch: 3/20... Training Step: 8571... Training loss: 1.6960... 0.0580 sec/batch
Epoch: 3/20... Training Step: 8572... Training loss: 1.5766... 0.0417 sec/batch
Epoch: 3/20... Training Step: 8573... Training loss: 1.6149... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8574... Training loss: 1.7368... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8575... Training loss: 1.8755... 0.0473 sec/batch
Epoch: 3/20... Training Step: 8576... Training loss: 1.7977... 0.0453 sec/batch
Epoch: 3/20... Training Step: 8577... Training loss: 1.7584... 0.0613 sec/batch
Epoch: 3/20... Training Step: 8578... Training loss: 1.7151... 0.0565 sec/batch
Epoch: 3/20... Training Step: 8579... Training loss: 1.6948... 0.0498 sec/batch
Epoch: 3/20... Training Step: 8580... Training loss: 1.6401... 0.0577 sec/batch
Epoch: 3/20... Training Step: 8581... Training loss: 1.7192... 0.0454 sec/batch
Epoch: 3/20... Training Step: 8582... Training loss: 1.6987... 0.0462 sec/batch
Epoch: 3/20... Training Step: 8583... Training loss: 1.6397... 0.0484 sec/batch
Epoch: 3/20... Training Step: 8584... Training loss: 1.5741... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8585... Training loss: 1.6459... 0.0552 sec/batch
Epoch: 3/20... Training Step: 8586... Training loss: 1.8379... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8587... Training loss: 1.6694... 0.0539 sec/batch
Epoch: 3/20... Training Step: 8588... Training loss: 1.7589... 0.0470 sec/batch
Epoch: 3/20... Training Step: 8589... Training loss: 1.7150... 0.0519 sec/batch
Epoch: 3/20... Training Step: 8590... Training loss: 1.5614... 0.0472 sec/batch
Epoch: 3/20... Training Step: 8591... Training loss: 1.7510... 0.0602 sec/batch
Epoch: 3/20... Training Step: 8592... Training loss: 1.8082... 0.0630 sec/batch
Epoch: 3/20... Training Step: 8593... Training loss: 1.8398... 0.0546 sec/batch
Epoch: 3/20... Training Step: 8594... Training loss: 1.6284... 0.0495 sec/batch
Epoch: 3/20... Training Step: 8595... Training loss: 1.7156... 0.0560 sec/batch
Epoch: 3/20... Training Step: 8596... Training loss: 1.6620... 0.0541 sec/batch
Epoch: 3/20... Training Step: 8597... Training loss: 1.6911... 0.0592 sec/batch
Epoch: 3/20... Training Step: 8598... Training loss: 1.7927... 0.0510 sec/batch
Epoch: 3/20... Training Step: 8599... Training loss: 1.8034... 0.0433 sec/batch
Epoch: 3/20... Training Step: 8600... Training loss: 1.7801... 0.0526 sec/batch
Epoch: 3/20... Training Step: 8601... Training loss: 1.7521... 0.0432 sec/batch
Epoch: 3/20... Training Step: 8602... Training loss: 1.7194... 0.0410 sec/batch
Epoch: 3/20... Training Step: 8603... Training loss: 1.7260... 0.0469 sec/batch
Epoch: 3/20... Training Step: 8604... Training loss: 1.9039... 0.0469 sec/batch
Epoch: 3/20... Training Step: 8605... Training loss: 1.7679... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8606... Training loss: 1.7448... 0.0503 sec/batch
Epoch: 3/20... Training Step: 8607... Training loss: 1.7865... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8608... Training loss: 1.6299... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8609... Training loss: 1.6175... 0.0437 sec/batch
Epoch: 3/20... Training Step: 8610... Training loss: 1.7469... 0.0430 sec/batch
Epoch: 3/20... Training Step: 8611... Training loss: 1.7331... 0.0434 sec/batch
Epoch: 3/20... Training Step: 8612... Training loss: 1.7302... 0.0436 sec/batch
Epoch: 3/20... Training Step: 8613... Training loss: 1.7753... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8614... Training loss: 1.7560... 0.0429 sec/batch
Epoch: 3/20... Training Step: 8615... Training loss: 1.6295... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8616... Training loss: 1.6507... 0.0455 sec/batch
Epoch: 3/20... Training Step: 8617... Training loss: 1.8696... 0.0489 sec/batch
Epoch: 3/20... Training Step: 8618... Training loss: 1.7841... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8619... Training loss: 1.7365... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8620... Training loss: 1.8684... 0.0411 sec/batch
Epoch: 3/20... Training Step: 8621... Training loss: 1.6767... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8622... Training loss: 1.6251... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8623... Training loss: 1.7512... 0.0616 sec/batch
Epoch: 3/20... Training Step: 8624... Training loss: 1.6607... 0.0475 sec/batch
Epoch: 3/20... Training Step: 8625... Training loss: 1.6714... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8626... Training loss: 1.6755... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8627... Training loss: 1.8156... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8628... Training loss: 1.6172... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8629... Training loss: 1.6067... 0.0609 sec/batch
Epoch: 3/20... Training Step: 8630... Training loss: 1.7134... 0.0475 sec/batch
Epoch: 3/20... Training Step: 8631... Training loss: 1.7406... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8632... Training loss: 1.7679... 0.0506 sec/batch
Epoch: 3/20... Training Step: 8633... Training loss: 1.6713... 0.0492 sec/batch
Epoch: 3/20... Training Step: 8634... Training loss: 1.7929... 0.0514 sec/batch
Epoch: 3/20... Training Step: 8635... Training loss: 1.6458... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8636... Training loss: 1.8360... 0.0421 sec/batch
Epoch: 3/20... Training Step: 8637... Training loss: 1.7919... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8638... Training loss: 1.7301... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8639... Training loss: 1.7350... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8640... Training loss: 1.6845... 0.0441 sec/batch
Epoch: 3/20... Training Step: 8641... Training loss: 1.7788... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8642... Training loss: 1.7001... 0.0584 sec/batch
Epoch: 3/20... Training Step: 8643... Training loss: 1.6708... 0.0581 sec/batch
Epoch: 3/20... Training Step: 8644... Training loss: 1.7211... 0.0485 sec/batch
Epoch: 3/20... Training Step: 8645... Training loss: 1.6831... 0.0528 sec/batch
Epoch: 3/20... Training Step: 8646... Training loss: 1.7763... 0.0658 sec/batch
Epoch: 3/20... Training Step: 8647... Training loss: 1.7482... 0.0540 sec/batch
Epoch: 3/20... Training Step: 8648... Training loss: 1.7144... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8649... Training loss: 1.7260... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8650... Training loss: 1.7914... 0.0437 sec/batch
Epoch: 3/20... Training Step: 8651... Training loss: 1.6918... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8652... Training loss: 1.7698... 0.0560 sec/batch
Epoch: 3/20... Training Step: 8653... Training loss: 1.8347... 0.0435 sec/batch
Epoch: 3/20... Training Step: 8654... Training loss: 1.6353... 0.0465 sec/batch
Epoch: 3/20... Training Step: 8655... Training loss: 1.6570... 0.0434 sec/batch
Epoch: 3/20... Training Step: 8656... Training loss: 1.6166... 0.0521 sec/batch
Epoch: 3/20... Training Step: 8657... Training loss: 1.7691... 0.0593 sec/batch
Epoch: 3/20... Training Step: 8658... Training loss: 1.5925... 0.0457 sec/batch
Epoch: 3/20... Training Step: 8659... Training loss: 1.6934... 0.0520 sec/batch
Epoch: 3/20... Training Step: 8660... Training loss: 1.8151... 0.0432 sec/batch
Epoch: 3/20... Training Step: 8661... Training loss: 1.7138... 0.0446 sec/batch
Epoch: 3/20... Training Step: 8662... Training loss: 1.5648... 0.0462 sec/batch
Epoch: 3/20... Training Step: 8663... Training loss: 1.6980... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8664... Training loss: 1.7853... 0.0519 sec/batch
Epoch: 3/20... Training Step: 8665... Training loss: 1.6386... 0.0452 sec/batch
Epoch: 3/20... Training Step: 8666... Training loss: 1.6815... 0.0473 sec/batch
Epoch: 3/20... Training Step: 8667... Training loss: 1.7009... 0.0491 sec/batch
Epoch: 3/20... Training Step: 8668... Training loss: 1.6796... 0.0455 sec/batch
Epoch: 3/20... Training Step: 8669... Training loss: 1.6763... 0.0521 sec/batch
Epoch: 3/20... Training Step: 8670... Training loss: 1.5835... 0.0515 sec/batch
Epoch: 3/20... Training Step: 8671... Training loss: 1.7194... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8672... Training loss: 1.6480... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8673... Training loss: 1.7010... 0.0506 sec/batch
Epoch: 3/20... Training Step: 8674... Training loss: 1.6188... 0.0415 sec/batch
Epoch: 3/20... Training Step: 8675... Training loss: 1.5727... 0.0461 sec/batch
Epoch: 3/20... Training Step: 8676... Training loss: 1.8416... 0.0550 sec/batch
Epoch: 3/20... Training Step: 8677... Training loss: 1.5835... 0.0457 sec/batch
Epoch: 3/20... Training Step: 8678... Training loss: 1.7792... 0.0444 sec/batch
Epoch: 3/20... Training Step: 8679... Training loss: 1.7645... 0.0492 sec/batch
Epoch: 3/20... Training Step: 8680... Training loss: 1.7042... 0.0587 sec/batch
Epoch: 3/20... Training Step: 8681... Training loss: 1.7021... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8682... Training loss: 1.5655... 0.0450 sec/batch
Epoch: 3/20... Training Step: 8683... Training loss: 1.5698... 0.0634 sec/batch
Epoch: 3/20... Training Step: 8684... Training loss: 1.6468... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8685... Training loss: 1.6480... 0.0537 sec/batch
Epoch: 3/20... Training Step: 8686... Training loss: 1.7627... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8687... Training loss: 1.7172... 0.0462 sec/batch
Epoch: 3/20... Training Step: 8688... Training loss: 1.7261... 0.0462 sec/batch
Epoch: 3/20... Training Step: 8689... Training loss: 1.6050... 0.0456 sec/batch
Epoch: 3/20... Training Step: 8690... Training loss: 1.7122... 0.0668 sec/batch
Epoch: 3/20... Training Step: 8691... Training loss: 1.6429... 0.0535 sec/batch
Epoch: 3/20... Training Step: 8692... Training loss: 1.9514... 0.0460 sec/batch
Epoch: 3/20... Training Step: 8693... Training loss: 1.7182... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8694... Training loss: 1.6192... 0.0469 sec/batch
Epoch: 3/20... Training Step: 8695... Training loss: 1.6852... 0.0577 sec/batch
Epoch: 3/20... Training Step: 8696... Training loss: 1.6327... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8697... Training loss: 1.6615... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8698... Training loss: 1.5979... 0.0616 sec/batch
Epoch: 3/20... Training Step: 8699... Training loss: 1.7503... 0.0519 sec/batch
Epoch: 3/20... Training Step: 8700... Training loss: 1.6480... 0.0443 sec/batch
Epoch: 3/20... Training Step: 8701... Training loss: 1.6082... 0.0454 sec/batch
Epoch: 3/20... Training Step: 8702... Training loss: 1.8381... 0.0535 sec/batch
Epoch: 3/20... Training Step: 8703... Training loss: 1.7622... 0.0546 sec/batch
Epoch: 3/20... Training Step: 8704... Training loss: 1.7654... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8705... Training loss: 1.7362... 0.0485 sec/batch
Epoch: 3/20... Training Step: 8706... Training loss: 1.7940... 0.0515 sec/batch
Epoch: 3/20... Training Step: 8707... Training loss: 1.7458... 0.0552 sec/batch
Epoch: 3/20... Training Step: 8708... Training loss: 1.7459... 0.0543 sec/batch
Epoch: 3/20... Training Step: 8709... Training loss: 1.8367... 0.0536 sec/batch
Epoch: 3/20... Training Step: 8710... Training loss: 1.7542... 0.0542 sec/batch
Epoch: 3/20... Training Step: 8711... Training loss: 1.6723... 0.0523 sec/batch
Epoch: 3/20... Training Step: 8712... Training loss: 1.6792... 0.0563 sec/batch
Epoch: 3/20... Training Step: 8713... Training loss: 1.7264... 0.0472 sec/batch
Epoch: 3/20... Training Step: 8714... Training loss: 1.8493... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8715... Training loss: 1.6846... 0.0478 sec/batch
Epoch: 3/20... Training Step: 8716... Training loss: 1.7943... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8717... Training loss: 1.7883... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8718... Training loss: 1.7609... 0.0421 sec/batch
Epoch: 3/20... Training Step: 8719... Training loss: 1.6833... 0.0448 sec/batch
Epoch: 3/20... Training Step: 8720... Training loss: 1.7931... 0.0504 sec/batch
Epoch: 3/20... Training Step: 8721... Training loss: 1.7102... 0.0474 sec/batch
Epoch: 3/20... Training Step: 8722... Training loss: 1.7273... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8723... Training loss: 1.7184... 0.0547 sec/batch
Epoch: 3/20... Training Step: 8724... Training loss: 1.6883... 0.0440 sec/batch
Epoch: 3/20... Training Step: 8725... Training loss: 1.6873... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8726... Training loss: 1.7254... 0.0442 sec/batch
Epoch: 3/20... Training Step: 8727... Training loss: 1.7405... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8728... Training loss: 1.5366... 0.0486 sec/batch
Epoch: 3/20... Training Step: 8729... Training loss: 1.7469... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8730... Training loss: 1.6931... 0.0676 sec/batch
Epoch: 3/20... Training Step: 8731... Training loss: 1.7341... 0.0518 sec/batch
Epoch: 3/20... Training Step: 8732... Training loss: 1.5944... 0.0478 sec/batch
Epoch: 3/20... Training Step: 8733... Training loss: 1.7166... 0.0513 sec/batch
Epoch: 3/20... Training Step: 8734... Training loss: 1.6578... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8735... Training loss: 1.6516... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8736... Training loss: 1.6059... 0.0505 sec/batch
Epoch: 3/20... Training Step: 8737... Training loss: 1.7702... 0.0565 sec/batch
Epoch: 3/20... Training Step: 8738... Training loss: 1.5640... 0.0478 sec/batch
Epoch: 3/20... Training Step: 8739... Training loss: 1.6961... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8740... Training loss: 1.6736... 0.0537 sec/batch
Epoch: 3/20... Training Step: 8741... Training loss: 1.6834... 0.0586 sec/batch
Epoch: 3/20... Training Step: 8742... Training loss: 1.6367... 0.0514 sec/batch
Epoch: 3/20... Training Step: 8743... Training loss: 1.7584... 0.0505 sec/batch
Epoch: 3/20... Training Step: 8744... Training loss: 1.8764... 0.0464 sec/batch
Epoch: 3/20... Training Step: 8745... Training loss: 1.8000... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8746... Training loss: 1.7617... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8747... Training loss: 1.7188... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8748... Training loss: 1.7120... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8749... Training loss: 1.6610... 0.0555 sec/batch
Epoch: 3/20... Training Step: 8750... Training loss: 1.6419... 0.0489 sec/batch
Epoch: 3/20... Training Step: 8751... Training loss: 1.6524... 0.0429 sec/batch
Epoch: 3/20... Training Step: 8752... Training loss: 1.7072... 0.0523 sec/batch
Epoch: 3/20... Training Step: 8753... Training loss: 1.6612... 0.0579 sec/batch
Epoch: 3/20... Training Step: 8754... Training loss: 1.6559... 0.0450 sec/batch
Epoch: 3/20... Training Step: 8755... Training loss: 1.6736... 0.0546 sec/batch
Epoch: 3/20... Training Step: 8756... Training loss: 1.6045... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8757... Training loss: 1.5621... 0.0520 sec/batch
Epoch: 3/20... Training Step: 8758... Training loss: 1.6265... 0.0507 sec/batch
Epoch: 3/20... Training Step: 8759... Training loss: 1.6115... 0.0435 sec/batch
Epoch: 3/20... Training Step: 8760... Training loss: 1.7237... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8761... Training loss: 1.7571... 0.0530 sec/batch
Epoch: 3/20... Training Step: 8762... Training loss: 1.7796... 0.0491 sec/batch
Epoch: 3/20... Training Step: 8763... Training loss: 1.6162... 0.0466 sec/batch
Epoch: 3/20... Training Step: 8764... Training loss: 1.6272... 0.0425 sec/batch
Epoch: 3/20... Training Step: 8765... Training loss: 1.5294... 0.0578 sec/batch
Epoch: 3/20... Training Step: 8766... Training loss: 1.6903... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8767... Training loss: 1.7272... 0.0547 sec/batch
Epoch: 3/20... Training Step: 8768... Training loss: 1.7143... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8769... Training loss: 1.6198... 0.0430 sec/batch
Epoch: 3/20... Training Step: 8770... Training loss: 1.6778... 0.0564 sec/batch
Epoch: 3/20... Training Step: 8771... Training loss: 1.7784... 0.0574 sec/batch
Epoch: 3/20... Training Step: 8772... Training loss: 1.6881... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8773... Training loss: 1.6366... 0.0532 sec/batch
Epoch: 3/20... Training Step: 8774... Training loss: 1.6591... 0.0590 sec/batch
Epoch: 3/20... Training Step: 8775... Training loss: 1.7787... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8776... Training loss: 1.5123... 0.0442 sec/batch
Epoch: 3/20... Training Step: 8777... Training loss: 1.6776... 0.0520 sec/batch
Epoch: 3/20... Training Step: 8778... Training loss: 1.5633... 0.0558 sec/batch
Epoch: 3/20... Training Step: 8779... Training loss: 1.8054... 0.0552 sec/batch
Epoch: 3/20... Training Step: 8780... Training loss: 1.6601... 0.0464 sec/batch
Epoch: 3/20... Training Step: 8781... Training loss: 1.6989... 0.0468 sec/batch
Epoch: 3/20... Training Step: 8782... Training loss: 1.6728... 0.0402 sec/batch
Epoch: 3/20... Training Step: 8783... Training loss: 1.7314... 0.0443 sec/batch
Epoch: 3/20... Training Step: 8784... Training loss: 1.6905... 0.0450 sec/batch
Epoch: 3/20... Training Step: 8785... Training loss: 1.6118... 0.0426 sec/batch
Epoch: 3/20... Training Step: 8786... Training loss: 1.7481... 0.0471 sec/batch
Epoch: 3/20... Training Step: 8787... Training loss: 1.7550... 0.0495 sec/batch
Epoch: 3/20... Training Step: 8788... Training loss: 1.7240... 0.0457 sec/batch
Epoch: 3/20... Training Step: 8789... Training loss: 1.6755... 0.0544 sec/batch
Epoch: 3/20... Training Step: 8790... Training loss: 1.7735... 0.0515 sec/batch
Epoch: 3/20... Training Step: 8791... Training loss: 1.8782... 0.0426 sec/batch
Epoch: 3/20... Training Step: 8792... Training loss: 1.5355... 0.0527 sec/batch
Epoch: 3/20... Training Step: 8793... Training loss: 1.7463... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8794... Training loss: 1.5640... 0.0502 sec/batch
Epoch: 3/20... Training Step: 8795... Training loss: 1.7991... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8796... Training loss: 1.7339... 0.0505 sec/batch
Epoch: 3/20... Training Step: 8797... Training loss: 1.8784... 0.0495 sec/batch
Epoch: 3/20... Training Step: 8798... Training loss: 1.6767... 0.0412 sec/batch
Epoch: 3/20... Training Step: 8799... Training loss: 1.7191... 0.0530 sec/batch
Epoch: 3/20... Training Step: 8800... Training loss: 1.8542... 0.0553 sec/batch
Epoch: 3/20... Training Step: 8801... Training loss: 1.7113... 0.0588 sec/batch
Epoch: 3/20... Training Step: 8802... Training loss: 1.7825... 0.0411 sec/batch
Epoch: 3/20... Training Step: 8803... Training loss: 1.8164... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8804... Training loss: 1.7084... 0.0428 sec/batch
Epoch: 3/20... Training Step: 8805... Training loss: 1.8148... 0.0524 sec/batch
Epoch: 3/20... Training Step: 8806... Training loss: 1.8965... 0.0470 sec/batch
Epoch: 3/20... Training Step: 8807... Training loss: 1.7683... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8808... Training loss: 1.7498... 0.0448 sec/batch
Epoch: 3/20... Training Step: 8809... Training loss: 1.8013... 0.0533 sec/batch
Epoch: 3/20... Training Step: 8810... Training loss: 1.7769... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8811... Training loss: 1.6538... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8812... Training loss: 1.6691... 0.0550 sec/batch
Epoch: 3/20... Training Step: 8813... Training loss: 1.6788... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8814... Training loss: 1.6898... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8815... Training loss: 1.8140... 0.0631 sec/batch
Epoch: 3/20... Training Step: 8816... Training loss: 1.6255... 0.0576 sec/batch
Epoch: 3/20... Training Step: 8817... Training loss: 1.7042... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8818... Training loss: 1.8856... 0.0520 sec/batch
Epoch: 3/20... Training Step: 8819... Training loss: 1.6295... 0.0497 sec/batch
Epoch: 3/20... Training Step: 8820... Training loss: 1.5928... 0.0622 sec/batch
Epoch: 3/20... Training Step: 8821... Training loss: 1.9519... 0.0578 sec/batch
Epoch: 3/20... Training Step: 8822... Training loss: 1.8708... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8823... Training loss: 1.7143... 0.0440 sec/batch
Epoch: 3/20... Training Step: 8824... Training loss: 1.7717... 0.0456 sec/batch
Epoch: 3/20... Training Step: 8825... Training loss: 1.6276... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8826... Training loss: 1.7199... 0.0610 sec/batch
Epoch: 3/20... Training Step: 8827... Training loss: 1.7288... 0.0434 sec/batch
Epoch: 3/20... Training Step: 8828... Training loss: 1.6037... 0.0627 sec/batch
Epoch: 3/20... Training Step: 8829... Training loss: 1.7036... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8830... Training loss: 1.6891... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8831... Training loss: 1.7506... 0.0540 sec/batch
Epoch: 3/20... Training Step: 8832... Training loss: 1.7514... 0.0462 sec/batch
Epoch: 3/20... Training Step: 8833... Training loss: 1.8694... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8834... Training loss: 1.8175... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8835... Training loss: 1.7417... 0.0671 sec/batch
Epoch: 3/20... Training Step: 8836... Training loss: 1.6868... 0.0530 sec/batch
Epoch: 3/20... Training Step: 8837... Training loss: 1.8354... 0.0445 sec/batch
Epoch: 3/20... Training Step: 8838... Training loss: 1.6732... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8839... Training loss: 1.7264... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8840... Training loss: 1.7627... 0.0489 sec/batch
Epoch: 3/20... Training Step: 8841... Training loss: 1.5642... 0.0587 sec/batch
Epoch: 3/20... Training Step: 8842... Training loss: 1.5879... 0.0478 sec/batch
Epoch: 3/20... Training Step: 8843... Training loss: 1.7069... 0.0560 sec/batch
Epoch: 3/20... Training Step: 8844... Training loss: 1.6429... 0.0484 sec/batch
Epoch: 3/20... Training Step: 8845... Training loss: 1.6819... 0.0535 sec/batch
Epoch: 3/20... Training Step: 8846... Training loss: 1.8368... 0.0463 sec/batch
Epoch: 3/20... Training Step: 8847... Training loss: 1.7918... 0.0463 sec/batch
Epoch: 3/20... Training Step: 8848... Training loss: 1.8541... 0.0555 sec/batch
Epoch: 3/20... Training Step: 8849... Training loss: 1.8340... 0.0614 sec/batch
Epoch: 3/20... Training Step: 8850... Training loss: 1.8242... 0.0495 sec/batch
Epoch: 3/20... Training Step: 8851... Training loss: 1.7466... 0.0517 sec/batch
Epoch: 3/20... Training Step: 8852... Training loss: 1.7401... 0.0449 sec/batch
Epoch: 3/20... Training Step: 8853... Training loss: 1.8552... 0.0486 sec/batch
Epoch: 3/20... Training Step: 8854... Training loss: 1.6812... 0.0501 sec/batch
Epoch: 3/20... Training Step: 8855... Training loss: 1.6628... 0.0529 sec/batch
Epoch: 3/20... Training Step: 8856... Training loss: 1.7422... 0.0485 sec/batch
Epoch: 3/20... Training Step: 8857... Training loss: 1.7404... 0.0470 sec/batch
Epoch: 3/20... Training Step: 8858... Training loss: 1.5852... 0.0555 sec/batch
Epoch: 3/20... Training Step: 8859... Training loss: 1.5864... 0.0481 sec/batch
Epoch: 3/20... Training Step: 8860... Training loss: 1.6882... 0.0583 sec/batch
Epoch: 3/20... Training Step: 8861... Training loss: 1.6364... 0.0636 sec/batch
Epoch: 3/20... Training Step: 8862... Training loss: 1.7492... 0.0538 sec/batch
Epoch: 3/20... Training Step: 8863... Training loss: 1.7893... 0.0461 sec/batch
Epoch: 3/20... Training Step: 8864... Training loss: 1.7023... 0.0575 sec/batch
Epoch: 3/20... Training Step: 8865... Training loss: 1.7850... 0.0541 sec/batch
Epoch: 3/20... Training Step: 8866... Training loss: 1.7277... 0.0514 sec/batch
Epoch: 3/20... Training Step: 8867... Training loss: 1.6846... 0.0558 sec/batch
Epoch: 3/20... Training Step: 8868... Training loss: 1.6618... 0.0439 sec/batch
Epoch: 3/20... Training Step: 8869... Training loss: 1.6218... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8870... Training loss: 1.7921... 0.0500 sec/batch
Epoch: 3/20... Training Step: 8871... Training loss: 1.7082... 0.0512 sec/batch
Epoch: 3/20... Training Step: 8872... Training loss: 1.6161... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8873... Training loss: 1.6209... 0.0513 sec/batch
Epoch: 3/20... Training Step: 8874... Training loss: 1.7159... 0.0575 sec/batch
Epoch: 3/20... Training Step: 8875... Training loss: 1.6123... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8876... Training loss: 1.7083... 0.0501 sec/batch
Epoch: 3/20... Training Step: 8877... Training loss: 1.6010... 0.0459 sec/batch
Epoch: 3/20... Training Step: 8878... Training loss: 1.6069... 0.0469 sec/batch
Epoch: 3/20... Training Step: 8879... Training loss: 1.6611... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8880... Training loss: 1.7093... 0.0502 sec/batch
Epoch: 3/20... Training Step: 8881... Training loss: 1.5885... 0.0475 sec/batch
Epoch: 3/20... Training Step: 8882... Training loss: 1.5229... 0.0511 sec/batch
Epoch: 3/20... Training Step: 8883... Training loss: 1.5923... 0.0434 sec/batch
Epoch: 3/20... Training Step: 8884... Training loss: 1.7516... 0.0553 sec/batch
Epoch: 3/20... Training Step: 8885... Training loss: 1.6773... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8886... Training loss: 1.6796... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8887... Training loss: 1.5765... 0.0448 sec/batch
Epoch: 3/20... Training Step: 8888... Training loss: 1.6924... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8889... Training loss: 1.7414... 0.0542 sec/batch
Epoch: 3/20... Training Step: 8890... Training loss: 1.7144... 0.0491 sec/batch
Epoch: 3/20... Training Step: 8891... Training loss: 1.6813... 0.0544 sec/batch
Epoch: 3/20... Training Step: 8892... Training loss: 1.7917... 0.0516 sec/batch
Epoch: 3/20... Training Step: 8893... Training loss: 1.7144... 0.0613 sec/batch
Epoch: 3/20... Training Step: 8894... Training loss: 1.6848... 0.0611 sec/batch
Epoch: 3/20... Training Step: 8895... Training loss: 1.7107... 0.0461 sec/batch
Epoch: 3/20... Training Step: 8896... Training loss: 1.6131... 0.0682 sec/batch
Epoch: 3/20... Training Step: 8897... Training loss: 1.7785... 0.0435 sec/batch
Epoch: 3/20... Training Step: 8898... Training loss: 1.7577... 0.0439 sec/batch
Epoch: 3/20... Training Step: 8899... Training loss: 1.6206... 0.0508 sec/batch
Epoch: 3/20... Training Step: 8900... Training loss: 1.7774... 0.0491 sec/batch
Epoch: 3/20... Training Step: 8901... Training loss: 1.5564... 0.0447 sec/batch
Epoch: 3/20... Training Step: 8902... Training loss: 1.6908... 0.0575 sec/batch
Epoch: 3/20... Training Step: 8903... Training loss: 1.5749... 0.0490 sec/batch
Epoch: 3/20... Training Step: 8904... Training loss: 1.6153... 0.0552 sec/batch
Epoch: 3/20... Training Step: 8905... Training loss: 1.7169... 0.0530 sec/batch
Epoch: 3/20... Training Step: 8906... Training loss: 1.7710... 0.0610 sec/batch
Epoch: 3/20... Training Step: 8907... Training loss: 1.6208... 0.0534 sec/batch
Epoch: 3/20... Training Step: 8908... Training loss: 1.7027... 0.0475 sec/batch
Epoch: 3/20... Training Step: 8909... Training loss: 1.6785... 0.0464 sec/batch
Epoch: 3/20... Training Step: 8910... Training loss: 1.5542... 0.0444 sec/batch
Epoch: 3/20... Training Step: 8911... Training loss: 1.6551... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8912... Training loss: 1.7668... 0.0514 sec/batch
Epoch: 3/20... Training Step: 8913... Training loss: 1.5724... 0.0479 sec/batch
Epoch: 3/20... Training Step: 8914... Training loss: 1.6368... 0.0492 sec/batch
Epoch: 3/20... Training Step: 8915... Training loss: 1.6648... 0.0428 sec/batch
Epoch: 3/20... Training Step: 8916... Training loss: 1.8393... 0.0439 sec/batch
Epoch: 3/20... Training Step: 8917... Training loss: 1.9226... 0.0458 sec/batch
Epoch: 3/20... Training Step: 8918... Training loss: 1.6747... 0.0435 sec/batch
Epoch: 3/20... Training Step: 8919... Training loss: 1.7382... 0.0417 sec/batch
Epoch: 3/20... Training Step: 8920... Training loss: 1.6677... 0.0595 sec/batch
Epoch: 3/20... Training Step: 8921... Training loss: 1.7508... 0.0541 sec/batch
Epoch: 3/20... Training Step: 8922... Training loss: 1.5686... 0.0477 sec/batch
Epoch: 3/20... Training Step: 8923... Training loss: 1.8753... 0.0540 sec/batch
Epoch: 3/20... Training Step: 8924... Training loss: 1.7038... 0.0513 sec/batch
Epoch: 3/20... Training Step: 8925... Training loss: 1.5976... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8926... Training loss: 1.6822... 0.0562 sec/batch
Epoch: 3/20... Training Step: 8927... Training loss: 1.6206... 0.0521 sec/batch
Epoch: 3/20... Training Step: 8928... Training loss: 1.6345... 0.0510 sec/batch
Epoch: 3/20... Training Step: 8929... Training loss: 1.7254... 0.0522 sec/batch
Epoch: 3/20... Training Step: 8930... Training loss: 1.6111... 0.0497 sec/batch
Epoch: 3/20... Training Step: 8931... Training loss: 1.5060... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8932... Training loss: 1.7352... 0.0467 sec/batch
Epoch: 3/20... Training Step: 8933... Training loss: 1.7672... 0.0571 sec/batch
Epoch: 3/20... Training Step: 8934... Training loss: 1.7645... 0.0526 sec/batch
Epoch: 3/20... Training Step: 8935... Training loss: 1.7159... 0.0476 sec/batch
Epoch: 3/20... Training Step: 8936... Training loss: 1.7912... 0.0476 sec/batch
Epoch: 3/20... Training Step: 8937... Training loss: 1.6429... 0.0428 sec/batch
Epoch: 3/20... Training Step: 8938... Training loss: 1.7515... 0.0539 sec/batch
Epoch: 3/20... Training Step: 8939... Training loss: 1.6602... 0.0494 sec/batch
Epoch: 3/20... Training Step: 8940... Training loss: 1.7699... 0.0430 sec/batch
Epoch: 3/20... Training Step: 8941... Training loss: 1.7064... 0.0595 sec/batch
Epoch: 3/20... Training Step: 8942... Training loss: 1.5681... 0.0472 sec/batch
Epoch: 3/20... Training Step: 8943... Training loss: 1.6515... 0.0604 sec/batch
Epoch: 3/20... Training Step: 8944... Training loss: 1.5214... 0.0588 sec/batch
Epoch: 3/20... Training Step: 8945... Training loss: 1.5714... 0.0529 sec/batch
Epoch: 3/20... Training Step: 8946... Training loss: 1.6531... 0.0422 sec/batch
Epoch: 3/20... Training Step: 8947... Training loss: 1.6981... 0.0453 sec/batch
Epoch: 3/20... Training Step: 8948... Training loss: 1.7799... 0.0570 sec/batch
Epoch: 3/20... Training Step: 8949... Training loss: 1.7218... 0.0544 sec/batch
Epoch: 3/20... Training Step: 8950... Training loss: 1.7670... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8951... Training loss: 1.6792... 0.0501 sec/batch
Epoch: 3/20... Training Step: 8952... Training loss: 1.7261... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8953... Training loss: 1.7016... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8954... Training loss: 1.7676... 0.0483 sec/batch
Epoch: 3/20... Training Step: 8955... Training loss: 1.5688... 0.0475 sec/batch
Epoch: 3/20... Training Step: 8956... Training loss: 1.8267... 0.0403 sec/batch
Epoch: 3/20... Training Step: 8957... Training loss: 1.7445... 0.0488 sec/batch
Epoch: 3/20... Training Step: 8958... Training loss: 1.6656... 0.0618 sec/batch
Epoch: 3/20... Training Step: 8959... Training loss: 1.6285... 0.0597 sec/batch
Epoch: 3/20... Training Step: 8960... Training loss: 1.5341... 0.0444 sec/batch
Epoch: 3/20... Training Step: 8961... Training loss: 1.6161... 0.0545 sec/batch
Epoch: 3/20... Training Step: 8962... Training loss: 1.6871... 0.0487 sec/batch
Epoch: 3/20... Training Step: 8963... Training loss: 1.5469... 0.0559 sec/batch
Epoch: 3/20... Training Step: 8964... Training loss: 1.6369... 0.0433 sec/batch
Epoch: 3/20... Training Step: 8965... Training loss: 1.7183... 0.0431 sec/batch
Epoch: 3/20... Training Step: 8966... Training loss: 1.6933... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8967... Training loss: 1.7616... 0.0571 sec/batch
Epoch: 3/20... Training Step: 8968... Training loss: 1.6019... 0.0419 sec/batch
Epoch: 3/20... Training Step: 8969... Training loss: 1.7964... 0.0457 sec/batch
Epoch: 3/20... Training Step: 8970... Training loss: 1.6829... 0.0414 sec/batch
Epoch: 3/20... Training Step: 8971... Training loss: 1.6849... 0.0425 sec/batch
Epoch: 3/20... Training Step: 8972... Training loss: 1.6385... 0.0499 sec/batch
Epoch: 3/20... Training Step: 8973... Training loss: 1.6250... 0.0425 sec/batch
Epoch: 3/20... Training Step: 8974... Training loss: 1.6493... 0.0461 sec/batch
Epoch: 3/20... Training Step: 8975... Training loss: 1.7641... 0.0549 sec/batch
Epoch: 3/20... Training Step: 8976... Training loss: 1.7926... 0.0480 sec/batch
Epoch: 3/20... Training Step: 8977... Training loss: 1.6582... 0.0420 sec/batch
Epoch: 3/20... Training Step: 8978... Training loss: 1.6115... 0.0536 sec/batch
Epoch: 3/20... Training Step: 8979... Training loss: 1.6587... 0.0426 sec/batch
Epoch: 3/20... Training Step: 8980... Training loss: 1.7045... 0.0602 sec/batch
Epoch: 3/20... Training Step: 8981... Training loss: 1.6661... 0.0551 sec/batch
Epoch: 3/20... Training Step: 8982... Training loss: 1.6390... 0.0442 sec/batch
Epoch: 3/20... Training Step: 8983... Training loss: 1.7695... 0.0525 sec/batch
Epoch: 3/20... Training Step: 8984... Training loss: 1.7597... 0.0497 sec/batch
Epoch: 3/20... Training Step: 8985... Training loss: 1.8034... 0.0402 sec/batch
Epoch: 3/20... Training Step: 8986... Training loss: 1.6941... 0.0552 sec/batch
Epoch: 3/20... Training Step: 8987... Training loss: 1.6425... 0.0474 sec/batch
Epoch: 3/20... Training Step: 8988... Training loss: 1.5888... 0.0510 sec/batch
Epoch: 3/20... Training Step: 8989... Training loss: 1.5754... 0.0440 sec/batch
Epoch: 3/20... Training Step: 8990... Training loss: 1.7113... 0.0510 sec/batch
Epoch: 3/20... Training Step: 8991... Training loss: 1.7696... 0.0453 sec/batch
Epoch: 3/20... Training Step: 8992... Training loss: 1.7069... 0.0440 sec/batch
Epoch: 3/20... Training Step: 8993... Training loss: 1.6596... 0.0534 sec/batch
Epoch: 3/20... Training Step: 8994... Training loss: 1.6667... 0.0503 sec/batch
Epoch: 3/20... Training Step: 8995... Training loss: 1.6758... 0.0564 sec/batch
Epoch: 3/20... Training Step: 8996... Training loss: 1.6487... 0.0540 sec/batch
Epoch: 3/20... Training Step: 8997... Training loss: 1.5127... 0.0497 sec/batch
Epoch: 3/20... Training Step: 8998... Training loss: 1.7127... 0.0531 sec/batch
Epoch: 3/20... Training Step: 8999... Training loss: 1.7662... 0.0503 sec/batch
Epoch: 3/20... Training Step: 9000... Training loss: 1.7434... 0.0486 sec/batch
Epoch: 3/20... Training Step: 9001... Training loss: 1.6845... 0.0608 sec/batch
Epoch: 3/20... Training Step: 9002... Training loss: 1.6505... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9003... Training loss: 1.6238... 0.0617 sec/batch
Epoch: 3/20... Training Step: 9004... Training loss: 1.6557... 0.0605 sec/batch
Epoch: 3/20... Training Step: 9005... Training loss: 1.6550... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9006... Training loss: 1.7603... 0.0463 sec/batch
Epoch: 3/20... Training Step: 9007... Training loss: 1.6288... 0.0522 sec/batch
Epoch: 3/20... Training Step: 9008... Training loss: 1.7864... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9009... Training loss: 1.6487... 0.0556 sec/batch
Epoch: 3/20... Training Step: 9010... Training loss: 1.6184... 0.0458 sec/batch
Epoch: 3/20... Training Step: 9011... Training loss: 1.6723... 0.0478 sec/batch
Epoch: 3/20... Training Step: 9012... Training loss: 1.5883... 0.0501 sec/batch
Epoch: 3/20... Training Step: 9013... Training loss: 1.4856... 0.0523 sec/batch
Epoch: 3/20... Training Step: 9014... Training loss: 1.6053... 0.0514 sec/batch
Epoch: 3/20... Training Step: 9015... Training loss: 1.6179... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9016... Training loss: 1.6232... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9017... Training loss: 1.6163... 0.0504 sec/batch
Epoch: 3/20... Training Step: 9018... Training loss: 1.7304... 0.0454 sec/batch
Epoch: 3/20... Training Step: 9019... Training loss: 1.6362... 0.0580 sec/batch
Epoch: 3/20... Training Step: 9020... Training loss: 1.5657... 0.0478 sec/batch
Epoch: 3/20... Training Step: 9021... Training loss: 1.6414... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9022... Training loss: 1.5939... 0.0579 sec/batch
Epoch: 3/20... Training Step: 9023... Training loss: 1.7885... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9024... Training loss: 1.4221... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9025... Training loss: 1.5612... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9026... Training loss: 1.6000... 0.0589 sec/batch
Epoch: 3/20... Training Step: 9027... Training loss: 1.6205... 0.0477 sec/batch
Epoch: 3/20... Training Step: 9028... Training loss: 1.5250... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9029... Training loss: 1.6852... 0.0446 sec/batch
Epoch: 3/20... Training Step: 9030... Training loss: 1.8861... 0.0526 sec/batch
Epoch: 3/20... Training Step: 9031... Training loss: 1.7722... 0.0577 sec/batch
Epoch: 3/20... Training Step: 9032... Training loss: 1.8336... 0.0520 sec/batch
Epoch: 3/20... Training Step: 9033... Training loss: 1.7004... 0.0507 sec/batch
Epoch: 3/20... Training Step: 9034... Training loss: 1.7131... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9035... Training loss: 1.6955... 0.0539 sec/batch
Epoch: 3/20... Training Step: 9036... Training loss: 1.7611... 0.0539 sec/batch
Epoch: 3/20... Training Step: 9037... Training loss: 1.6097... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9038... Training loss: 1.5188... 0.0448 sec/batch
Epoch: 3/20... Training Step: 9039... Training loss: 1.7423... 0.0516 sec/batch
Epoch: 3/20... Training Step: 9040... Training loss: 1.7061... 0.0599 sec/batch
Epoch: 3/20... Training Step: 9041... Training loss: 1.5958... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9042... Training loss: 1.5931... 0.0522 sec/batch
Epoch: 3/20... Training Step: 9043... Training loss: 1.5223... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9044... Training loss: 1.7723... 0.0457 sec/batch
Epoch: 3/20... Training Step: 9045... Training loss: 1.6686... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9046... Training loss: 1.6137... 0.0432 sec/batch
Epoch: 3/20... Training Step: 9047... Training loss: 1.6239... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9048... Training loss: 1.7679... 0.0475 sec/batch
Epoch: 3/20... Training Step: 9049... Training loss: 1.5618... 0.0536 sec/batch
Epoch: 3/20... Training Step: 9050... Training loss: 1.6454... 0.0670 sec/batch
Epoch: 3/20... Training Step: 9051... Training loss: 1.7663... 0.0465 sec/batch
Epoch: 3/20... Training Step: 9052... Training loss: 1.7084... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9053... Training loss: 1.5943... 0.0561 sec/batch
Epoch: 3/20... Training Step: 9054... Training loss: 1.6742... 0.0525 sec/batch
Epoch: 3/20... Training Step: 9055... Training loss: 1.6982... 0.0494 sec/batch
Epoch: 3/20... Training Step: 9056... Training loss: 1.8270... 0.0527 sec/batch
Epoch: 3/20... Training Step: 9057... Training loss: 1.6639... 0.0532 sec/batch
Epoch: 3/20... Training Step: 9058... Training loss: 1.6910... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9059... Training loss: 1.5995... 0.0557 sec/batch
Epoch: 3/20... Training Step: 9060... Training loss: 1.5382... 0.0528 sec/batch
Epoch: 3/20... Training Step: 9061... Training loss: 1.6997... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9062... Training loss: 1.7131... 0.0603 sec/batch
Epoch: 3/20... Training Step: 9063... Training loss: 1.6548... 0.0476 sec/batch
Epoch: 3/20... Training Step: 9064... Training loss: 1.5776... 0.0571 sec/batch
Epoch: 3/20... Training Step: 9065... Training loss: 1.7231... 0.0490 sec/batch
Epoch: 3/20... Training Step: 9066... Training loss: 1.6372... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9067... Training loss: 1.6888... 0.0588 sec/batch
Epoch: 3/20... Training Step: 9068... Training loss: 1.5777... 0.0566 sec/batch
Epoch: 3/20... Training Step: 9069... Training loss: 1.5436... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9070... Training loss: 1.5526... 0.0568 sec/batch
Epoch: 3/20... Training Step: 9071... Training loss: 1.6718... 0.0445 sec/batch
Epoch: 3/20... Training Step: 9072... Training loss: 1.5022... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9073... Training loss: 1.6543... 0.0501 sec/batch
Epoch: 3/20... Training Step: 9074... Training loss: 1.7437... 0.0529 sec/batch
Epoch: 3/20... Training Step: 9075... Training loss: 1.6079... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9076... Training loss: 1.7585... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9077... Training loss: 1.7224... 0.0477 sec/batch
Epoch: 3/20... Training Step: 9078... Training loss: 1.8218... 0.0527 sec/batch
Epoch: 3/20... Training Step: 9079... Training loss: 1.7511... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9080... Training loss: 1.8587... 0.0511 sec/batch
Epoch: 3/20... Training Step: 9081... Training loss: 1.6367... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9082... Training loss: 1.7257... 0.0433 sec/batch
Epoch: 3/20... Training Step: 9083... Training loss: 1.5680... 0.0432 sec/batch
Epoch: 3/20... Training Step: 9084... Training loss: 1.9690... 0.0531 sec/batch
Epoch: 3/20... Training Step: 9085... Training loss: 1.8368... 0.0482 sec/batch
Epoch: 3/20... Training Step: 9086... Training loss: 1.6676... 0.0492 sec/batch
Epoch: 3/20... Training Step: 9087... Training loss: 1.8429... 0.0526 sec/batch
Epoch: 3/20... Training Step: 9088... Training loss: 1.7558... 0.0444 sec/batch
Epoch: 3/20... Training Step: 9089... Training loss: 1.7814... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9090... Training loss: 1.6720... 0.0516 sec/batch
Epoch: 3/20... Training Step: 9091... Training loss: 1.7068... 0.0529 sec/batch
Epoch: 3/20... Training Step: 9092... Training loss: 1.7190... 0.0456 sec/batch
Epoch: 3/20... Training Step: 9093... Training loss: 1.7391... 0.0438 sec/batch
Epoch: 3/20... Training Step: 9094... Training loss: 1.7488... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9095... Training loss: 1.7206... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9096... Training loss: 1.5961... 0.0480 sec/batch
Epoch: 3/20... Training Step: 9097... Training loss: 1.6775... 0.0476 sec/batch
Epoch: 3/20... Training Step: 9098... Training loss: 1.5542... 0.0463 sec/batch
Epoch: 3/20... Training Step: 9099... Training loss: 1.7380... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9100... Training loss: 1.5366... 0.0500 sec/batch
Epoch: 3/20... Training Step: 9101... Training loss: 1.6744... 0.0435 sec/batch
Epoch: 3/20... Training Step: 9102... Training loss: 1.6946... 0.0605 sec/batch
Epoch: 3/20... Training Step: 9103... Training loss: 1.7230... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9104... Training loss: 1.7489... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9105... Training loss: 1.7422... 0.0406 sec/batch
Epoch: 3/20... Training Step: 9106... Training loss: 1.7845... 0.0634 sec/batch
Epoch: 3/20... Training Step: 9107... Training loss: 1.6549... 0.0504 sec/batch
Epoch: 3/20... Training Step: 9108... Training loss: 1.7455... 0.0440 sec/batch
Epoch: 3/20... Training Step: 9109... Training loss: 1.7337... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9110... Training loss: 1.6660... 0.0518 sec/batch
Epoch: 3/20... Training Step: 9111... Training loss: 1.5026... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9112... Training loss: 1.7384... 0.0484 sec/batch
Epoch: 3/20... Training Step: 9113... Training loss: 1.6070... 0.0504 sec/batch
Epoch: 3/20... Training Step: 9114... Training loss: 1.5547... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9115... Training loss: 1.7053... 0.0529 sec/batch
Epoch: 3/20... Training Step: 9116... Training loss: 1.6493... 0.0447 sec/batch
Epoch: 3/20... Training Step: 9117... Training loss: 1.6931... 0.0503 sec/batch
Epoch: 3/20... Training Step: 9118... Training loss: 1.4894... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9119... Training loss: 1.4891... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9120... Training loss: 1.6647... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9121... Training loss: 1.7630... 0.0522 sec/batch
Epoch: 3/20... Training Step: 9122... Training loss: 1.6819... 0.0562 sec/batch
Epoch: 3/20... Training Step: 9123... Training loss: 1.6336... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9124... Training loss: 1.7636... 0.0570 sec/batch
Epoch: 3/20... Training Step: 9125... Training loss: 1.6818... 0.0463 sec/batch
Epoch: 3/20... Training Step: 9126... Training loss: 1.6681... 0.0475 sec/batch
Epoch: 3/20... Training Step: 9127... Training loss: 1.7986... 0.0531 sec/batch
Epoch: 3/20... Training Step: 9128... Training loss: 1.7891... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9129... Training loss: 1.5895... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9130... Training loss: 1.7197... 0.0479 sec/batch
Epoch: 3/20... Training Step: 9131... Training loss: 1.6971... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9132... Training loss: 1.7170... 0.0551 sec/batch
Epoch: 3/20... Training Step: 9133... Training loss: 1.6632... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9134... Training loss: 1.7126... 0.0424 sec/batch
Epoch: 3/20... Training Step: 9135... Training loss: 1.8528... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9136... Training loss: 1.7134... 0.0585 sec/batch
Epoch: 3/20... Training Step: 9137... Training loss: 1.5459... 0.0458 sec/batch
Epoch: 3/20... Training Step: 9138... Training loss: 1.6589... 0.0525 sec/batch
Epoch: 3/20... Training Step: 9139... Training loss: 1.7209... 0.0499 sec/batch
Epoch: 3/20... Training Step: 9140... Training loss: 1.6051... 0.0444 sec/batch
Epoch: 3/20... Training Step: 9141... Training loss: 1.6446... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9142... Training loss: 1.7595... 0.0463 sec/batch
Epoch: 3/20... Training Step: 9143... Training loss: 1.6736... 0.0520 sec/batch
Epoch: 3/20... Training Step: 9144... Training loss: 1.6318... 0.0474 sec/batch
Epoch: 3/20... Training Step: 9145... Training loss: 1.6934... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9146... Training loss: 1.6231... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9147... Training loss: 1.5637... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9148... Training loss: 1.7264... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9149... Training loss: 1.7743... 0.0467 sec/batch
Epoch: 3/20... Training Step: 9150... Training loss: 1.6406... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9151... Training loss: 1.7109... 0.0500 sec/batch
Epoch: 3/20... Training Step: 9152... Training loss: 1.7274... 0.0519 sec/batch
Epoch: 3/20... Training Step: 9153... Training loss: 1.8688... 0.0522 sec/batch
Epoch: 3/20... Training Step: 9154... Training loss: 1.6560... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9155... Training loss: 1.6298... 0.0584 sec/batch
Epoch: 3/20... Training Step: 9156... Training loss: 1.7993... 0.0527 sec/batch
Epoch: 3/20... Training Step: 9157... Training loss: 1.8732... 0.0475 sec/batch
Epoch: 3/20... Training Step: 9158... Training loss: 1.8075... 0.0433 sec/batch
Epoch: 3/20... Training Step: 9159... Training loss: 1.7020... 0.0475 sec/batch
Epoch: 3/20... Training Step: 9160... Training loss: 1.6004... 0.0533 sec/batch
Epoch: 3/20... Training Step: 9161... Training loss: 1.7396... 0.0507 sec/batch
Epoch: 3/20... Training Step: 9162... Training loss: 1.7714... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9163... Training loss: 1.5617... 0.0584 sec/batch
Epoch: 3/20... Training Step: 9164... Training loss: 1.6141... 0.0486 sec/batch
Epoch: 3/20... Training Step: 9165... Training loss: 1.7185... 0.0564 sec/batch
Epoch: 3/20... Training Step: 9166... Training loss: 1.6910... 0.0492 sec/batch
Epoch: 3/20... Training Step: 9167... Training loss: 1.8753... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9168... Training loss: 1.7513... 0.0514 sec/batch
Epoch: 3/20... Training Step: 9169... Training loss: 1.5859... 0.0455 sec/batch
Epoch: 3/20... Training Step: 9170... Training loss: 1.7752... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9171... Training loss: 1.5534... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9172... Training loss: 1.7898... 0.0478 sec/batch
Epoch: 3/20... Training Step: 9173... Training loss: 1.7092... 0.0477 sec/batch
Epoch: 3/20... Training Step: 9174... Training loss: 1.8111... 0.0535 sec/batch
Epoch: 3/20... Training Step: 9175... Training loss: 1.6125... 0.0565 sec/batch
Epoch: 3/20... Training Step: 9176... Training loss: 1.6626... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9177... Training loss: 1.6139... 0.0484 sec/batch
Epoch: 3/20... Training Step: 9178... Training loss: 1.4972... 0.0515 sec/batch
Epoch: 3/20... Training Step: 9179... Training loss: 1.5696... 0.0530 sec/batch
Epoch: 3/20... Training Step: 9180... Training loss: 1.6021... 0.0454 sec/batch
Epoch: 3/20... Training Step: 9181... Training loss: 1.5023... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9182... Training loss: 1.6215... 0.0533 sec/batch
Epoch: 3/20... Training Step: 9183... Training loss: 1.5809... 0.0486 sec/batch
Epoch: 3/20... Training Step: 9184... Training loss: 1.7900... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9185... Training loss: 1.6030... 0.0570 sec/batch
Epoch: 3/20... Training Step: 9186... Training loss: 1.6619... 0.0549 sec/batch
Epoch: 3/20... Training Step: 9187... Training loss: 1.7437... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9188... Training loss: 1.6356... 0.0545 sec/batch
Epoch: 3/20... Training Step: 9189... Training loss: 1.6963... 0.0605 sec/batch
Epoch: 3/20... Training Step: 9190... Training loss: 1.5916... 0.0528 sec/batch
Epoch: 3/20... Training Step: 9191... Training loss: 1.7773... 0.0599 sec/batch
Epoch: 3/20... Training Step: 9192... Training loss: 1.6062... 0.0548 sec/batch
Epoch: 3/20... Training Step: 9193... Training loss: 1.6219... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9194... Training loss: 1.7582... 0.0427 sec/batch
Epoch: 3/20... Training Step: 9195... Training loss: 1.6320... 0.0536 sec/batch
Epoch: 3/20... Training Step: 9196... Training loss: 1.6835... 0.0491 sec/batch
Epoch: 3/20... Training Step: 9197... Training loss: 1.7744... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9198... Training loss: 1.6597... 0.0605 sec/batch
Epoch: 3/20... Training Step: 9199... Training loss: 1.6946... 0.0496 sec/batch
Epoch: 3/20... Training Step: 9200... Training loss: 1.6248... 0.0577 sec/batch
Epoch: 3/20... Training Step: 9201... Training loss: 1.8501... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9202... Training loss: 1.7723... 0.0454 sec/batch
Epoch: 3/20... Training Step: 9203... Training loss: 1.6090... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9204... Training loss: 1.4894... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9205... Training loss: 1.5746... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9206... Training loss: 1.6608... 0.0435 sec/batch
Epoch: 3/20... Training Step: 9207... Training loss: 1.6660... 0.0517 sec/batch
Epoch: 3/20... Training Step: 9208... Training loss: 1.6532... 0.0555 sec/batch
Epoch: 3/20... Training Step: 9209... Training loss: 1.7151... 0.0506 sec/batch
Epoch: 3/20... Training Step: 9210... Training loss: 1.6892... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9211... Training loss: 1.6465... 0.0503 sec/batch
Epoch: 3/20... Training Step: 9212... Training loss: 1.6977... 0.0530 sec/batch
Epoch: 3/20... Training Step: 9213... Training loss: 1.7384... 0.0530 sec/batch
Epoch: 3/20... Training Step: 9214... Training loss: 1.6097... 0.0496 sec/batch
Epoch: 3/20... Training Step: 9215... Training loss: 1.7380... 0.0503 sec/batch
Epoch: 3/20... Training Step: 9216... Training loss: 1.6348... 0.0519 sec/batch
Epoch: 3/20... Training Step: 9217... Training loss: 1.7194... 0.0462 sec/batch
Epoch: 3/20... Training Step: 9218... Training loss: 1.8056... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9219... Training loss: 1.7887... 0.0520 sec/batch
Epoch: 3/20... Training Step: 9220... Training loss: 1.6039... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9221... Training loss: 1.6733... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9222... Training loss: 1.8183... 0.0447 sec/batch
Epoch: 3/20... Training Step: 9223... Training loss: 1.7431... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9224... Training loss: 1.8296... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9225... Training loss: 1.8503... 0.0460 sec/batch
Epoch: 3/20... Training Step: 9226... Training loss: 1.6356... 0.0427 sec/batch
Epoch: 3/20... Training Step: 9227... Training loss: 1.7396... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9228... Training loss: 1.6750... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9229... Training loss: 1.6981... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9230... Training loss: 1.8873... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9231... Training loss: 1.7815... 0.0414 sec/batch
Epoch: 3/20... Training Step: 9232... Training loss: 1.7873... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9233... Training loss: 1.7454... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9234... Training loss: 1.7633... 0.0577 sec/batch
Epoch: 3/20... Training Step: 9235... Training loss: 1.7638... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9236... Training loss: 1.7804... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9237... Training loss: 1.6557... 0.0462 sec/batch
Epoch: 3/20... Training Step: 9238... Training loss: 1.5843... 0.0559 sec/batch
Epoch: 3/20... Training Step: 9239... Training loss: 1.5878... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9240... Training loss: 1.5964... 0.0414 sec/batch
Epoch: 3/20... Training Step: 9241... Training loss: 1.7641... 0.0424 sec/batch
Epoch: 3/20... Training Step: 9242... Training loss: 1.7520... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9243... Training loss: 1.7158... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9244... Training loss: 1.7608... 0.0442 sec/batch
Epoch: 3/20... Training Step: 9245... Training loss: 1.6723... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9246... Training loss: 1.8954... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9247... Training loss: 1.8721... 0.0541 sec/batch
Epoch: 3/20... Training Step: 9248... Training loss: 1.8152... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9249... Training loss: 1.7421... 0.0490 sec/batch
Epoch: 3/20... Training Step: 9250... Training loss: 1.7157... 0.0551 sec/batch
Epoch: 3/20... Training Step: 9251... Training loss: 1.6890... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9252... Training loss: 1.7580... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9253... Training loss: 1.6506... 0.0502 sec/batch
Epoch: 3/20... Training Step: 9254... Training loss: 1.8173... 0.0467 sec/batch
Epoch: 3/20... Training Step: 9255... Training loss: 1.8334... 0.0456 sec/batch
Epoch: 3/20... Training Step: 9256... Training loss: 1.7724... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9257... Training loss: 1.7187... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9258... Training loss: 1.6473... 0.0482 sec/batch
Epoch: 3/20... Training Step: 9259... Training loss: 1.6753... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9260... Training loss: 1.8005... 0.0427 sec/batch
Epoch: 3/20... Training Step: 9261... Training loss: 1.7860... 0.0471 sec/batch
Epoch: 3/20... Training Step: 9262... Training loss: 1.7355... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9263... Training loss: 1.8173... 0.0457 sec/batch
Epoch: 3/20... Training Step: 9264... Training loss: 1.7516... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9265... Training loss: 1.7364... 0.0486 sec/batch
Epoch: 3/20... Training Step: 9266... Training loss: 1.7124... 0.0455 sec/batch
Epoch: 3/20... Training Step: 9267... Training loss: 1.7627... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9268... Training loss: 1.6978... 0.0559 sec/batch
Epoch: 3/20... Training Step: 9269... Training loss: 1.7579... 0.0507 sec/batch
Epoch: 3/20... Training Step: 9270... Training loss: 1.7394... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9271... Training loss: 1.7651... 0.0497 sec/batch
Epoch: 3/20... Training Step: 9272... Training loss: 1.6928... 0.0525 sec/batch
Epoch: 3/20... Training Step: 9273... Training loss: 1.7348... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9274... Training loss: 1.7401... 0.0516 sec/batch
Epoch: 3/20... Training Step: 9275... Training loss: 1.5897... 0.0547 sec/batch
Epoch: 3/20... Training Step: 9276... Training loss: 1.6594... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9277... Training loss: 1.5511... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9278... Training loss: 1.7037... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9279... Training loss: 1.7089... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9280... Training loss: 1.7337... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9281... Training loss: 1.6035... 0.0552 sec/batch
Epoch: 3/20... Training Step: 9282... Training loss: 1.6329... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9283... Training loss: 1.7359... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9284... Training loss: 1.7818... 0.0496 sec/batch
Epoch: 3/20... Training Step: 9285... Training loss: 1.7487... 0.0508 sec/batch
Epoch: 3/20... Training Step: 9286... Training loss: 1.6953... 0.0454 sec/batch
Epoch: 3/20... Training Step: 9287... Training loss: 1.5945... 0.0491 sec/batch
Epoch: 3/20... Training Step: 9288... Training loss: 1.7304... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9289... Training loss: 1.7973... 0.0552 sec/batch
Epoch: 3/20... Training Step: 9290... Training loss: 1.5644... 0.0459 sec/batch
Epoch: 3/20... Training Step: 9291... Training loss: 1.5164... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9292... Training loss: 1.5606... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9293... Training loss: 1.6879... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9294... Training loss: 1.7144... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9295... Training loss: 1.6114... 0.0444 sec/batch
Epoch: 3/20... Training Step: 9296... Training loss: 1.6155... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9297... Training loss: 1.7251... 0.0393 sec/batch
Epoch: 3/20... Training Step: 9298... Training loss: 1.5873... 0.0489 sec/batch
Epoch: 3/20... Training Step: 9299... Training loss: 1.8200... 0.0529 sec/batch
Epoch: 3/20... Training Step: 9300... Training loss: 1.6207... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9301... Training loss: 1.6213... 0.0544 sec/batch
Epoch: 3/20... Training Step: 9302... Training loss: 1.6218... 0.0488 sec/batch
Epoch: 3/20... Training Step: 9303... Training loss: 1.4870... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9304... Training loss: 1.5768... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9305... Training loss: 1.6186... 0.0647 sec/batch
Epoch: 3/20... Training Step: 9306... Training loss: 1.6268... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9307... Training loss: 1.6295... 0.0465 sec/batch
Epoch: 3/20... Training Step: 9308... Training loss: 1.7217... 0.0479 sec/batch
Epoch: 3/20... Training Step: 9309... Training loss: 1.7912... 0.0457 sec/batch
Epoch: 3/20... Training Step: 9310... Training loss: 1.7308... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9311... Training loss: 1.7133... 0.0642 sec/batch
Epoch: 3/20... Training Step: 9312... Training loss: 1.6407... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9313... Training loss: 1.6845... 0.0528 sec/batch
Epoch: 3/20... Training Step: 9314... Training loss: 1.7474... 0.0477 sec/batch
Epoch: 3/20... Training Step: 9315... Training loss: 1.5764... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9316... Training loss: 1.6985... 0.0531 sec/batch
Epoch: 3/20... Training Step: 9317... Training loss: 1.7092... 0.0519 sec/batch
Epoch: 3/20... Training Step: 9318... Training loss: 1.7296... 0.0460 sec/batch
Epoch: 3/20... Training Step: 9319... Training loss: 1.5534... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9320... Training loss: 1.6364... 0.0547 sec/batch
Epoch: 3/20... Training Step: 9321... Training loss: 1.8474... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9322... Training loss: 1.7299... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9323... Training loss: 1.7974... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9324... Training loss: 1.7284... 0.0512 sec/batch
Epoch: 3/20... Training Step: 9325... Training loss: 1.8077... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9326... Training loss: 1.9594... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9327... Training loss: 1.6675... 0.0692 sec/batch
Epoch: 3/20... Training Step: 9328... Training loss: 1.7336... 0.0512 sec/batch
Epoch: 3/20... Training Step: 9329... Training loss: 1.6855... 0.0541 sec/batch
Epoch: 3/20... Training Step: 9330... Training loss: 1.6875... 0.0501 sec/batch
Epoch: 3/20... Training Step: 9331... Training loss: 1.6834... 0.0583 sec/batch
Epoch: 3/20... Training Step: 9332... Training loss: 1.7018... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9333... Training loss: 1.7168... 0.0613 sec/batch
Epoch: 3/20... Training Step: 9334... Training loss: 1.7240... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9335... Training loss: 1.7574... 0.0458 sec/batch
Epoch: 3/20... Training Step: 9336... Training loss: 1.6547... 0.0568 sec/batch
Epoch: 3/20... Training Step: 9337... Training loss: 1.6702... 0.0499 sec/batch
Epoch: 3/20... Training Step: 9338... Training loss: 1.5979... 0.0548 sec/batch
Epoch: 3/20... Training Step: 9339... Training loss: 1.7873... 0.0534 sec/batch
Epoch: 3/20... Training Step: 9340... Training loss: 1.8228... 0.0490 sec/batch
Epoch: 3/20... Training Step: 9341... Training loss: 1.5780... 0.0440 sec/batch
Epoch: 3/20... Training Step: 9342... Training loss: 1.8142... 0.0530 sec/batch
Epoch: 3/20... Training Step: 9343... Training loss: 1.6305... 0.0690 sec/batch
Epoch: 3/20... Training Step: 9344... Training loss: 1.8468... 0.0546 sec/batch
Epoch: 3/20... Training Step: 9345... Training loss: 1.7852... 0.0554 sec/batch
Epoch: 3/20... Training Step: 9346... Training loss: 1.5746... 0.0496 sec/batch
Epoch: 3/20... Training Step: 9347... Training loss: 1.7339... 0.0478 sec/batch
Epoch: 3/20... Training Step: 9348... Training loss: 1.7883... 0.0535 sec/batch
Epoch: 3/20... Training Step: 9349... Training loss: 1.7342... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9350... Training loss: 1.7177... 0.0551 sec/batch
Epoch: 3/20... Training Step: 9351... Training loss: 1.6440... 0.0524 sec/batch
Epoch: 3/20... Training Step: 9352... Training loss: 1.7619... 0.0502 sec/batch
Epoch: 3/20... Training Step: 9353... Training loss: 1.6263... 0.0612 sec/batch
Epoch: 3/20... Training Step: 9354... Training loss: 1.9954... 0.0665 sec/batch
Epoch: 3/20... Training Step: 9355... Training loss: 1.6828... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9356... Training loss: 1.7466... 0.0470 sec/batch
Epoch: 3/20... Training Step: 9357... Training loss: 1.8238... 0.0621 sec/batch
Epoch: 3/20... Training Step: 9358... Training loss: 1.7298... 0.0501 sec/batch
Epoch: 3/20... Training Step: 9359... Training loss: 1.6890... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9360... Training loss: 1.7776... 0.0662 sec/batch
Epoch: 3/20... Training Step: 9361... Training loss: 1.8700... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9362... Training loss: 1.7069... 0.0627 sec/batch
Epoch: 3/20... Training Step: 9363... Training loss: 1.7776... 0.0517 sec/batch
Epoch: 3/20... Training Step: 9364... Training loss: 1.6928... 0.0445 sec/batch
Epoch: 3/20... Training Step: 9365... Training loss: 1.6494... 0.0536 sec/batch
Epoch: 3/20... Training Step: 9366... Training loss: 1.7236... 0.0498 sec/batch
Epoch: 3/20... Training Step: 9367... Training loss: 1.6174... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9368... Training loss: 1.7618... 0.0446 sec/batch
Epoch: 3/20... Training Step: 9369... Training loss: 1.6964... 0.0475 sec/batch
Epoch: 3/20... Training Step: 9370... Training loss: 1.7093... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9371... Training loss: 1.7103... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9372... Training loss: 1.7784... 0.0449 sec/batch
Epoch: 3/20... Training Step: 9373... Training loss: 1.9081... 0.0583 sec/batch
Epoch: 3/20... Training Step: 9374... Training loss: 1.7129... 0.0533 sec/batch
Epoch: 3/20... Training Step: 9375... Training loss: 1.7460... 0.0445 sec/batch
Epoch: 3/20... Training Step: 9376... Training loss: 1.7005... 0.0479 sec/batch
Epoch: 3/20... Training Step: 9377... Training loss: 1.7032... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9378... Training loss: 1.6907... 0.0488 sec/batch
Epoch: 3/20... Training Step: 9379... Training loss: 1.8336... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9380... Training loss: 1.6897... 0.0588 sec/batch
Epoch: 3/20... Training Step: 9381... Training loss: 1.5535... 0.0502 sec/batch
Epoch: 3/20... Training Step: 9382... Training loss: 1.7675... 0.0473 sec/batch
Epoch: 3/20... Training Step: 9383... Training loss: 1.6745... 0.0574 sec/batch
Epoch: 3/20... Training Step: 9384... Training loss: 1.7326... 0.0453 sec/batch
Epoch: 3/20... Training Step: 9385... Training loss: 1.7307... 0.0590 sec/batch
Epoch: 3/20... Training Step: 9386... Training loss: 1.8434... 0.0530 sec/batch
Epoch: 3/20... Training Step: 9387... Training loss: 1.9319... 0.0474 sec/batch
Epoch: 3/20... Training Step: 9388... Training loss: 1.7071... 0.0495 sec/batch
Epoch: 3/20... Training Step: 9389... Training loss: 1.7589... 0.0467 sec/batch
Epoch: 3/20... Training Step: 9390... Training loss: 1.7323... 0.0505 sec/batch
Epoch: 3/20... Training Step: 9391... Training loss: 1.6979... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9392... Training loss: 1.6978... 0.0482 sec/batch
Epoch: 3/20... Training Step: 9393... Training loss: 1.8296... 0.0506 sec/batch
Epoch: 3/20... Training Step: 9394... Training loss: 1.8081... 0.0550 sec/batch
Epoch: 3/20... Training Step: 9395... Training loss: 1.7994... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9396... Training loss: 1.5336... 0.0813 sec/batch
Epoch: 3/20... Training Step: 9397... Training loss: 1.6754... 0.0457 sec/batch
Epoch: 3/20... Training Step: 9398... Training loss: 1.7232... 0.0474 sec/batch
Epoch: 3/20... Training Step: 9399... Training loss: 1.6812... 0.0561 sec/batch
Epoch: 3/20... Training Step: 9400... Training loss: 1.7375... 0.0492 sec/batch
Epoch: 3/20... Training Step: 9401... Training loss: 1.6889... 0.0462 sec/batch
Epoch: 3/20... Training Step: 9402... Training loss: 1.5944... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9403... Training loss: 1.6832... 0.0480 sec/batch
Epoch: 3/20... Training Step: 9404... Training loss: 1.7189... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9405... Training loss: 1.7284... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9406... Training loss: 1.7265... 0.0444 sec/batch
Epoch: 3/20... Training Step: 9407... Training loss: 1.6254... 0.0494 sec/batch
Epoch: 3/20... Training Step: 9408... Training loss: 1.6416... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9409... Training loss: 1.6725... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9410... Training loss: 1.6568... 0.0476 sec/batch
Epoch: 3/20... Training Step: 9411... Training loss: 1.6987... 0.0600 sec/batch
Epoch: 3/20... Training Step: 9412... Training loss: 1.8589... 0.0533 sec/batch
Epoch: 3/20... Training Step: 9413... Training loss: 1.7195... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9414... Training loss: 1.7630... 0.0592 sec/batch
Epoch: 3/20... Training Step: 9415... Training loss: 1.6502... 0.0590 sec/batch
Epoch: 3/20... Training Step: 9416... Training loss: 1.5620... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9417... Training loss: 1.8268... 0.0436 sec/batch
Epoch: 3/20... Training Step: 9418... Training loss: 1.7555... 0.0508 sec/batch
Epoch: 3/20... Training Step: 9419... Training loss: 1.7250... 0.0555 sec/batch
Epoch: 3/20... Training Step: 9420... Training loss: 1.7148... 0.0549 sec/batch
Epoch: 3/20... Training Step: 9421... Training loss: 1.6923... 0.0490 sec/batch
Epoch: 3/20... Training Step: 9422... Training loss: 1.6869... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9423... Training loss: 1.7174... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9424... Training loss: 1.8062... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9425... Training loss: 1.6959... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9426... Training loss: 1.8061... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9427... Training loss: 1.7201... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9428... Training loss: 1.7651... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9429... Training loss: 1.8574... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9430... Training loss: 1.7518... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9431... Training loss: 1.7502... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9432... Training loss: 1.6234... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9433... Training loss: 1.8511... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9434... Training loss: 1.5599... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9435... Training loss: 1.6710... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9436... Training loss: 1.5750... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9437... Training loss: 1.7074... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9438... Training loss: 1.6965... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9439... Training loss: 1.7503... 0.0543 sec/batch
Epoch: 3/20... Training Step: 9440... Training loss: 1.6861... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9441... Training loss: 1.7023... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9442... Training loss: 1.6765... 0.0511 sec/batch
Epoch: 3/20... Training Step: 9443... Training loss: 1.8084... 0.0445 sec/batch
Epoch: 3/20... Training Step: 9444... Training loss: 1.9218... 0.0490 sec/batch
Epoch: 3/20... Training Step: 9445... Training loss: 1.5842... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9446... Training loss: 1.6357... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9447... Training loss: 1.8037... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9448... Training loss: 1.6618... 0.0395 sec/batch
Epoch: 3/20... Training Step: 9449... Training loss: 1.6058... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9450... Training loss: 1.6768... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9451... Training loss: 1.6589... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9452... Training loss: 1.5635... 0.0391 sec/batch
Epoch: 3/20... Training Step: 9453... Training loss: 1.7004... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9454... Training loss: 1.6147... 0.0436 sec/batch
Epoch: 3/20... Training Step: 9455... Training loss: 1.6523... 0.0388 sec/batch
Epoch: 3/20... Training Step: 9456... Training loss: 1.8042... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9457... Training loss: 1.7737... 0.0389 sec/batch
Epoch: 3/20... Training Step: 9458... Training loss: 1.6915... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9459... Training loss: 1.7900... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9460... Training loss: 1.7262... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9461... Training loss: 1.7112... 0.0414 sec/batch
Epoch: 3/20... Training Step: 9462... Training loss: 1.8418... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9463... Training loss: 1.7258... 0.0389 sec/batch
Epoch: 3/20... Training Step: 9464... Training loss: 1.7280... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9465... Training loss: 1.7851... 0.0394 sec/batch
Epoch: 3/20... Training Step: 9466... Training loss: 1.6797... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9467... Training loss: 1.4955... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9468... Training loss: 1.7465... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9469... Training loss: 1.7593... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9470... Training loss: 1.7147... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9471... Training loss: 1.7980... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9472... Training loss: 1.8853... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9473... Training loss: 1.6551... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9474... Training loss: 1.6764... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9475... Training loss: 1.7602... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9476... Training loss: 1.6057... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9477... Training loss: 1.6764... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9478... Training loss: 1.7554... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9479... Training loss: 1.9572... 0.0419 sec/batch
Epoch: 3/20... Training Step: 9480... Training loss: 1.7964... 0.0394 sec/batch
Epoch: 3/20... Training Step: 9481... Training loss: 1.8062... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9482... Training loss: 1.6553... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9483... Training loss: 1.7249... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9484... Training loss: 1.8746... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9485... Training loss: 1.5031... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9486... Training loss: 1.8436... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9487... Training loss: 1.6031... 0.0548 sec/batch
Epoch: 3/20... Training Step: 9488... Training loss: 1.4869... 0.0473 sec/batch
Epoch: 3/20... Training Step: 9489... Training loss: 1.4483... 0.0458 sec/batch
Epoch: 3/20... Training Step: 9490... Training loss: 1.6392... 0.0532 sec/batch
Epoch: 3/20... Training Step: 9491... Training loss: 1.5902... 0.0449 sec/batch
Epoch: 3/20... Training Step: 9492... Training loss: 1.7206... 0.0394 sec/batch
Epoch: 3/20... Training Step: 9493... Training loss: 1.7440... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9494... Training loss: 1.7748... 0.0392 sec/batch
Epoch: 3/20... Training Step: 9495... Training loss: 1.7520... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9496... Training loss: 1.7735... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9497... Training loss: 1.7674... 0.0392 sec/batch
Epoch: 3/20... Training Step: 9498... Training loss: 1.7218... 0.0395 sec/batch
Epoch: 3/20... Training Step: 9499... Training loss: 1.7103... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9500... Training loss: 1.7520... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9501... Training loss: 1.6439... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9502... Training loss: 1.7755... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9503... Training loss: 1.7130... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9504... Training loss: 1.7142... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9505... Training loss: 1.7236... 0.0390 sec/batch
Epoch: 3/20... Training Step: 9506... Training loss: 1.8683... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9507... Training loss: 1.6996... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9508... Training loss: 1.6229... 0.0395 sec/batch
Epoch: 3/20... Training Step: 9509... Training loss: 1.6484... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9510... Training loss: 1.6990... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9511... Training loss: 1.6961... 0.0393 sec/batch
Epoch: 3/20... Training Step: 9512... Training loss: 1.6540... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9513... Training loss: 1.7227... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9514... Training loss: 1.5978... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9515... Training loss: 1.7486... 0.0392 sec/batch
Epoch: 3/20... Training Step: 9516... Training loss: 1.7380... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9517... Training loss: 1.6399... 0.0441 sec/batch
Epoch: 3/20... Training Step: 9518... Training loss: 1.9184... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9519... Training loss: 1.7974... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9520... Training loss: 1.7097... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9521... Training loss: 1.8536... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9522... Training loss: 1.7934... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9523... Training loss: 1.7209... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9524... Training loss: 1.6771... 0.0406 sec/batch
Epoch: 3/20... Training Step: 9525... Training loss: 1.7271... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9526... Training loss: 1.6623... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9527... Training loss: 1.7203... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9528... Training loss: 1.6406... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9529... Training loss: 1.7983... 0.0435 sec/batch
Epoch: 3/20... Training Step: 9530... Training loss: 1.7665... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9531... Training loss: 1.6406... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9532... Training loss: 1.7478... 0.0442 sec/batch
Epoch: 3/20... Training Step: 9533... Training loss: 1.6784... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9534... Training loss: 1.7571... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9535... Training loss: 1.7625... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9536... Training loss: 1.6995... 0.0438 sec/batch
Epoch: 3/20... Training Step: 9537... Training loss: 1.6142... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9538... Training loss: 1.7018... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9539... Training loss: 1.6953... 0.0395 sec/batch
Epoch: 3/20... Training Step: 9540... Training loss: 1.6798... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9541... Training loss: 1.8094... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9542... Training loss: 1.7266... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9543... Training loss: 1.7958... 0.0386 sec/batch
Epoch: 3/20... Training Step: 9544... Training loss: 1.8337... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9545... Training loss: 1.9142... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9546... Training loss: 1.7169... 0.0389 sec/batch
Epoch: 3/20... Training Step: 9547... Training loss: 1.9351... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9548... Training loss: 1.9124... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9549... Training loss: 1.8231... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9550... Training loss: 1.7346... 0.0386 sec/batch
Epoch: 3/20... Training Step: 9551... Training loss: 1.9221... 0.0390 sec/batch
Epoch: 3/20... Training Step: 9552... Training loss: 1.7291... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9553... Training loss: 1.7750... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9554... Training loss: 1.7675... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9555... Training loss: 1.8226... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9556... Training loss: 1.7268... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9557... Training loss: 1.7987... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9558... Training loss: 1.8676... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9559... Training loss: 1.7779... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9560... Training loss: 1.7865... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9561... Training loss: 1.7707... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9562... Training loss: 1.6658... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9563... Training loss: 1.6912... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9564... Training loss: 1.6396... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9565... Training loss: 1.8000... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9566... Training loss: 1.7364... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9567... Training loss: 1.5856... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9568... Training loss: 1.8110... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9569... Training loss: 1.8380... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9570... Training loss: 1.7313... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9571... Training loss: 1.7385... 0.0393 sec/batch
Epoch: 3/20... Training Step: 9572... Training loss: 1.5125... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9573... Training loss: 1.6113... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9574... Training loss: 1.6489... 0.0388 sec/batch
Epoch: 3/20... Training Step: 9575... Training loss: 1.6629... 0.0387 sec/batch
Epoch: 3/20... Training Step: 9576... Training loss: 1.7299... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9577... Training loss: 1.7008... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9578... Training loss: 1.6430... 0.0393 sec/batch
Epoch: 3/20... Training Step: 9579... Training loss: 1.6884... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9580... Training loss: 1.7292... 0.0391 sec/batch
Epoch: 3/20... Training Step: 9581... Training loss: 1.5618... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9582... Training loss: 1.6231... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9583... Training loss: 1.6746... 0.0466 sec/batch
Epoch: 3/20... Training Step: 9584... Training loss: 1.6197... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9585... Training loss: 1.6148... 0.0627 sec/batch
Epoch: 3/20... Training Step: 9586... Training loss: 1.6290... 0.0436 sec/batch
Epoch: 3/20... Training Step: 9587... Training loss: 1.6215... 0.0447 sec/batch
Epoch: 3/20... Training Step: 9588... Training loss: 1.8443... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9589... Training loss: 1.5595... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9590... Training loss: 1.6167... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9591... Training loss: 1.6728... 0.0387 sec/batch
Epoch: 3/20... Training Step: 9592... Training loss: 1.8447... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9593... Training loss: 1.6704... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9594... Training loss: 1.6203... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9595... Training loss: 1.6165... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9596... Training loss: 1.5723... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9597... Training loss: 1.6658... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9598... Training loss: 1.7492... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9599... Training loss: 1.8241... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9600... Training loss: 1.6089... 0.0406 sec/batch
Epoch: 3/20... Training Step: 9601... Training loss: 1.7630... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9602... Training loss: 1.6057... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9603... Training loss: 1.7256... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9604... Training loss: 1.8506... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9605... Training loss: 2.0162... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9606... Training loss: 1.8252... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9607... Training loss: 1.6942... 0.0424 sec/batch
Epoch: 3/20... Training Step: 9608... Training loss: 1.8364... 0.0433 sec/batch
Epoch: 3/20... Training Step: 9609... Training loss: 1.6767... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9610... Training loss: 1.8221... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9611... Training loss: 1.6337... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9612... Training loss: 1.7023... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9613... Training loss: 1.6735... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9614... Training loss: 1.7479... 0.0395 sec/batch
Epoch: 3/20... Training Step: 9615... Training loss: 1.8192... 0.0387 sec/batch
Epoch: 3/20... Training Step: 9616... Training loss: 1.7390... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9617... Training loss: 1.7892... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9618... Training loss: 1.7841... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9619... Training loss: 1.7497... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9620... Training loss: 1.6511... 0.0449 sec/batch
Epoch: 3/20... Training Step: 9621... Training loss: 1.7763... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9622... Training loss: 1.6796... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9623... Training loss: 1.7253... 0.0380 sec/batch
Epoch: 3/20... Training Step: 9624... Training loss: 1.6366... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9625... Training loss: 1.7356... 0.0441 sec/batch
Epoch: 3/20... Training Step: 9626... Training loss: 1.7222... 0.0388 sec/batch
Epoch: 3/20... Training Step: 9627... Training loss: 1.7139... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9628... Training loss: 1.8259... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9629... Training loss: 1.7786... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9630... Training loss: 1.6567... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9631... Training loss: 1.6572... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9632... Training loss: 1.7397... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9633... Training loss: 1.5887... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9634... Training loss: 1.6857... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9635... Training loss: 1.7417... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9636... Training loss: 1.6858... 0.0386 sec/batch
Epoch: 3/20... Training Step: 9637... Training loss: 1.6193... 0.0383 sec/batch
Epoch: 3/20... Training Step: 9638... Training loss: 1.7086... 0.0388 sec/batch
Epoch: 3/20... Training Step: 9639... Training loss: 1.7244... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9640... Training loss: 1.6899... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9641... Training loss: 1.5654... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9642... Training loss: 1.7542... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9643... Training loss: 1.6412... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9644... Training loss: 1.6793... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9645... Training loss: 1.7597... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9646... Training loss: 1.7908... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9647... Training loss: 1.7163... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9648... Training loss: 1.6481... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9649... Training loss: 1.7632... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9650... Training loss: 1.6928... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9651... Training loss: 1.7307... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9652... Training loss: 1.6815... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9653... Training loss: 1.5759... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9654... Training loss: 1.7120... 0.0388 sec/batch
Epoch: 3/20... Training Step: 9655... Training loss: 1.8073... 0.0387 sec/batch
Epoch: 3/20... Training Step: 9656... Training loss: 1.8732... 0.0445 sec/batch
Epoch: 3/20... Training Step: 9657... Training loss: 1.6057... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9658... Training loss: 1.6911... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9659... Training loss: 1.6648... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9660... Training loss: 1.6882... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9661... Training loss: 1.7057... 0.0392 sec/batch
Epoch: 3/20... Training Step: 9662... Training loss: 1.7438... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9663... Training loss: 1.6454... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9664... Training loss: 1.5619... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9665... Training loss: 1.6742... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9666... Training loss: 1.7036... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9667... Training loss: 1.6086... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9668... Training loss: 1.7210... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9669... Training loss: 1.7017... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9670... Training loss: 1.6018... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9671... Training loss: 1.7217... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9672... Training loss: 1.8903... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9673... Training loss: 1.7645... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9674... Training loss: 1.9049... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9675... Training loss: 1.8181... 0.0438 sec/batch
Epoch: 3/20... Training Step: 9676... Training loss: 1.6842... 0.0419 sec/batch
Epoch: 3/20... Training Step: 9677... Training loss: 1.6609... 0.0424 sec/batch
Epoch: 3/20... Training Step: 9678... Training loss: 1.6662... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9679... Training loss: 1.7101... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9680... Training loss: 1.8104... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9681... Training loss: 1.6801... 0.0465 sec/batch
Epoch: 3/20... Training Step: 9682... Training loss: 1.7160... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9683... Training loss: 1.7592... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9684... Training loss: 1.5645... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9685... Training loss: 1.6422... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9686... Training loss: 1.5220... 0.0474 sec/batch
Epoch: 3/20... Training Step: 9687... Training loss: 1.7306... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9688... Training loss: 1.7237... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9689... Training loss: 1.5545... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9690... Training loss: 1.7038... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9691... Training loss: 1.7613... 0.0438 sec/batch
Epoch: 3/20... Training Step: 9692... Training loss: 1.5947... 0.0474 sec/batch
Epoch: 3/20... Training Step: 9693... Training loss: 1.6500... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9694... Training loss: 1.7003... 0.0427 sec/batch
Epoch: 3/20... Training Step: 9695... Training loss: 1.6767... 0.0449 sec/batch
Epoch: 3/20... Training Step: 9696... Training loss: 1.5406... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9697... Training loss: 1.6277... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9698... Training loss: 1.5709... 0.0389 sec/batch
Epoch: 3/20... Training Step: 9699... Training loss: 1.7730... 0.0504 sec/batch
Epoch: 3/20... Training Step: 9700... Training loss: 1.6333... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9701... Training loss: 1.6952... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9702... Training loss: 1.7338... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9703... Training loss: 1.6206... 0.0788 sec/batch
Epoch: 3/20... Training Step: 9704... Training loss: 1.7305... 0.0598 sec/batch
Epoch: 3/20... Training Step: 9705... Training loss: 1.7211... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9706... Training loss: 1.7550... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9707... Training loss: 1.5662... 0.0510 sec/batch
Epoch: 3/20... Training Step: 9708... Training loss: 1.6733... 0.0453 sec/batch
Epoch: 3/20... Training Step: 9709... Training loss: 1.8215... 0.0470 sec/batch
Epoch: 3/20... Training Step: 9710... Training loss: 1.7946... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9711... Training loss: 1.5224... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9712... Training loss: 1.6167... 0.0446 sec/batch
Epoch: 3/20... Training Step: 9713... Training loss: 1.6354... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9714... Training loss: 1.8145... 0.0551 sec/batch
Epoch: 3/20... Training Step: 9715... Training loss: 1.8306... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9716... Training loss: 1.4756... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9717... Training loss: 1.7305... 0.0451 sec/batch
Epoch: 3/20... Training Step: 9718... Training loss: 1.6115... 0.0488 sec/batch
Epoch: 3/20... Training Step: 9719... Training loss: 1.6324... 0.0419 sec/batch
Epoch: 3/20... Training Step: 9720... Training loss: 1.7549... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9721... Training loss: 1.7494... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9722... Training loss: 1.6489... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9723... Training loss: 1.6863... 0.0440 sec/batch
Epoch: 3/20... Training Step: 9724... Training loss: 1.6208... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9725... Training loss: 1.6918... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9726... Training loss: 1.6964... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9727... Training loss: 1.7769... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9728... Training loss: 1.5941... 0.0407 sec/batch
Epoch: 3/20... Training Step: 9729... Training loss: 1.6779... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9730... Training loss: 1.6477... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9731... Training loss: 1.7206... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9732... Training loss: 1.8658... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9733... Training loss: 1.6380... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9734... Training loss: 1.5136... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9735... Training loss: 1.4266... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9736... Training loss: 1.7010... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9737... Training loss: 1.4618... 0.0436 sec/batch
Epoch: 3/20... Training Step: 9738... Training loss: 1.6418... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9739... Training loss: 1.6458... 0.0500 sec/batch
Epoch: 3/20... Training Step: 9740... Training loss: 1.6798... 0.0517 sec/batch
Epoch: 3/20... Training Step: 9741... Training loss: 1.6551... 0.0626 sec/batch
Epoch: 3/20... Training Step: 9742... Training loss: 1.6346... 0.0608 sec/batch
Epoch: 3/20... Training Step: 9743... Training loss: 1.6750... 0.0487 sec/batch
Epoch: 3/20... Training Step: 9744... Training loss: 1.5914... 0.0470 sec/batch
Epoch: 3/20... Training Step: 9745... Training loss: 1.5629... 0.0504 sec/batch
Epoch: 3/20... Training Step: 9746... Training loss: 1.6043... 0.0614 sec/batch
Epoch: 3/20... Training Step: 9747... Training loss: 1.5906... 0.0477 sec/batch
Epoch: 3/20... Training Step: 9748... Training loss: 1.6624... 0.0461 sec/batch
Epoch: 3/20... Training Step: 9749... Training loss: 1.5989... 0.0451 sec/batch
Epoch: 3/20... Training Step: 9750... Training loss: 1.4670... 0.0760 sec/batch
Epoch: 3/20... Training Step: 9751... Training loss: 1.5753... 0.0539 sec/batch
Epoch: 3/20... Training Step: 9752... Training loss: 1.7434... 0.0460 sec/batch
Epoch: 3/20... Training Step: 9753... Training loss: 1.5665... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9754... Training loss: 1.7246... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9755... Training loss: 1.6901... 0.0575 sec/batch
Epoch: 3/20... Training Step: 9756... Training loss: 1.8049... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9757... Training loss: 1.6306... 0.0459 sec/batch
Epoch: 3/20... Training Step: 9758... Training loss: 1.5618... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9759... Training loss: 1.6291... 0.0427 sec/batch
Epoch: 3/20... Training Step: 9760... Training loss: 1.7541... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9761... Training loss: 1.7197... 0.0716 sec/batch
Epoch: 3/20... Training Step: 9762... Training loss: 1.7516... 0.0525 sec/batch
Epoch: 3/20... Training Step: 9763... Training loss: 1.6704... 0.0626 sec/batch
Epoch: 3/20... Training Step: 9764... Training loss: 1.6829... 0.0489 sec/batch
Epoch: 3/20... Training Step: 9765... Training loss: 1.7187... 0.0612 sec/batch
Epoch: 3/20... Training Step: 9766... Training loss: 1.6740... 0.0558 sec/batch
Epoch: 3/20... Training Step: 9767... Training loss: 1.8077... 0.0478 sec/batch
Epoch: 3/20... Training Step: 9768... Training loss: 1.6921... 0.0524 sec/batch
Epoch: 3/20... Training Step: 9769... Training loss: 1.7797... 0.0448 sec/batch
Epoch: 3/20... Training Step: 9770... Training loss: 1.6528... 0.0644 sec/batch
Epoch: 3/20... Training Step: 9771... Training loss: 1.7575... 0.0570 sec/batch
Epoch: 3/20... Training Step: 9772... Training loss: 1.6794... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9773... Training loss: 1.7235... 0.0444 sec/batch
Epoch: 3/20... Training Step: 9774... Training loss: 1.7486... 0.0492 sec/batch
Epoch: 3/20... Training Step: 9775... Training loss: 1.6342... 0.0594 sec/batch
Epoch: 3/20... Training Step: 9776... Training loss: 1.7590... 0.0589 sec/batch
Epoch: 3/20... Training Step: 9777... Training loss: 1.7406... 0.0684 sec/batch
Epoch: 3/20... Training Step: 9778... Training loss: 1.6209... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9779... Training loss: 1.6362... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9780... Training loss: 1.6543... 0.0597 sec/batch
Epoch: 3/20... Training Step: 9781... Training loss: 1.6250... 0.0679 sec/batch
Epoch: 3/20... Training Step: 9782... Training loss: 1.7443... 0.0472 sec/batch
Epoch: 3/20... Training Step: 9783... Training loss: 1.4602... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9784... Training loss: 1.5690... 0.0502 sec/batch
Epoch: 3/20... Training Step: 9785... Training loss: 1.7102... 0.0552 sec/batch
Epoch: 3/20... Training Step: 9786... Training loss: 1.7755... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9787... Training loss: 1.6978... 0.0441 sec/batch
Epoch: 3/20... Training Step: 9788... Training loss: 1.6649... 0.0437 sec/batch
Epoch: 3/20... Training Step: 9789... Training loss: 1.5789... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9790... Training loss: 1.8111... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9791... Training loss: 1.6431... 0.0693 sec/batch
Epoch: 3/20... Training Step: 9792... Training loss: 1.8510... 0.0483 sec/batch
Epoch: 3/20... Training Step: 9793... Training loss: 1.8628... 0.0438 sec/batch
Epoch: 3/20... Training Step: 9794... Training loss: 1.8894... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9795... Training loss: 1.7584... 0.0484 sec/batch
Epoch: 3/20... Training Step: 9796... Training loss: 1.5183... 0.0646 sec/batch
Epoch: 3/20... Training Step: 9797... Training loss: 1.9194... 0.0550 sec/batch
Epoch: 3/20... Training Step: 9798... Training loss: 1.8257... 0.0549 sec/batch
Epoch: 3/20... Training Step: 9799... Training loss: 1.6430... 0.0585 sec/batch
Epoch: 3/20... Training Step: 9800... Training loss: 1.7718... 0.0736 sec/batch
Epoch: 3/20... Training Step: 9801... Training loss: 1.7128... 0.0481 sec/batch
Epoch: 3/20... Training Step: 9802... Training loss: 1.7629... 0.0430 sec/batch
Epoch: 3/20... Training Step: 9803... Training loss: 1.7507... 0.0432 sec/batch
Epoch: 3/20... Training Step: 9804... Training loss: 1.8292... 0.0506 sec/batch
Epoch: 3/20... Training Step: 9805... Training loss: 1.6448... 0.0524 sec/batch
Epoch: 3/20... Training Step: 9806... Training loss: 1.6023... 0.0534 sec/batch
Epoch: 3/20... Training Step: 9807... Training loss: 1.7967... 0.0566 sec/batch
Epoch: 3/20... Training Step: 9808... Training loss: 1.6704... 0.0640 sec/batch
Epoch: 3/20... Training Step: 9809... Training loss: 1.6790... 0.0610 sec/batch
Epoch: 3/20... Training Step: 9810... Training loss: 1.6216... 0.0622 sec/batch
Epoch: 3/20... Training Step: 9811... Training loss: 1.6684... 0.0582 sec/batch
Epoch: 3/20... Training Step: 9812... Training loss: 1.5326... 0.0582 sec/batch
Epoch: 3/20... Training Step: 9813... Training loss: 1.6141... 0.0738 sec/batch
Epoch: 3/20... Training Step: 9814... Training loss: 1.6654... 0.0591 sec/batch
Epoch: 3/20... Training Step: 9815... Training loss: 1.6727... 0.0655 sec/batch
Epoch: 3/20... Training Step: 9816... Training loss: 1.8257... 0.0598 sec/batch
Epoch: 3/20... Training Step: 9817... Training loss: 1.6923... 0.0604 sec/batch
Epoch: 3/20... Training Step: 9818... Training loss: 1.6675... 0.0615 sec/batch
Epoch: 3/20... Training Step: 9819... Training loss: 1.6109... 0.0497 sec/batch
Epoch: 3/20... Training Step: 9820... Training loss: 1.7865... 0.0440 sec/batch
Epoch: 3/20... Training Step: 9821... Training loss: 1.8714... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9822... Training loss: 1.5893... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9823... Training loss: 1.8479... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9824... Training loss: 1.8500... 0.0499 sec/batch
Epoch: 3/20... Training Step: 9825... Training loss: 1.5450... 0.0434 sec/batch
Epoch: 3/20... Training Step: 9826... Training loss: 1.7253... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9827... Training loss: 1.8109... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9828... Training loss: 1.8382... 0.0423 sec/batch
Epoch: 3/20... Training Step: 9829... Training loss: 1.6200... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9830... Training loss: 1.7382... 0.0432 sec/batch
Epoch: 3/20... Training Step: 9831... Training loss: 1.7808... 0.0550 sec/batch
Epoch: 3/20... Training Step: 9832... Training loss: 1.7739... 0.0603 sec/batch
Epoch: 3/20... Training Step: 9833... Training loss: 1.6765... 0.0549 sec/batch
Epoch: 3/20... Training Step: 9834... Training loss: 1.7028... 0.0607 sec/batch
Epoch: 3/20... Training Step: 9835... Training loss: 1.5700... 0.0523 sec/batch
Epoch: 3/20... Training Step: 9836... Training loss: 1.6420... 0.0409 sec/batch
Epoch: 3/20... Training Step: 9837... Training loss: 1.7769... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9838... Training loss: 1.7607... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9839... Training loss: 1.6828... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9840... Training loss: 1.5488... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9841... Training loss: 1.6888... 0.0531 sec/batch
Epoch: 3/20... Training Step: 9842... Training loss: 1.7365... 0.0514 sec/batch
Epoch: 3/20... Training Step: 9843... Training loss: 1.6927... 0.0585 sec/batch
Epoch: 3/20... Training Step: 9844... Training loss: 1.4811... 0.0571 sec/batch
Epoch: 3/20... Training Step: 9845... Training loss: 1.5830... 0.0543 sec/batch
Epoch: 3/20... Training Step: 9846... Training loss: 1.7842... 0.0579 sec/batch
Epoch: 3/20... Training Step: 9847... Training loss: 1.7235... 0.0446 sec/batch
Epoch: 3/20... Training Step: 9848... Training loss: 1.7401... 0.0442 sec/batch
Epoch: 3/20... Training Step: 9849... Training loss: 1.7268... 0.0402 sec/batch
Epoch: 3/20... Training Step: 9850... Training loss: 1.6131... 0.0523 sec/batch
Epoch: 3/20... Training Step: 9851... Training loss: 1.6396... 0.0588 sec/batch
Epoch: 3/20... Training Step: 9852... Training loss: 1.6856... 0.0634 sec/batch
Epoch: 3/20... Training Step: 9853... Training loss: 1.6399... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9854... Training loss: 1.7163... 0.0513 sec/batch
Epoch: 3/20... Training Step: 9855... Training loss: 1.6127... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9856... Training loss: 1.6824... 0.0568 sec/batch
Epoch: 3/20... Training Step: 9857... Training loss: 1.7329... 0.0451 sec/batch
Epoch: 3/20... Training Step: 9858... Training loss: 1.7509... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9859... Training loss: 1.6449... 0.0525 sec/batch
Epoch: 3/20... Training Step: 9860... Training loss: 1.6711... 0.0591 sec/batch
Epoch: 3/20... Training Step: 9861... Training loss: 1.7457... 0.0570 sec/batch
Epoch: 3/20... Training Step: 9862... Training loss: 1.5192... 0.0558 sec/batch
Epoch: 3/20... Training Step: 9863... Training loss: 1.6972... 0.0583 sec/batch
Epoch: 3/20... Training Step: 9864... Training loss: 1.6515... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9865... Training loss: 1.6132... 0.0459 sec/batch
Epoch: 3/20... Training Step: 9866... Training loss: 1.5466... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9867... Training loss: 1.6449... 0.0406 sec/batch
Epoch: 3/20... Training Step: 9868... Training loss: 1.7639... 0.0414 sec/batch
Epoch: 3/20... Training Step: 9869... Training loss: 1.7298... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9870... Training loss: 1.7568... 0.0396 sec/batch
Epoch: 3/20... Training Step: 9871... Training loss: 1.6927... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9872... Training loss: 1.7101... 0.0482 sec/batch
Epoch: 3/20... Training Step: 9873... Training loss: 1.6802... 0.0500 sec/batch
Epoch: 3/20... Training Step: 9874... Training loss: 1.9291... 0.0557 sec/batch
Epoch: 3/20... Training Step: 9875... Training loss: 1.6768... 0.0635 sec/batch
Epoch: 3/20... Training Step: 9876... Training loss: 1.7330... 0.0572 sec/batch
Epoch: 3/20... Training Step: 9877... Training loss: 1.8330... 0.0484 sec/batch
Epoch: 3/20... Training Step: 9878... Training loss: 2.0640... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9879... Training loss: 1.8333... 0.0394 sec/batch
Epoch: 3/20... Training Step: 9880... Training loss: 1.9384... 0.0514 sec/batch
Epoch: 3/20... Training Step: 9881... Training loss: 1.6856... 0.0562 sec/batch
Epoch: 3/20... Training Step: 9882... Training loss: 1.6429... 0.0552 sec/batch
Epoch: 3/20... Training Step: 9883... Training loss: 1.7133... 0.0518 sec/batch
Epoch: 3/20... Training Step: 9884... Training loss: 1.6315... 0.0521 sec/batch
Epoch: 3/20... Training Step: 9885... Training loss: 1.7829... 0.0469 sec/batch
Epoch: 3/20... Training Step: 9886... Training loss: 1.5652... 0.0393 sec/batch
Epoch: 3/20... Training Step: 9887... Training loss: 1.6663... 0.0552 sec/batch
Epoch: 3/20... Training Step: 9888... Training loss: 1.6881... 0.0611 sec/batch
Epoch: 3/20... Training Step: 9889... Training loss: 1.6485... 0.0602 sec/batch
Epoch: 3/20... Training Step: 9890... Training loss: 1.6944... 0.0582 sec/batch
Epoch: 3/20... Training Step: 9891... Training loss: 1.6998... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9892... Training loss: 1.7515... 0.0410 sec/batch
Epoch: 3/20... Training Step: 9893... Training loss: 1.6992... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9894... Training loss: 1.5676... 0.0435 sec/batch
Epoch: 3/20... Training Step: 9895... Training loss: 1.8137... 0.0403 sec/batch
Epoch: 3/20... Training Step: 9896... Training loss: 1.8733... 0.0431 sec/batch
Epoch: 3/20... Training Step: 9897... Training loss: 1.6957... 0.0400 sec/batch
Epoch: 3/20... Training Step: 9898... Training loss: 1.6491... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9899... Training loss: 1.7030... 0.0397 sec/batch
Epoch: 3/20... Training Step: 9900... Training loss: 1.7235... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9901... Training loss: 1.6437... 0.0385 sec/batch
Epoch: 3/20... Training Step: 9902... Training loss: 1.7685... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9903... Training loss: 1.5715... 0.0436 sec/batch
Epoch: 3/20... Training Step: 9904... Training loss: 1.5157... 0.0426 sec/batch
Epoch: 3/20... Training Step: 9905... Training loss: 1.6749... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9906... Training loss: 1.5307... 0.0394 sec/batch
Epoch: 3/20... Training Step: 9907... Training loss: 1.6826... 0.0398 sec/batch
Epoch: 3/20... Training Step: 9908... Training loss: 1.7594... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9909... Training loss: 1.8825... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9910... Training loss: 1.6931... 0.0528 sec/batch
Epoch: 3/20... Training Step: 9911... Training loss: 1.6529... 0.0564 sec/batch
Epoch: 3/20... Training Step: 9912... Training loss: 1.6738... 0.0595 sec/batch
Epoch: 3/20... Training Step: 9913... Training loss: 1.7448... 0.0549 sec/batch
Epoch: 3/20... Training Step: 9914... Training loss: 1.8007... 0.0587 sec/batch
Epoch: 3/20... Training Step: 9915... Training loss: 1.8181... 0.0513 sec/batch
Epoch: 3/20... Training Step: 9916... Training loss: 1.5681... 0.0433 sec/batch
Epoch: 3/20... Training Step: 9917... Training loss: 1.7219... 0.0418 sec/batch
Epoch: 3/20... Training Step: 9918... Training loss: 1.6112... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9919... Training loss: 1.7755... 0.0456 sec/batch
Epoch: 3/20... Training Step: 9920... Training loss: 1.6175... 0.0412 sec/batch
Epoch: 3/20... Training Step: 9921... Training loss: 1.7697... 0.0622 sec/batch
Epoch: 3/20... Training Step: 9922... Training loss: 1.6095... 0.0563 sec/batch
Epoch: 3/20... Training Step: 9923... Training loss: 1.6583... 0.0563 sec/batch
Epoch: 3/20... Training Step: 9924... Training loss: 1.8183... 0.0527 sec/batch
Epoch: 3/20... Training Step: 9925... Training loss: 1.9249... 0.0569 sec/batch
Epoch: 3/20... Training Step: 9926... Training loss: 1.6460... 0.0579 sec/batch
Epoch: 3/20... Training Step: 9927... Training loss: 1.6139... 0.0408 sec/batch
Epoch: 3/20... Training Step: 9928... Training loss: 1.7240... 0.0428 sec/batch
Epoch: 3/20... Training Step: 9929... Training loss: 1.5964... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9930... Training loss: 1.6526... 0.0448 sec/batch
Epoch: 3/20... Training Step: 9931... Training loss: 1.5899... 0.0518 sec/batch
Epoch: 3/20... Training Step: 9932... Training loss: 1.7700... 0.0577 sec/batch
Epoch: 3/20... Training Step: 9933... Training loss: 1.9519... 0.0550 sec/batch
Epoch: 3/20... Training Step: 9934... Training loss: 1.8014... 0.0560 sec/batch
Epoch: 3/20... Training Step: 9935... Training loss: 1.6982... 0.0620 sec/batch
Epoch: 3/20... Training Step: 9936... Training loss: 1.8170... 0.0696 sec/batch
Epoch: 3/20... Training Step: 9937... Training loss: 1.6885... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9938... Training loss: 1.6823... 0.0568 sec/batch
Epoch: 3/20... Training Step: 9939... Training loss: 1.6552... 0.0425 sec/batch
Epoch: 3/20... Training Step: 9940... Training loss: 1.6947... 0.0452 sec/batch
Epoch: 3/20... Training Step: 9941... Training loss: 1.6040... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9942... Training loss: 1.5365... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9943... Training loss: 1.5630... 0.0567 sec/batch
Epoch: 3/20... Training Step: 9944... Training loss: 1.6390... 0.0586 sec/batch
Epoch: 3/20... Training Step: 9945... Training loss: 1.7168... 0.0573 sec/batch
Epoch: 3/20... Training Step: 9946... Training loss: 1.5123... 0.0523 sec/batch
Epoch: 3/20... Training Step: 9947... Training loss: 1.6575... 0.0573 sec/batch
Epoch: 3/20... Training Step: 9948... Training loss: 1.6300... 0.0539 sec/batch
Epoch: 3/20... Training Step: 9949... Training loss: 1.5518... 0.0542 sec/batch
Epoch: 3/20... Training Step: 9950... Training loss: 1.7346... 0.0624 sec/batch
Epoch: 3/20... Training Step: 9951... Training loss: 1.5684... 0.0697 sec/batch
Epoch: 3/20... Training Step: 9952... Training loss: 1.6254... 0.0769 sec/batch
Epoch: 3/20... Training Step: 9953... Training loss: 1.6852... 0.0571 sec/batch
Epoch: 3/20... Training Step: 9954... Training loss: 1.5784... 0.0596 sec/batch
Epoch: 3/20... Training Step: 9955... Training loss: 1.7825... 0.0650 sec/batch
Epoch: 3/20... Training Step: 9956... Training loss: 1.6205... 0.0553 sec/batch
Epoch: 3/20... Training Step: 9957... Training loss: 1.5712... 0.0513 sec/batch
Epoch: 3/20... Training Step: 9958... Training loss: 1.5228... 0.0528 sec/batch
Epoch: 3/20... Training Step: 9959... Training loss: 1.5667... 0.0593 sec/batch
Epoch: 3/20... Training Step: 9960... Training loss: 1.6192... 0.0440 sec/batch
Epoch: 3/20... Training Step: 9961... Training loss: 1.7651... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9962... Training loss: 1.5847... 0.0411 sec/batch
Epoch: 3/20... Training Step: 9963... Training loss: 1.7924... 0.0439 sec/batch
Epoch: 3/20... Training Step: 9964... Training loss: 1.6461... 0.0422 sec/batch
Epoch: 3/20... Training Step: 9965... Training loss: 1.6672... 0.0442 sec/batch
Epoch: 3/20... Training Step: 9966... Training loss: 1.5589... 0.0450 sec/batch
Epoch: 3/20... Training Step: 9967... Training loss: 1.6598... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9968... Training loss: 1.5236... 0.0406 sec/batch
Epoch: 3/20... Training Step: 9969... Training loss: 1.7438... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9970... Training loss: 1.5224... 0.0413 sec/batch
Epoch: 3/20... Training Step: 9971... Training loss: 1.6811... 0.0464 sec/batch
Epoch: 3/20... Training Step: 9972... Training loss: 1.7724... 0.0433 sec/batch
Epoch: 3/20... Training Step: 9973... Training loss: 1.6730... 0.0420 sec/batch
Epoch: 3/20... Training Step: 9974... Training loss: 1.7263... 0.0473 sec/batch
Epoch: 3/20... Training Step: 9975... Training loss: 1.6750... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9976... Training loss: 1.7167... 0.0419 sec/batch
Epoch: 3/20... Training Step: 9977... Training loss: 1.5414... 0.0419 sec/batch
Epoch: 3/20... Training Step: 9978... Training loss: 1.6939... 0.0454 sec/batch
Epoch: 3/20... Training Step: 9979... Training loss: 1.6865... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9980... Training loss: 1.7763... 0.0415 sec/batch
Epoch: 3/20... Training Step: 9981... Training loss: 1.6397... 0.0476 sec/batch
Epoch: 3/20... Training Step: 9982... Training loss: 1.7395... 0.0417 sec/batch
Epoch: 3/20... Training Step: 9983... Training loss: 1.7068... 0.0416 sec/batch
Epoch: 3/20... Training Step: 9984... Training loss: 1.7421... 0.0443 sec/batch
Epoch: 3/20... Training Step: 9985... Training loss: 1.7704... 0.0493 sec/batch
Epoch: 3/20... Training Step: 9986... Training loss: 1.6580... 0.0429 sec/batch
Epoch: 3/20... Training Step: 9987... Training loss: 1.7111... 0.0449 sec/batch
Epoch: 3/20... Training Step: 9988... Training loss: 1.5693... 0.0401 sec/batch
Epoch: 3/20... Training Step: 9989... Training loss: 1.7098... 0.0468 sec/batch
Epoch: 3/20... Training Step: 9990... Training loss: 1.8496... 0.0482 sec/batch
Epoch: 3/20... Training Step: 9991... Training loss: 1.6815... 0.0405 sec/batch
Epoch: 3/20... Training Step: 9992... Training loss: 1.7366... 0.0399 sec/batch
Epoch: 3/20... Training Step: 9993... Training loss: 1.5752... 0.0453 sec/batch
Epoch: 3/20... Training Step: 9994... Training loss: 1.5709... 0.0421 sec/batch
Epoch: 3/20... Training Step: 9995... Training loss: 1.6198... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9996... Training loss: 1.6766... 0.0404 sec/batch
Epoch: 3/20... Training Step: 9997... Training loss: 1.6476... 0.0435 sec/batch
Epoch: 3/20... Training Step: 9998... Training loss: 1.6364... 0.0512 sec/batch
Epoch: 3/20... Training Step: 9999... Training loss: 1.6279... 0.0578 sec/batch
Epoch: 3/20... Training Step: 10000... Training loss: 1.6391... 0.0608 sec/batch
Epoch: 3/20... Training Step: 10001... Training loss: 1.7142... 0.0577 sec/batch
Epoch: 3/20... Training Step: 10002... Training loss: 1.6497... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10003... Training loss: 1.6641... 0.0391 sec/batch
Epoch: 3/20... Training Step: 10004... Training loss: 1.7182... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10005... Training loss: 1.6589... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10006... Training loss: 1.7539... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10007... Training loss: 1.6501... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10008... Training loss: 1.8470... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10009... Training loss: 1.7987... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10010... Training loss: 1.5924... 0.0387 sec/batch
Epoch: 3/20... Training Step: 10011... Training loss: 1.7433... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10012... Training loss: 1.8237... 0.0531 sec/batch
Epoch: 3/20... Training Step: 10013... Training loss: 1.6306... 0.0588 sec/batch
Epoch: 3/20... Training Step: 10014... Training loss: 1.8950... 0.0551 sec/batch
Epoch: 3/20... Training Step: 10015... Training loss: 1.5992... 0.0575 sec/batch
Epoch: 3/20... Training Step: 10016... Training loss: 1.7514... 0.0571 sec/batch
Epoch: 3/20... Training Step: 10017... Training loss: 1.6426... 0.0560 sec/batch
Epoch: 3/20... Training Step: 10018... Training loss: 1.7180... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10019... Training loss: 1.7617... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10020... Training loss: 1.6482... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10021... Training loss: 1.6983... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10022... Training loss: 1.6738... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10023... Training loss: 1.6973... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10024... Training loss: 1.7030... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10025... Training loss: 1.6269... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10026... Training loss: 1.7155... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10027... Training loss: 1.6905... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10028... Training loss: 1.8987... 0.0509 sec/batch
Epoch: 3/20... Training Step: 10029... Training loss: 1.6773... 0.0576 sec/batch
Epoch: 3/20... Training Step: 10030... Training loss: 1.6247... 0.0562 sec/batch
Epoch: 3/20... Training Step: 10031... Training loss: 1.7742... 0.0575 sec/batch
Epoch: 3/20... Training Step: 10032... Training loss: 1.6512... 0.0577 sec/batch
Epoch: 3/20... Training Step: 10033... Training loss: 1.6854... 0.0538 sec/batch
Epoch: 3/20... Training Step: 10034... Training loss: 1.6959... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10035... Training loss: 1.6963... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10036... Training loss: 1.7617... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10037... Training loss: 1.7306... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10038... Training loss: 1.7254... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10039... Training loss: 1.5070... 0.0384 sec/batch
Epoch: 3/20... Training Step: 10040... Training loss: 1.6992... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10041... Training loss: 1.5954... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10042... Training loss: 1.6445... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10043... Training loss: 1.8185... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10044... Training loss: 1.7384... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10045... Training loss: 1.6853... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10046... Training loss: 1.7432... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10047... Training loss: 1.7144... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10048... Training loss: 1.7373... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10049... Training loss: 1.7074... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10050... Training loss: 1.6254... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10051... Training loss: 1.6109... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10052... Training loss: 1.7165... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10053... Training loss: 1.6797... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10054... Training loss: 1.6444... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10055... Training loss: 1.5408... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10056... Training loss: 1.6890... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10057... Training loss: 1.6849... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10058... Training loss: 1.6369... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10059... Training loss: 1.6131... 0.0391 sec/batch
Epoch: 3/20... Training Step: 10060... Training loss: 1.5176... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10061... Training loss: 1.7751... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10062... Training loss: 1.8304... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10063... Training loss: 1.6085... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10064... Training loss: 1.6728... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10065... Training loss: 1.7225... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10066... Training loss: 1.6880... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10067... Training loss: 1.7551... 0.0518 sec/batch
Epoch: 3/20... Training Step: 10068... Training loss: 1.5983... 0.0563 sec/batch
Epoch: 3/20... Training Step: 10069... Training loss: 1.6056... 0.0580 sec/batch
Epoch: 3/20... Training Step: 10070... Training loss: 1.7806... 0.0576 sec/batch
Epoch: 3/20... Training Step: 10071... Training loss: 1.6111... 0.0542 sec/batch
Epoch: 3/20... Training Step: 10072... Training loss: 1.6014... 0.0455 sec/batch
Epoch: 3/20... Training Step: 10073... Training loss: 1.7353... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10074... Training loss: 1.6752... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10075... Training loss: 1.6364... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10076... Training loss: 1.7143... 0.0568 sec/batch
Epoch: 3/20... Training Step: 10077... Training loss: 1.6589... 0.0566 sec/batch
Epoch: 3/20... Training Step: 10078... Training loss: 1.6889... 0.0536 sec/batch
Epoch: 3/20... Training Step: 10079... Training loss: 1.6976... 0.0544 sec/batch
Epoch: 3/20... Training Step: 10080... Training loss: 1.7245... 0.0599 sec/batch
Epoch: 3/20... Training Step: 10081... Training loss: 1.6906... 0.0598 sec/batch
Epoch: 3/20... Training Step: 10082... Training loss: 1.7432... 0.0520 sec/batch
Epoch: 3/20... Training Step: 10083... Training loss: 1.6602... 0.0685 sec/batch
Epoch: 3/20... Training Step: 10084... Training loss: 1.6753... 0.0674 sec/batch
Epoch: 3/20... Training Step: 10085... Training loss: 1.7324... 0.0547 sec/batch
Epoch: 3/20... Training Step: 10086... Training loss: 1.7294... 0.0499 sec/batch
Epoch: 3/20... Training Step: 10087... Training loss: 1.5505... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10088... Training loss: 1.9482... 0.0541 sec/batch
Epoch: 3/20... Training Step: 10089... Training loss: 1.6822... 0.0645 sec/batch
Epoch: 3/20... Training Step: 10090... Training loss: 1.6718... 0.0573 sec/batch
Epoch: 3/20... Training Step: 10091... Training loss: 1.8447... 0.0535 sec/batch
Epoch: 3/20... Training Step: 10092... Training loss: 1.6334... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10093... Training loss: 1.5785... 0.0545 sec/batch
Epoch: 3/20... Training Step: 10094... Training loss: 1.7662... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10095... Training loss: 1.6536... 0.0561 sec/batch
Epoch: 3/20... Training Step: 10096... Training loss: 1.5137... 0.0541 sec/batch
Epoch: 3/20... Training Step: 10097... Training loss: 1.7502... 0.0530 sec/batch
Epoch: 3/20... Training Step: 10098... Training loss: 1.6867... 0.0528 sec/batch
Epoch: 3/20... Training Step: 10099... Training loss: 1.8109... 0.0598 sec/batch
Epoch: 3/20... Training Step: 10100... Training loss: 1.7602... 0.0521 sec/batch
Epoch: 3/20... Training Step: 10101... Training loss: 1.6690... 0.0529 sec/batch
Epoch: 3/20... Training Step: 10102... Training loss: 1.6425... 0.0557 sec/batch
Epoch: 3/20... Training Step: 10103... Training loss: 1.5689... 0.0647 sec/batch
Epoch: 3/20... Training Step: 10104... Training loss: 1.7106... 0.0606 sec/batch
Epoch: 3/20... Training Step: 10105... Training loss: 1.7915... 0.0515 sec/batch
Epoch: 3/20... Training Step: 10106... Training loss: 1.6585... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10107... Training loss: 1.6673... 0.0555 sec/batch
Epoch: 3/20... Training Step: 10108... Training loss: 1.5365... 0.0581 sec/batch
Epoch: 3/20... Training Step: 10109... Training loss: 1.6729... 0.0570 sec/batch
Epoch: 3/20... Training Step: 10110... Training loss: 1.6120... 0.0523 sec/batch
Epoch: 3/20... Training Step: 10111... Training loss: 1.5888... 0.0581 sec/batch
Epoch: 3/20... Training Step: 10112... Training loss: 1.6867... 0.0580 sec/batch
Epoch: 3/20... Training Step: 10113... Training loss: 1.7786... 0.0587 sec/batch
Epoch: 3/20... Training Step: 10114... Training loss: 1.6559... 0.0659 sec/batch
Epoch: 3/20... Training Step: 10115... Training loss: 1.7879... 0.0593 sec/batch
Epoch: 3/20... Training Step: 10116... Training loss: 1.7711... 0.0579 sec/batch
Epoch: 3/20... Training Step: 10117... Training loss: 1.6860... 0.0526 sec/batch
Epoch: 3/20... Training Step: 10118... Training loss: 1.7438... 0.0627 sec/batch
Epoch: 3/20... Training Step: 10119... Training loss: 1.7228... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10120... Training loss: 1.7180... 0.0538 sec/batch
Epoch: 3/20... Training Step: 10121... Training loss: 1.6836... 0.0530 sec/batch
Epoch: 3/20... Training Step: 10122... Training loss: 1.8811... 0.0440 sec/batch
Epoch: 3/20... Training Step: 10123... Training loss: 1.5885... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10124... Training loss: 1.7494... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10125... Training loss: 1.7701... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10126... Training loss: 1.7422... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10127... Training loss: 1.7806... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10128... Training loss: 1.6124... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10129... Training loss: 1.6826... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10130... Training loss: 1.6857... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10131... Training loss: 1.7460... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10132... Training loss: 1.6293... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10133... Training loss: 1.6281... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10134... Training loss: 1.5531... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10135... Training loss: 1.7113... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10136... Training loss: 1.7941... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10137... Training loss: 1.6998... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10138... Training loss: 1.7518... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10139... Training loss: 1.6539... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10140... Training loss: 1.6691... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10141... Training loss: 1.7710... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10142... Training loss: 1.8121... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10143... Training loss: 1.9024... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10144... Training loss: 1.5385... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10145... Training loss: 1.7186... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10146... Training loss: 1.6465... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10147... Training loss: 1.8249... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10148... Training loss: 1.6759... 0.0608 sec/batch
Epoch: 3/20... Training Step: 10149... Training loss: 1.6776... 0.0559 sec/batch
Epoch: 3/20... Training Step: 10150... Training loss: 1.8357... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10151... Training loss: 1.7600... 0.0618 sec/batch
Epoch: 3/20... Training Step: 10152... Training loss: 1.5259... 0.0544 sec/batch
Epoch: 3/20... Training Step: 10153... Training loss: 1.5980... 0.0585 sec/batch
Epoch: 3/20... Training Step: 10154... Training loss: 1.7652... 0.0607 sec/batch
Epoch: 3/20... Training Step: 10155... Training loss: 1.7417... 0.0462 sec/batch
Epoch: 3/20... Training Step: 10156... Training loss: 1.9520... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10157... Training loss: 1.7463... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10158... Training loss: 1.5911... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10159... Training loss: 1.6838... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10160... Training loss: 1.7351... 0.0553 sec/batch
Epoch: 3/20... Training Step: 10161... Training loss: 1.6403... 0.0577 sec/batch
Epoch: 3/20... Training Step: 10162... Training loss: 1.7141... 0.0606 sec/batch
Epoch: 3/20... Training Step: 10163... Training loss: 1.6111... 0.0560 sec/batch
Epoch: 3/20... Training Step: 10164... Training loss: 1.5551... 0.0621 sec/batch
Epoch: 3/20... Training Step: 10165... Training loss: 1.8546... 0.0603 sec/batch
Epoch: 3/20... Training Step: 10166... Training loss: 1.5377... 0.0574 sec/batch
Epoch: 3/20... Training Step: 10167... Training loss: 1.6494... 0.0570 sec/batch
Epoch: 3/20... Training Step: 10168... Training loss: 1.6863... 0.0562 sec/batch
Epoch: 3/20... Training Step: 10169... Training loss: 1.7447... 0.0615 sec/batch
Epoch: 3/20... Training Step: 10170... Training loss: 1.6663... 0.0489 sec/batch
Epoch: 3/20... Training Step: 10171... Training loss: 1.7238... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10172... Training loss: 1.6784... 0.0516 sec/batch
Epoch: 3/20... Training Step: 10173... Training loss: 1.6616... 0.0603 sec/batch
Epoch: 3/20... Training Step: 10174... Training loss: 1.6688... 0.0602 sec/batch
Epoch: 3/20... Training Step: 10175... Training loss: 1.5708... 0.0596 sec/batch
Epoch: 3/20... Training Step: 10176... Training loss: 1.5255... 0.0686 sec/batch
Epoch: 3/20... Training Step: 10177... Training loss: 1.6605... 0.0545 sec/batch
Epoch: 3/20... Training Step: 10178... Training loss: 1.5370... 0.0577 sec/batch
Epoch: 3/20... Training Step: 10179... Training loss: 1.8106... 0.0622 sec/batch
Epoch: 3/20... Training Step: 10180... Training loss: 1.9634... 0.0643 sec/batch
Epoch: 3/20... Training Step: 10181... Training loss: 1.8024... 0.0565 sec/batch
Epoch: 3/20... Training Step: 10182... Training loss: 1.7560... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10183... Training loss: 1.7804... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10184... Training loss: 1.7571... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10185... Training loss: 2.0072... 0.0465 sec/batch
Epoch: 3/20... Training Step: 10186... Training loss: 1.8045... 0.0598 sec/batch
Epoch: 3/20... Training Step: 10187... Training loss: 1.6322... 0.0588 sec/batch
Epoch: 3/20... Training Step: 10188... Training loss: 1.7599... 0.0609 sec/batch
Epoch: 3/20... Training Step: 10189... Training loss: 1.5893... 0.0676 sec/batch
Epoch: 3/20... Training Step: 10190... Training loss: 1.7642... 0.0481 sec/batch
Epoch: 3/20... Training Step: 10191... Training loss: 1.6541... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10192... Training loss: 1.7587... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10193... Training loss: 1.5584... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10194... Training loss: 1.6806... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10195... Training loss: 1.7232... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10196... Training loss: 1.6153... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10197... Training loss: 1.7157... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10198... Training loss: 1.6264... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10199... Training loss: 1.5495... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10200... Training loss: 1.9108... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10201... Training loss: 1.9010... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10202... Training loss: 1.5791... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10203... Training loss: 1.5401... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10204... Training loss: 1.6688... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10205... Training loss: 1.6187... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10206... Training loss: 1.6438... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10207... Training loss: 1.6196... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10208... Training loss: 1.7340... 0.0456 sec/batch
Epoch: 3/20... Training Step: 10209... Training loss: 1.7404... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10210... Training loss: 1.5869... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10211... Training loss: 1.6266... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10212... Training loss: 1.7687... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10213... Training loss: 1.5518... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10214... Training loss: 1.5385... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10215... Training loss: 1.5322... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10216... Training loss: 1.6428... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10217... Training loss: 1.6537... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10218... Training loss: 1.7706... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10219... Training loss: 1.7617... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10220... Training loss: 1.6634... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10221... Training loss: 1.6398... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10222... Training loss: 1.6662... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10223... Training loss: 1.6896... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10224... Training loss: 1.5544... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10225... Training loss: 1.6183... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10226... Training loss: 1.5505... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10227... Training loss: 1.6033... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10228... Training loss: 1.6052... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10229... Training loss: 1.6937... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10230... Training loss: 1.6932... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10231... Training loss: 1.5929... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10232... Training loss: 1.6510... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10233... Training loss: 1.5794... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10234... Training loss: 1.6453... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10235... Training loss: 1.6975... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10236... Training loss: 1.5651... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10237... Training loss: 1.6526... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10238... Training loss: 1.5731... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10239... Training loss: 1.7451... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10240... Training loss: 1.5800... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10241... Training loss: 1.8161... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10242... Training loss: 1.7712... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10243... Training loss: 1.6534... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10244... Training loss: 1.6418... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10245... Training loss: 1.6573... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10246... Training loss: 1.7529... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10247... Training loss: 1.6567... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10248... Training loss: 1.7457... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10249... Training loss: 1.6513... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10250... Training loss: 1.7384... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10251... Training loss: 1.5909... 0.0387 sec/batch
Epoch: 3/20... Training Step: 10252... Training loss: 1.6262... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10253... Training loss: 1.6748... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10254... Training loss: 1.6978... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10255... Training loss: 1.6201... 0.0388 sec/batch
Epoch: 3/20... Training Step: 10256... Training loss: 1.6619... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10257... Training loss: 1.5165... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10258... Training loss: 1.7040... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10259... Training loss: 1.6474... 0.0460 sec/batch
Epoch: 3/20... Training Step: 10260... Training loss: 1.6152... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10261... Training loss: 1.6192... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10262... Training loss: 1.6434... 0.0377 sec/batch
Epoch: 3/20... Training Step: 10263... Training loss: 1.5562... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10264... Training loss: 1.7669... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10265... Training loss: 1.8495... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10266... Training loss: 1.7575... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10267... Training loss: 1.7042... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10268... Training loss: 1.7660... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10269... Training loss: 1.7746... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10270... Training loss: 1.6244... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10271... Training loss: 1.5267... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10272... Training loss: 1.7695... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10273... Training loss: 1.5704... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10274... Training loss: 1.5566... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10275... Training loss: 1.6722... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10276... Training loss: 1.6795... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10277... Training loss: 1.8676... 0.0388 sec/batch
Epoch: 3/20... Training Step: 10278... Training loss: 1.8648... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10279... Training loss: 1.7655... 0.0448 sec/batch
Epoch: 3/20... Training Step: 10280... Training loss: 1.8090... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10281... Training loss: 1.7089... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10282... Training loss: 1.6340... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10283... Training loss: 1.7209... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10284... Training loss: 1.6131... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10285... Training loss: 1.6420... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10286... Training loss: 1.4987... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10287... Training loss: 1.7175... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10288... Training loss: 1.6100... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10289... Training loss: 1.6711... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10290... Training loss: 1.7625... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10291... Training loss: 1.8654... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10292... Training loss: 1.7322... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10293... Training loss: 1.8507... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10294... Training loss: 1.7730... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10295... Training loss: 1.6414... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10296... Training loss: 1.7211... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10297... Training loss: 1.6518... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10298... Training loss: 1.7125... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10299... Training loss: 1.6204... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10300... Training loss: 1.8588... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10301... Training loss: 1.6706... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10302... Training loss: 1.8407... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10303... Training loss: 1.6595... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10304... Training loss: 1.8518... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10305... Training loss: 1.5923... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10306... Training loss: 1.6208... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10307... Training loss: 1.6081... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10308... Training loss: 1.6190... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10309... Training loss: 1.7884... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10310... Training loss: 1.6773... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10311... Training loss: 1.7565... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10312... Training loss: 1.7175... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10313... Training loss: 1.6168... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10314... Training loss: 1.6782... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10315... Training loss: 1.6988... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10316... Training loss: 1.7355... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10317... Training loss: 1.8103... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10318... Training loss: 1.6529... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10319... Training loss: 1.6908... 0.0467 sec/batch
Epoch: 3/20... Training Step: 10320... Training loss: 1.7186... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10321... Training loss: 1.6686... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10322... Training loss: 1.7568... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10323... Training loss: 1.8120... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10324... Training loss: 1.7199... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10325... Training loss: 1.7790... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10326... Training loss: 1.5739... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10327... Training loss: 1.5747... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10328... Training loss: 1.7672... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10329... Training loss: 1.6837... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10330... Training loss: 1.7229... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10331... Training loss: 1.8236... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10332... Training loss: 1.7417... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10333... Training loss: 1.9059... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10334... Training loss: 1.8522... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10335... Training loss: 1.7321... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10336... Training loss: 1.7266... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10337... Training loss: 1.7343... 0.0437 sec/batch
Epoch: 3/20... Training Step: 10338... Training loss: 1.6856... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10339... Training loss: 1.7569... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10340... Training loss: 1.7459... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10341... Training loss: 1.6697... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10342... Training loss: 1.6146... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10343... Training loss: 1.6498... 0.0460 sec/batch
Epoch: 3/20... Training Step: 10344... Training loss: 1.6635... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10345... Training loss: 1.5349... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10346... Training loss: 1.6746... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10347... Training loss: 1.5856... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10348... Training loss: 1.6360... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10349... Training loss: 1.5637... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10350... Training loss: 1.6418... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10351... Training loss: 1.6536... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10352... Training loss: 1.7451... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10353... Training loss: 1.8278... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10354... Training loss: 1.6897... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10355... Training loss: 1.6652... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10356... Training loss: 1.6085... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10357... Training loss: 1.6391... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10358... Training loss: 1.6667... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10359... Training loss: 1.6350... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10360... Training loss: 1.8145... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10361... Training loss: 1.7305... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10362... Training loss: 1.7711... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10363... Training loss: 1.7030... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10364... Training loss: 1.6450... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10365... Training loss: 1.5560... 0.0385 sec/batch
Epoch: 3/20... Training Step: 10366... Training loss: 1.7064... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10367... Training loss: 1.6571... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10368... Training loss: 1.7938... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10369... Training loss: 1.7324... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10370... Training loss: 1.6493... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10371... Training loss: 1.7115... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10372... Training loss: 1.8116... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10373... Training loss: 1.7479... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10374... Training loss: 1.6510... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10375... Training loss: 1.5584... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10376... Training loss: 1.7286... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10377... Training loss: 1.6672... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10378... Training loss: 1.7418... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10379... Training loss: 1.7453... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10380... Training loss: 1.7139... 0.0386 sec/batch
Epoch: 3/20... Training Step: 10381... Training loss: 1.6418... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10382... Training loss: 1.6961... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10383... Training loss: 1.6326... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10384... Training loss: 1.8242... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10385... Training loss: 1.6813... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10386... Training loss: 1.6106... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10387... Training loss: 1.7929... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10388... Training loss: 1.6079... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10389... Training loss: 1.6433... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10390... Training loss: 1.8151... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10391... Training loss: 1.8457... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10392... Training loss: 1.7388... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10393... Training loss: 1.7721... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10394... Training loss: 1.5979... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10395... Training loss: 1.6851... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10396... Training loss: 1.7568... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10397... Training loss: 1.8742... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10398... Training loss: 1.7360... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10399... Training loss: 1.7978... 0.0447 sec/batch
Epoch: 3/20... Training Step: 10400... Training loss: 1.7963... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10401... Training loss: 1.5843... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10402... Training loss: 1.7153... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10403... Training loss: 1.6446... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10404... Training loss: 1.5031... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10405... Training loss: 1.8166... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10406... Training loss: 1.6211... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10407... Training loss: 1.7903... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10408... Training loss: 1.7262... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10409... Training loss: 1.7049... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10410... Training loss: 1.7515... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10411... Training loss: 1.6681... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10412... Training loss: 1.6354... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10413... Training loss: 1.5705... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10414... Training loss: 1.6171... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10415... Training loss: 1.5005... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10416... Training loss: 1.5666... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10417... Training loss: 1.5565... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10418... Training loss: 1.6518... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10419... Training loss: 1.6138... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10420... Training loss: 1.7568... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10421... Training loss: 1.5591... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10422... Training loss: 1.6512... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10423... Training loss: 1.9211... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10424... Training loss: 1.6496... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10425... Training loss: 1.6305... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10426... Training loss: 1.6336... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10427... Training loss: 1.6167... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10428... Training loss: 1.5802... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10429... Training loss: 1.7410... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10430... Training loss: 1.7579... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10431... Training loss: 1.8134... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10432... Training loss: 1.7458... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10433... Training loss: 1.6883... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10434... Training loss: 1.6881... 0.0387 sec/batch
Epoch: 3/20... Training Step: 10435... Training loss: 1.6179... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10436... Training loss: 1.6486... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10437... Training loss: 1.6988... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10438... Training loss: 1.6316... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10439... Training loss: 1.7469... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10440... Training loss: 1.6775... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10441... Training loss: 1.5938... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10442... Training loss: 1.5865... 0.0453 sec/batch
Epoch: 3/20... Training Step: 10443... Training loss: 1.6045... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10444... Training loss: 1.6296... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10445... Training loss: 1.7924... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10446... Training loss: 1.7747... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10447... Training loss: 1.6596... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10448... Training loss: 1.7178... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10449... Training loss: 1.7127... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10450... Training loss: 1.7589... 0.0454 sec/batch
Epoch: 3/20... Training Step: 10451... Training loss: 1.6844... 0.0551 sec/batch
Epoch: 3/20... Training Step: 10452... Training loss: 1.5893... 0.0514 sec/batch
Epoch: 3/20... Training Step: 10453... Training loss: 1.5634... 0.0622 sec/batch
Epoch: 3/20... Training Step: 10454... Training loss: 1.7269... 0.0587 sec/batch
Epoch: 3/20... Training Step: 10455... Training loss: 1.5735... 0.0550 sec/batch
Epoch: 3/20... Training Step: 10456... Training loss: 1.7057... 0.0620 sec/batch
Epoch: 3/20... Training Step: 10457... Training loss: 1.6799... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10458... Training loss: 1.7396... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10459... Training loss: 1.7094... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10460... Training loss: 1.7588... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10461... Training loss: 1.5951... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10462... Training loss: 1.6354... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10463... Training loss: 1.6327... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10464... Training loss: 1.6977... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10465... Training loss: 1.8163... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10466... Training loss: 1.6055... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10467... Training loss: 1.8380... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10468... Training loss: 1.6544... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10469... Training loss: 1.7961... 0.0387 sec/batch
Epoch: 3/20... Training Step: 10470... Training loss: 1.6084... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10471... Training loss: 1.6321... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10472... Training loss: 1.6394... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10473... Training loss: 1.7578... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10474... Training loss: 1.6260... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10475... Training loss: 1.7196... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10476... Training loss: 1.6600... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10477... Training loss: 1.6540... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10478... Training loss: 1.7176... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10479... Training loss: 1.8265... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10480... Training loss: 1.7592... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10481... Training loss: 1.7040... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10482... Training loss: 1.7051... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10483... Training loss: 1.6124... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10484... Training loss: 1.7109... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10485... Training loss: 1.7017... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10486... Training loss: 1.7260... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10487... Training loss: 1.7260... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10488... Training loss: 1.7400... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10489... Training loss: 1.6949... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10490... Training loss: 1.7004... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10491... Training loss: 1.6454... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10492... Training loss: 1.7703... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10493... Training loss: 1.5302... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10494... Training loss: 1.5882... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10495... Training loss: 1.7014... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10496... Training loss: 1.7495... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10497... Training loss: 1.7552... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10498... Training loss: 1.6901... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10499... Training loss: 1.7354... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10500... Training loss: 1.6012... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10501... Training loss: 1.5482... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10502... Training loss: 1.5974... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10503... Training loss: 1.6692... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10504... Training loss: 1.5554... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10505... Training loss: 1.5261... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10506... Training loss: 1.7068... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10507... Training loss: 1.6448... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10508... Training loss: 1.7297... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10509... Training loss: 1.5688... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10510... Training loss: 1.6424... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10511... Training loss: 1.6493... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10512... Training loss: 1.6711... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10513... Training loss: 1.7620... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10514... Training loss: 1.6867... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10515... Training loss: 1.6089... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10516... Training loss: 1.6668... 0.0388 sec/batch
Epoch: 3/20... Training Step: 10517... Training loss: 1.7703... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10518... Training loss: 1.7215... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10519... Training loss: 1.6451... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10520... Training loss: 1.7544... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10521... Training loss: 1.6400... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10522... Training loss: 1.7418... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10523... Training loss: 1.6056... 0.0437 sec/batch
Epoch: 3/20... Training Step: 10524... Training loss: 1.6507... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10525... Training loss: 1.7429... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10526... Training loss: 1.8775... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10527... Training loss: 1.7343... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10528... Training loss: 1.8970... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10529... Training loss: 1.7464... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10530... Training loss: 1.7721... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10531... Training loss: 1.7130... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10532... Training loss: 1.6996... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10533... Training loss: 1.7213... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10534... Training loss: 1.6635... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10535... Training loss: 1.6107... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10536... Training loss: 1.6757... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10537... Training loss: 1.5407... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10538... Training loss: 1.6028... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10539... Training loss: 1.6152... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10540... Training loss: 1.6270... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10541... Training loss: 1.6151... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10542... Training loss: 1.6475... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10543... Training loss: 1.7707... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10544... Training loss: 1.7044... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10545... Training loss: 1.6473... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10546... Training loss: 1.6436... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10547... Training loss: 1.7097... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10548... Training loss: 1.7742... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10549... Training loss: 1.5936... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10550... Training loss: 1.7704... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10551... Training loss: 1.7585... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10552... Training loss: 1.6869... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10553... Training loss: 1.7129... 0.0437 sec/batch
Epoch: 3/20... Training Step: 10554... Training loss: 1.6780... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10555... Training loss: 1.4488... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10556... Training loss: 1.5919... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10557... Training loss: 1.6844... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10558... Training loss: 1.7053... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10559... Training loss: 1.5388... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10560... Training loss: 1.5653... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10561... Training loss: 1.8124... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10562... Training loss: 1.5101... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10563... Training loss: 1.7040... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10564... Training loss: 1.8213... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10565... Training loss: 1.7550... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10566... Training loss: 1.7146... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10567... Training loss: 1.7306... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10568... Training loss: 1.6561... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10569... Training loss: 1.5444... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10570... Training loss: 1.8185... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10571... Training loss: 1.7634... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10572... Training loss: 1.6416... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10573... Training loss: 1.7301... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10574... Training loss: 1.7129... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10575... Training loss: 1.5419... 0.0491 sec/batch
Epoch: 3/20... Training Step: 10576... Training loss: 1.8216... 0.0549 sec/batch
Epoch: 3/20... Training Step: 10577... Training loss: 1.7041... 0.0623 sec/batch
Epoch: 3/20... Training Step: 10578... Training loss: 1.6151... 0.0534 sec/batch
Epoch: 3/20... Training Step: 10579... Training loss: 1.7078... 0.0567 sec/batch
Epoch: 3/20... Training Step: 10580... Training loss: 1.6005... 0.0551 sec/batch
Epoch: 3/20... Training Step: 10581... Training loss: 1.7671... 0.0481 sec/batch
Epoch: 3/20... Training Step: 10582... Training loss: 1.7268... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10583... Training loss: 1.6349... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10584... Training loss: 1.6822... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10585... Training loss: 1.7115... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10586... Training loss: 1.6765... 0.0661 sec/batch
Epoch: 3/20... Training Step: 10587... Training loss: 1.6267... 0.0513 sec/batch
Epoch: 3/20... Training Step: 10588... Training loss: 1.7592... 0.0573 sec/batch
Epoch: 3/20... Training Step: 10589... Training loss: 1.6895... 0.0565 sec/batch
Epoch: 3/20... Training Step: 10590... Training loss: 1.8395... 0.0631 sec/batch
Epoch: 3/20... Training Step: 10591... Training loss: 1.8119... 0.0533 sec/batch
Epoch: 3/20... Training Step: 10592... Training loss: 1.7798... 0.0559 sec/batch
Epoch: 3/20... Training Step: 10593... Training loss: 1.8151... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10594... Training loss: 1.7526... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10595... Training loss: 1.6971... 0.0507 sec/batch
Epoch: 3/20... Training Step: 10596... Training loss: 1.7182... 0.0541 sec/batch
Epoch: 3/20... Training Step: 10597... Training loss: 1.7962... 0.0534 sec/batch
Epoch: 3/20... Training Step: 10598... Training loss: 1.6526... 0.0519 sec/batch
Epoch: 3/20... Training Step: 10599... Training loss: 1.6109... 0.0555 sec/batch
Epoch: 3/20... Training Step: 10600... Training loss: 1.7912... 0.0591 sec/batch
Epoch: 3/20... Training Step: 10601... Training loss: 1.5445... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10602... Training loss: 1.6108... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10603... Training loss: 1.5938... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10604... Training loss: 1.6503... 0.0471 sec/batch
Epoch: 3/20... Training Step: 10605... Training loss: 1.9098... 0.0614 sec/batch
Epoch: 3/20... Training Step: 10606... Training loss: 1.6828... 0.0598 sec/batch
Epoch: 3/20... Training Step: 10607... Training loss: 1.6769... 0.0558 sec/batch
Epoch: 3/20... Training Step: 10608... Training loss: 1.5549... 0.0582 sec/batch
Epoch: 3/20... Training Step: 10609... Training loss: 1.8110... 0.0621 sec/batch
Epoch: 3/20... Training Step: 10610... Training loss: 1.6817... 0.0552 sec/batch
Epoch: 3/20... Training Step: 10611... Training loss: 1.9701... 0.0542 sec/batch
Epoch: 3/20... Training Step: 10612... Training loss: 1.7363... 0.0643 sec/batch
Epoch: 3/20... Training Step: 10613... Training loss: 1.8192... 0.0576 sec/batch
Epoch: 3/20... Training Step: 10614... Training loss: 1.7960... 0.0647 sec/batch
Epoch: 3/20... Training Step: 10615... Training loss: 1.6015... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10616... Training loss: 1.7205... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10617... Training loss: 1.6656... 0.0580 sec/batch
Epoch: 3/20... Training Step: 10618... Training loss: 1.6842... 0.0558 sec/batch
Epoch: 3/20... Training Step: 10619... Training loss: 1.7283... 0.0656 sec/batch
Epoch: 3/20... Training Step: 10620... Training loss: 1.6229... 0.0519 sec/batch
Epoch: 3/20... Training Step: 10621... Training loss: 1.7857... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10622... Training loss: 1.6219... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10623... Training loss: 1.7648... 0.0470 sec/batch
Epoch: 3/20... Training Step: 10624... Training loss: 1.6056... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10625... Training loss: 1.6283... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10626... Training loss: 1.5724... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10627... Training loss: 1.6499... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10628... Training loss: 1.6049... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10629... Training loss: 1.5944... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10630... Training loss: 1.7233... 0.0608 sec/batch
Epoch: 3/20... Training Step: 10631... Training loss: 1.7586... 0.0622 sec/batch
Epoch: 3/20... Training Step: 10632... Training loss: 1.6948... 0.0585 sec/batch
Epoch: 3/20... Training Step: 10633... Training loss: 1.5709... 0.0576 sec/batch
Epoch: 3/20... Training Step: 10634... Training loss: 1.7958... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10635... Training loss: 1.9188... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10636... Training loss: 1.7028... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10637... Training loss: 1.6788... 0.0515 sec/batch
Epoch: 3/20... Training Step: 10638... Training loss: 1.7009... 0.0576 sec/batch
Epoch: 3/20... Training Step: 10639... Training loss: 1.8108... 0.0553 sec/batch
Epoch: 3/20... Training Step: 10640... Training loss: 1.7971... 0.0637 sec/batch
Epoch: 3/20... Training Step: 10641... Training loss: 1.7667... 0.0561 sec/batch
Epoch: 3/20... Training Step: 10642... Training loss: 1.6746... 0.0566 sec/batch
Epoch: 3/20... Training Step: 10643... Training loss: 1.7760... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10644... Training loss: 1.8243... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10645... Training loss: 1.8499... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10646... Training loss: 1.7309... 0.0566 sec/batch
Epoch: 3/20... Training Step: 10647... Training loss: 1.6452... 0.0620 sec/batch
Epoch: 3/20... Training Step: 10648... Training loss: 1.7210... 0.0604 sec/batch
Epoch: 3/20... Training Step: 10649... Training loss: 1.6630... 0.0546 sec/batch
Epoch: 3/20... Training Step: 10650... Training loss: 1.6441... 0.0558 sec/batch
Epoch: 3/20... Training Step: 10651... Training loss: 1.4979... 0.0548 sec/batch
Epoch: 3/20... Training Step: 10652... Training loss: 1.7866... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10653... Training loss: 1.7447... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10654... Training loss: 1.6675... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10655... Training loss: 1.5865... 0.0460 sec/batch
Epoch: 3/20... Training Step: 10656... Training loss: 1.7120... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10657... Training loss: 1.5518... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10658... Training loss: 1.6356... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10659... Training loss: 1.6757... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10660... Training loss: 1.5372... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10661... Training loss: 1.7266... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10662... Training loss: 1.6020... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10663... Training loss: 1.6600... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10664... Training loss: 1.7080... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10665... Training loss: 1.6649... 0.0472 sec/batch
Epoch: 3/20... Training Step: 10666... Training loss: 1.8811... 0.0468 sec/batch
Epoch: 3/20... Training Step: 10667... Training loss: 1.7409... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10668... Training loss: 1.8494... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10669... Training loss: 1.7203... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10670... Training loss: 1.6936... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10671... Training loss: 1.7316... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10672... Training loss: 1.6587... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10673... Training loss: 1.6750... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10674... Training loss: 1.6999... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10675... Training loss: 1.6828... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10676... Training loss: 1.7400... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10677... Training loss: 1.6079... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10678... Training loss: 1.6716... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10679... Training loss: 1.9015... 0.0465 sec/batch
Epoch: 3/20... Training Step: 10680... Training loss: 1.5976... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10681... Training loss: 1.7504... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10682... Training loss: 1.8046... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10683... Training loss: 1.8287... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10684... Training loss: 1.7547... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10685... Training loss: 1.7562... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10686... Training loss: 1.7995... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10687... Training loss: 1.6432... 0.0452 sec/batch
Epoch: 3/20... Training Step: 10688... Training loss: 1.8235... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10689... Training loss: 1.8518... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10690... Training loss: 1.6251... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10691... Training loss: 1.5973... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10692... Training loss: 1.8130... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10693... Training loss: 1.6059... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10694... Training loss: 1.6745... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10695... Training loss: 1.8806... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10696... Training loss: 1.7268... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10697... Training loss: 1.7455... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10698... Training loss: 1.7241... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10699... Training loss: 1.6718... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10700... Training loss: 1.7445... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10701... Training loss: 1.6679... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10702... Training loss: 1.6681... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10703... Training loss: 1.7978... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10704... Training loss: 1.5603... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10705... Training loss: 1.8563... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10706... Training loss: 1.6202... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10707... Training loss: 1.5828... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10708... Training loss: 1.6841... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10709... Training loss: 1.8308... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10710... Training loss: 1.7930... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10711... Training loss: 1.7007... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10712... Training loss: 1.8290... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10713... Training loss: 1.6391... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10714... Training loss: 1.7738... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10715... Training loss: 1.6714... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10716... Training loss: 1.6472... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10717... Training loss: 1.7161... 0.0388 sec/batch
Epoch: 3/20... Training Step: 10718... Training loss: 1.5938... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10719... Training loss: 1.6388... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10720... Training loss: 1.6529... 0.0438 sec/batch
Epoch: 3/20... Training Step: 10721... Training loss: 1.6688... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10722... Training loss: 1.9020... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10723... Training loss: 1.6028... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10724... Training loss: 1.5418... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10725... Training loss: 1.6344... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10726... Training loss: 1.7511... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10727... Training loss: 1.6414... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10728... Training loss: 1.6475... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10729... Training loss: 1.6164... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10730... Training loss: 1.6142... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10731... Training loss: 1.6774... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10732... Training loss: 1.5728... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10733... Training loss: 1.5342... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10734... Training loss: 1.4961... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10735... Training loss: 1.5454... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10736... Training loss: 1.7888... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10737... Training loss: 1.6027... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10738... Training loss: 1.7477... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10739... Training loss: 1.6648... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10740... Training loss: 1.5958... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10741... Training loss: 1.7080... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10742... Training loss: 1.7849... 0.0448 sec/batch
Epoch: 3/20... Training Step: 10743... Training loss: 1.6939... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10744... Training loss: 1.6804... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10745... Training loss: 1.7052... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10746... Training loss: 1.7597... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10747... Training loss: 1.7539... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10748... Training loss: 1.6727... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10749... Training loss: 1.7331... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10750... Training loss: 1.7568... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10751... Training loss: 1.6786... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10752... Training loss: 1.6235... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10753... Training loss: 1.7128... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10754... Training loss: 1.6802... 0.0391 sec/batch
Epoch: 3/20... Training Step: 10755... Training loss: 1.7975... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10756... Training loss: 1.8002... 0.0440 sec/batch
Epoch: 3/20... Training Step: 10757... Training loss: 1.6672... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10758... Training loss: 1.7755... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10759... Training loss: 1.6103... 0.0433 sec/batch
Epoch: 3/20... Training Step: 10760... Training loss: 1.7988... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10761... Training loss: 1.7631... 0.0464 sec/batch
Epoch: 3/20... Training Step: 10762... Training loss: 1.6466... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10763... Training loss: 1.8185... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10764... Training loss: 1.5747... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10765... Training loss: 1.9235... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10766... Training loss: 1.7545... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10767... Training loss: 1.6953... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10768... Training loss: 1.7584... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10769... Training loss: 1.7358... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10770... Training loss: 1.7767... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10771... Training loss: 1.7118... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10772... Training loss: 1.8555... 0.0437 sec/batch
Epoch: 3/20... Training Step: 10773... Training loss: 1.7513... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10774... Training loss: 1.7485... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10775... Training loss: 1.7237... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10776... Training loss: 1.8590... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10777... Training loss: 1.8209... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10778... Training loss: 1.7125... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10779... Training loss: 1.7611... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10780... Training loss: 1.7365... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10781... Training loss: 1.8212... 0.0447 sec/batch
Epoch: 3/20... Training Step: 10782... Training loss: 1.7925... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10783... Training loss: 1.5527... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10784... Training loss: 1.5687... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10785... Training loss: 1.7206... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10786... Training loss: 1.6879... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10787... Training loss: 1.7331... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10788... Training loss: 1.7535... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10789... Training loss: 1.6449... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10790... Training loss: 1.8445... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10791... Training loss: 1.7578... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10792... Training loss: 1.8428... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10793... Training loss: 1.8017... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10794... Training loss: 1.6321... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10795... Training loss: 1.6658... 0.0512 sec/batch
Epoch: 3/20... Training Step: 10796... Training loss: 1.8480... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10797... Training loss: 1.7094... 0.0452 sec/batch
Epoch: 3/20... Training Step: 10798... Training loss: 1.8247... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10799... Training loss: 1.5211... 0.0484 sec/batch
Epoch: 3/20... Training Step: 10800... Training loss: 1.6725... 0.0465 sec/batch
Epoch: 3/20... Training Step: 10801... Training loss: 1.7559... 0.0391 sec/batch
Epoch: 3/20... Training Step: 10802... Training loss: 1.5549... 0.0383 sec/batch
Epoch: 3/20... Training Step: 10803... Training loss: 1.6725... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10804... Training loss: 1.6369... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10805... Training loss: 1.5966... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10806... Training loss: 1.5872... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10807... Training loss: 1.6463... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10808... Training loss: 1.6821... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10809... Training loss: 1.7652... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10810... Training loss: 1.7103... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10811... Training loss: 1.6176... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10812... Training loss: 1.5747... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10813... Training loss: 1.7114... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10814... Training loss: 1.6615... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10815... Training loss: 1.7842... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10816... Training loss: 1.8493... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10817... Training loss: 1.7110... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10818... Training loss: 1.6264... 0.0392 sec/batch
Epoch: 3/20... Training Step: 10819... Training loss: 1.8720... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10820... Training loss: 1.6763... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10821... Training loss: 1.8390... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10822... Training loss: 1.7859... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10823... Training loss: 1.7665... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10824... Training loss: 1.6703... 0.0435 sec/batch
Epoch: 3/20... Training Step: 10825... Training loss: 1.7702... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10826... Training loss: 1.5520... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10827... Training loss: 1.6460... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10828... Training loss: 1.7428... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10829... Training loss: 1.6984... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10830... Training loss: 1.5988... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10831... Training loss: 1.8115... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10832... Training loss: 1.5758... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10833... Training loss: 1.7644... 0.0390 sec/batch
Epoch: 3/20... Training Step: 10834... Training loss: 1.6837... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10835... Training loss: 1.5721... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10836... Training loss: 1.6319... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10837... Training loss: 1.5788... 0.0403 sec/batch
Epoch: 3/20... Training Step: 10838... Training loss: 1.7544... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10839... Training loss: 1.6716... 0.0391 sec/batch
Epoch: 3/20... Training Step: 10840... Training loss: 1.5524... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10841... Training loss: 1.4899... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10842... Training loss: 1.6520... 0.0448 sec/batch
Epoch: 3/20... Training Step: 10843... Training loss: 1.5814... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10844... Training loss: 1.7037... 0.0419 sec/batch
Epoch: 3/20... Training Step: 10845... Training loss: 1.7234... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10846... Training loss: 1.7077... 0.0400 sec/batch
Epoch: 3/20... Training Step: 10847... Training loss: 1.7114... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10848... Training loss: 1.6775... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10849... Training loss: 1.5564... 0.0399 sec/batch
Epoch: 3/20... Training Step: 10850... Training loss: 1.7127... 0.0406 sec/batch
Epoch: 3/20... Training Step: 10851... Training loss: 1.5755... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10852... Training loss: 1.7942... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10853... Training loss: 1.7310... 0.0385 sec/batch
Epoch: 3/20... Training Step: 10854... Training loss: 1.7322... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10855... Training loss: 1.8357... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10856... Training loss: 1.6723... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10857... Training loss: 1.7146... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10858... Training loss: 1.7337... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10859... Training loss: 1.7636... 0.0393 sec/batch
Epoch: 3/20... Training Step: 10860... Training loss: 1.8184... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10861... Training loss: 1.8198... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10862... Training loss: 1.6305... 0.0440 sec/batch
Epoch: 3/20... Training Step: 10863... Training loss: 1.6449... 0.0395 sec/batch
Epoch: 3/20... Training Step: 10864... Training loss: 1.8272... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10865... Training loss: 1.6047... 0.0389 sec/batch
Epoch: 3/20... Training Step: 10866... Training loss: 1.6288... 0.0423 sec/batch
Epoch: 3/20... Training Step: 10867... Training loss: 1.7138... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10868... Training loss: 1.7834... 0.0407 sec/batch
Epoch: 3/20... Training Step: 10869... Training loss: 1.5654... 0.0397 sec/batch
Epoch: 3/20... Training Step: 10870... Training loss: 1.5977... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10871... Training loss: 1.6495... 0.0411 sec/batch
Epoch: 3/20... Training Step: 10872... Training loss: 1.5513... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10873... Training loss: 1.7003... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10874... Training loss: 1.6273... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10875... Training loss: 1.6145... 0.0408 sec/batch
Epoch: 3/20... Training Step: 10876... Training loss: 1.7404... 0.0405 sec/batch
Epoch: 3/20... Training Step: 10877... Training loss: 1.7057... 0.0404 sec/batch
Epoch: 3/20... Training Step: 10878... Training loss: 1.6540... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10879... Training loss: 1.6043... 0.0417 sec/batch
Epoch: 3/20... Training Step: 10880... Training loss: 1.7431... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10881... Training loss: 1.7931... 0.0481 sec/batch
Epoch: 3/20... Training Step: 10882... Training loss: 1.8023... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10883... Training loss: 1.6849... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10884... Training loss: 1.6461... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10885... Training loss: 1.6672... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10886... Training loss: 1.7208... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10887... Training loss: 1.6305... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10888... Training loss: 1.6113... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10889... Training loss: 1.5318... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10890... Training loss: 1.5348... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10891... Training loss: 1.6380... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10892... Training loss: 1.6050... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10893... Training loss: 1.6226... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10894... Training loss: 1.5740... 0.0412 sec/batch
Epoch: 3/20... Training Step: 10895... Training loss: 1.5444... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10896... Training loss: 1.6351... 0.0437 sec/batch
Epoch: 3/20... Training Step: 10897... Training loss: 1.6086... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10898... Training loss: 1.6133... 0.0468 sec/batch
Epoch: 3/20... Training Step: 10899... Training loss: 1.6969... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10900... Training loss: 1.7441... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10901... Training loss: 1.6530... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10902... Training loss: 1.7602... 0.0421 sec/batch
Epoch: 3/20... Training Step: 10903... Training loss: 1.6330... 0.0428 sec/batch
Epoch: 3/20... Training Step: 10904... Training loss: 1.5441... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10905... Training loss: 1.5903... 0.0414 sec/batch
Epoch: 3/20... Training Step: 10906... Training loss: 1.6269... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10907... Training loss: 1.6149... 0.0466 sec/batch
Epoch: 3/20... Training Step: 10908... Training loss: 1.7460... 0.0416 sec/batch
Epoch: 3/20... Training Step: 10909... Training loss: 1.5953... 0.0455 sec/batch
Epoch: 3/20... Training Step: 10910... Training loss: 1.5909... 0.0418 sec/batch
Epoch: 3/20... Training Step: 10911... Training loss: 1.7967... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10912... Training loss: 1.6316... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10913... Training loss: 1.5757... 0.0429 sec/batch
Epoch: 3/20... Training Step: 10914... Training loss: 1.7808... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10915... Training loss: 1.6421... 0.0422 sec/batch
Epoch: 3/20... Training Step: 10916... Training loss: 1.7299... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10917... Training loss: 1.5803... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10918... Training loss: 1.7752... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10919... Training loss: 1.7283... 0.0444 sec/batch
Epoch: 3/20... Training Step: 10920... Training loss: 1.8574... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10921... Training loss: 1.9090... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10922... Training loss: 1.7716... 0.0427 sec/batch
Epoch: 3/20... Training Step: 10923... Training loss: 1.5125... 0.0443 sec/batch
Epoch: 3/20... Training Step: 10924... Training loss: 1.5733... 0.0430 sec/batch
Epoch: 3/20... Training Step: 10925... Training loss: 1.6329... 0.0447 sec/batch
Epoch: 3/20... Training Step: 10926... Training loss: 1.7059... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10927... Training loss: 1.6717... 0.0454 sec/batch
Epoch: 3/20... Training Step: 10928... Training loss: 1.8232... 0.0439 sec/batch
Epoch: 3/20... Training Step: 10929... Training loss: 1.7595... 0.0434 sec/batch
Epoch: 3/20... Training Step: 10930... Training loss: 1.7286... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10931... Training loss: 1.7918... 0.0464 sec/batch
Epoch: 3/20... Training Step: 10932... Training loss: 1.7241... 0.0409 sec/batch
Epoch: 3/20... Training Step: 10933... Training loss: 1.5540... 0.0468 sec/batch
Epoch: 3/20... Training Step: 10934... Training loss: 1.7481... 0.0432 sec/batch
Epoch: 3/20... Training Step: 10935... Training loss: 1.7821... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10936... Training loss: 1.8190... 0.0459 sec/batch
Epoch: 3/20... Training Step: 10937... Training loss: 1.8334... 0.0454 sec/batch
Epoch: 3/20... Training Step: 10938... Training loss: 1.7479... 0.0420 sec/batch
Epoch: 3/20... Training Step: 10939... Training loss: 1.5711... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10940... Training loss: 1.6784... 0.0445 sec/batch
Epoch: 3/20... Training Step: 10941... Training loss: 1.6313... 0.0442 sec/batch
Epoch: 3/20... Training Step: 10942... Training loss: 1.4975... 0.0465 sec/batch
Epoch: 3/20... Training Step: 10943... Training loss: 1.6682... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10944... Training loss: 1.6294... 0.0424 sec/batch
Epoch: 3/20... Training Step: 10945... Training loss: 1.8601... 0.0425 sec/batch
Epoch: 3/20... Training Step: 10946... Training loss: 1.6104... 0.0480 sec/batch
Epoch: 3/20... Training Step: 10947... Training loss: 1.5862... 0.0426 sec/batch
Epoch: 3/20... Training Step: 10948... Training loss: 1.7922... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10949... Training loss: 1.6261... 0.0462 sec/batch
Epoch: 3/20... Training Step: 10950... Training loss: 1.7882... 0.0458 sec/batch
Epoch: 3/20... Training Step: 10951... Training loss: 1.5919... 0.0495 sec/batch
Epoch: 3/20... Training Step: 10952... Training loss: 1.8168... 0.0472 sec/batch
Epoch: 3/20... Training Step: 10953... Training loss: 1.5992... 0.0466 sec/batch
Epoch: 3/20... Training Step: 10954... Training loss: 1.7165... 0.0454 sec/batch
Epoch: 3/20... Training Step: 10955... Training loss: 1.7691... 0.0458 sec/batch
Epoch: 3/20... Training Step: 10956... Training loss: 1.6569... 0.0496 sec/batch
Epoch: 3/20... Training Step: 10957... Training loss: 1.6496... 0.0480 sec/batch
Epoch: 3/20... Training Step: 10958... Training loss: 1.7186... 0.0482 sec/batch
Epoch: 3/20... Training Step: 10959... Training loss: 1.6003... 0.0480 sec/batch
Epoch: 3/20... Training Step: 10960... Training loss: 1.8414... 0.0473 sec/batch
Epoch: 3/20... Training Step: 10961... Training loss: 1.7327... 0.0541 sec/batch
Epoch: 3/20... Training Step: 10962... Training loss: 1.7131... 0.0704 sec/batch
Epoch: 3/20... Training Step: 10963... Training loss: 1.6538... 0.0555 sec/batch
Epoch: 3/20... Training Step: 10964... Training loss: 1.6488... 0.0413 sec/batch
Epoch: 3/20... Training Step: 10965... Training loss: 1.6296... 0.0431 sec/batch
Epoch: 3/20... Training Step: 10966... Training loss: 1.6271... 0.0451 sec/batch
Epoch: 3/20... Training Step: 10967... Training loss: 1.6658... 0.0460 sec/batch
Epoch: 3/20... Training Step: 10968... Training loss: 1.6849... 0.0394 sec/batch
Epoch: 3/20... Training Step: 10969... Training loss: 1.8742... 0.0398 sec/batch
Epoch: 3/20... Training Step: 10970... Training loss: 1.6289... 0.0441 sec/batch
Epoch: 3/20... Training Step: 10971... Training loss: 1.7785... 0.0401 sec/batch
Epoch: 3/20... Training Step: 10972... Training loss: 1.8128... 0.0462 sec/batch
Epoch: 3/20... Training Step: 10973... Training loss: 1.7094... 0.0499 sec/batch
Epoch: 3/20... Training Step: 10974... Training loss: 1.7535... 0.0475 sec/batch
Epoch: 3/20... Training Step: 10975... Training loss: 1.4754... 0.0563 sec/batch
Epoch: 3/20... Training Step: 10976... Training loss: 1.8679... 0.0474 sec/batch
Epoch: 3/20... Training Step: 10977... Training loss: 1.7276... 0.0481 sec/batch
Epoch: 3/20... Training Step: 10978... Training loss: 1.8044... 0.0476 sec/batch
Epoch: 3/20... Training Step: 10979... Training loss: 1.6364... 0.0470 sec/batch
Epoch: 3/20... Training Step: 10980... Training loss: 1.8307... 0.0497 sec/batch
Epoch: 3/20... Training Step: 10981... Training loss: 1.6060... 0.0690 sec/batch
Epoch: 3/20... Training Step: 10982... Training loss: 1.7024... 0.0607 sec/batch
Epoch: 3/20... Training Step: 10983... Training loss: 1.7251... 0.0562 sec/batch
Epoch: 3/20... Training Step: 10984... Training loss: 1.7287... 0.0493 sec/batch
Epoch: 3/20... Training Step: 10985... Training loss: 1.7405... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10986... Training loss: 1.7635... 0.0499 sec/batch
Epoch: 3/20... Training Step: 10987... Training loss: 1.5847... 0.0450 sec/batch
Epoch: 3/20... Training Step: 10988... Training loss: 1.8728... 0.0449 sec/batch
Epoch: 3/20... Training Step: 10989... Training loss: 1.6937... 0.0483 sec/batch
Epoch: 3/20... Training Step: 10990... Training loss: 1.8039... 0.0460 sec/batch
Epoch: 3/20... Training Step: 10991... Training loss: 1.7427... 0.0446 sec/batch
Epoch: 3/20... Training Step: 10992... Training loss: 1.7499... 0.0436 sec/batch
Epoch: 3/20... Training Step: 10993... Training loss: 1.8985... 0.0415 sec/batch
Epoch: 3/20... Training Step: 10994... Training loss: 1.6902... 0.0402 sec/batch
Epoch: 3/20... Training Step: 10995... Training loss: 1.6458... 0.0396 sec/batch
Epoch: 3/20... Training Step: 10996... Training loss: 1.7401... 0.0502 sec/batch
Epoch: 3/20... Training Step: 10997... Training loss: 1.7261... 0.0444 sec/batch
Epoch: 3/20... Training Step: 10998... Training loss: 1.6863... 0.0410 sec/batch
Epoch: 3/20... Training Step: 10999... Training loss: 1.8614... 0.0676 sec/batch
Epoch: 3/20... Training Step: 11000... Training loss: 1.6808... 0.0561 sec/batch
Epoch: 3/20... Training Step: 11001... Training loss: 1.8324... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11002... Training loss: 1.7341... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11003... Training loss: 1.7759... 0.0451 sec/batch
Epoch: 3/20... Training Step: 11004... Training loss: 1.6301... 0.0577 sec/batch
Epoch: 3/20... Training Step: 11005... Training loss: 1.6500... 0.0545 sec/batch
Epoch: 3/20... Training Step: 11006... Training loss: 1.6849... 0.0476 sec/batch
Epoch: 3/20... Training Step: 11007... Training loss: 1.7812... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11008... Training loss: 1.6943... 0.0421 sec/batch
Epoch: 3/20... Training Step: 11009... Training loss: 1.7428... 0.0409 sec/batch
Epoch: 3/20... Training Step: 11010... Training loss: 1.7114... 0.0528 sec/batch
Epoch: 3/20... Training Step: 11011... Training loss: 1.6110... 0.0449 sec/batch
Epoch: 3/20... Training Step: 11012... Training loss: 1.6679... 0.0404 sec/batch
Epoch: 3/20... Training Step: 11013... Training loss: 1.7626... 0.0440 sec/batch
Epoch: 3/20... Training Step: 11014... Training loss: 1.5601... 0.0428 sec/batch
Epoch: 3/20... Training Step: 11015... Training loss: 1.7828... 0.0419 sec/batch
Epoch: 3/20... Training Step: 11016... Training loss: 1.6937... 0.0429 sec/batch
Epoch: 3/20... Training Step: 11017... Training loss: 1.6566... 0.0408 sec/batch
Epoch: 3/20... Training Step: 11018... Training loss: 1.6515... 0.0414 sec/batch
Epoch: 3/20... Training Step: 11019... Training loss: 1.5429... 0.0407 sec/batch
Epoch: 3/20... Training Step: 11020... Training loss: 1.6706... 0.0406 sec/batch
Epoch: 3/20... Training Step: 11021... Training loss: 1.6798... 0.0429 sec/batch
Epoch: 3/20... Training Step: 11022... Training loss: 1.5793... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11023... Training loss: 1.7215... 0.0441 sec/batch
Epoch: 3/20... Training Step: 11024... Training loss: 1.5883... 0.0403 sec/batch
Epoch: 3/20... Training Step: 11025... Training loss: 1.7178... 0.0426 sec/batch
Epoch: 3/20... Training Step: 11026... Training loss: 1.5667... 0.0449 sec/batch
Epoch: 3/20... Training Step: 11027... Training loss: 1.6358... 0.0412 sec/batch
Epoch: 3/20... Training Step: 11028... Training loss: 1.7247... 0.0408 sec/batch
Epoch: 3/20... Training Step: 11029... Training loss: 1.6629... 0.0405 sec/batch
Epoch: 3/20... Training Step: 11030... Training loss: 1.6004... 0.0401 sec/batch
Epoch: 3/20... Training Step: 11031... Training loss: 1.5316... 0.0404 sec/batch
Epoch: 3/20... Training Step: 11032... Training loss: 1.6352... 0.0412 sec/batch
Epoch: 3/20... Training Step: 11033... Training loss: 1.7292... 0.0396 sec/batch
Epoch: 3/20... Training Step: 11034... Training loss: 1.8494... 0.0421 sec/batch
Epoch: 3/20... Training Step: 11035... Training loss: 1.7299... 0.0598 sec/batch
Epoch: 3/20... Training Step: 11036... Training loss: 1.6709... 0.0662 sec/batch
Epoch: 3/20... Training Step: 11037... Training loss: 1.7140... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11038... Training loss: 1.7277... 0.0488 sec/batch
Epoch: 3/20... Training Step: 11039... Training loss: 1.6189... 0.0593 sec/batch
Epoch: 3/20... Training Step: 11040... Training loss: 1.6123... 0.0585 sec/batch
Epoch: 3/20... Training Step: 11041... Training loss: 1.7503... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11042... Training loss: 1.6705... 0.0511 sec/batch
Epoch: 3/20... Training Step: 11043... Training loss: 1.5189... 0.0591 sec/batch
Epoch: 3/20... Training Step: 11044... Training loss: 1.6713... 0.0496 sec/batch
Epoch: 3/20... Training Step: 11045... Training loss: 1.5846... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11046... Training loss: 1.6064... 0.0535 sec/batch
Epoch: 3/20... Training Step: 11047... Training loss: 1.7588... 0.0627 sec/batch
Epoch: 3/20... Training Step: 11048... Training loss: 1.5687... 0.0517 sec/batch
Epoch: 3/20... Training Step: 11049... Training loss: 1.5657... 0.0504 sec/batch
Epoch: 3/20... Training Step: 11050... Training loss: 1.6811... 0.0563 sec/batch
Epoch: 3/20... Training Step: 11051... Training loss: 1.6614... 0.0534 sec/batch
Epoch: 3/20... Training Step: 11052... Training loss: 1.6938... 0.0585 sec/batch
Epoch: 3/20... Training Step: 11053... Training loss: 1.6991... 0.0477 sec/batch
Epoch: 3/20... Training Step: 11054... Training loss: 1.6872... 0.0544 sec/batch
Epoch: 3/20... Training Step: 11055... Training loss: 1.6468... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11056... Training loss: 1.6394... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11057... Training loss: 1.5428... 0.0462 sec/batch
Epoch: 3/20... Training Step: 11058... Training loss: 1.6952... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11059... Training loss: 1.7310... 0.0589 sec/batch
Epoch: 3/20... Training Step: 11060... Training loss: 1.6133... 0.0561 sec/batch
Epoch: 3/20... Training Step: 11061... Training loss: 1.6519... 0.0546 sec/batch
Epoch: 3/20... Training Step: 11062... Training loss: 1.5880... 0.0484 sec/batch
Epoch: 3/20... Training Step: 11063... Training loss: 1.6725... 0.0491 sec/batch
Epoch: 3/20... Training Step: 11064... Training loss: 1.7013... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11065... Training loss: 1.6390... 0.0455 sec/batch
Epoch: 3/20... Training Step: 11066... Training loss: 1.7070... 0.0496 sec/batch
Epoch: 3/20... Training Step: 11067... Training loss: 1.7349... 0.0481 sec/batch
Epoch: 3/20... Training Step: 11068... Training loss: 1.5989... 0.0517 sec/batch
Epoch: 3/20... Training Step: 11069... Training loss: 1.5316... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11070... Training loss: 1.6456... 0.0488 sec/batch
Epoch: 3/20... Training Step: 11071... Training loss: 1.6099... 0.0539 sec/batch
Epoch: 3/20... Training Step: 11072... Training loss: 1.7218... 0.0504 sec/batch
Epoch: 3/20... Training Step: 11073... Training loss: 1.8092... 0.0501 sec/batch
Epoch: 3/20... Training Step: 11074... Training loss: 1.6802... 0.0549 sec/batch
Epoch: 3/20... Training Step: 11075... Training loss: 1.8201... 0.0914 sec/batch
Epoch: 3/20... Training Step: 11076... Training loss: 1.7234... 0.0828 sec/batch
Epoch: 3/20... Training Step: 11077... Training loss: 1.6124... 0.0912 sec/batch
Epoch: 3/20... Training Step: 11078... Training loss: 1.6091... 0.0888 sec/batch
Epoch: 3/20... Training Step: 11079... Training loss: 1.7020... 0.0632 sec/batch
Epoch: 3/20... Training Step: 11080... Training loss: 1.8311... 0.0614 sec/batch
Epoch: 3/20... Training Step: 11081... Training loss: 1.7255... 0.0536 sec/batch
Epoch: 3/20... Training Step: 11082... Training loss: 1.5586... 0.0530 sec/batch
Epoch: 3/20... Training Step: 11083... Training loss: 1.6804... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11084... Training loss: 1.5423... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11085... Training loss: 1.7602... 0.0458 sec/batch
Epoch: 3/20... Training Step: 11086... Training loss: 1.6412... 0.0612 sec/batch
Epoch: 3/20... Training Step: 11087... Training loss: 1.6353... 0.0564 sec/batch
Epoch: 3/20... Training Step: 11088... Training loss: 1.6932... 0.0681 sec/batch
Epoch: 3/20... Training Step: 11089... Training loss: 1.6617... 0.0785 sec/batch
Epoch: 3/20... Training Step: 11090... Training loss: 1.5731... 0.0759 sec/batch
Epoch: 3/20... Training Step: 11091... Training loss: 1.6880... 0.0576 sec/batch
Epoch: 3/20... Training Step: 11092... Training loss: 1.7134... 0.0738 sec/batch
Epoch: 3/20... Training Step: 11093... Training loss: 1.6913... 0.0839 sec/batch
Epoch: 3/20... Training Step: 11094... Training loss: 1.5354... 0.0701 sec/batch
Epoch: 3/20... Training Step: 11095... Training loss: 1.7028... 0.0689 sec/batch
Epoch: 3/20... Training Step: 11096... Training loss: 1.7654... 0.0525 sec/batch
Epoch: 3/20... Training Step: 11097... Training loss: 1.5666... 0.0506 sec/batch
Epoch: 3/20... Training Step: 11098... Training loss: 1.6188... 0.0524 sec/batch
Epoch: 3/20... Training Step: 11099... Training loss: 1.5645... 0.0608 sec/batch
Epoch: 3/20... Training Step: 11100... Training loss: 1.6088... 0.0509 sec/batch
Epoch: 3/20... Training Step: 11101... Training loss: 1.7589... 0.0565 sec/batch
Epoch: 3/20... Training Step: 11102... Training loss: 1.7164... 0.0550 sec/batch
Epoch: 3/20... Training Step: 11103... Training loss: 1.7889... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11104... Training loss: 1.6591... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11105... Training loss: 1.5582... 0.0628 sec/batch
Epoch: 3/20... Training Step: 11106... Training loss: 1.6323... 0.0510 sec/batch
Epoch: 3/20... Training Step: 11107... Training loss: 1.6200... 0.0436 sec/batch
Epoch: 3/20... Training Step: 11108... Training loss: 1.6219... 0.0682 sec/batch
Epoch: 3/20... Training Step: 11109... Training loss: 1.6514... 0.0597 sec/batch
Epoch: 3/20... Training Step: 11110... Training loss: 1.7429... 0.0496 sec/batch
Epoch: 3/20... Training Step: 11111... Training loss: 1.6854... 0.0566 sec/batch
Epoch: 3/20... Training Step: 11112... Training loss: 1.5999... 0.0722 sec/batch
Epoch: 3/20... Training Step: 11113... Training loss: 1.7146... 0.0678 sec/batch
Epoch: 3/20... Training Step: 11114... Training loss: 1.5754... 0.0595 sec/batch
Epoch: 3/20... Training Step: 11115... Training loss: 1.6184... 0.0482 sec/batch
Epoch: 3/20... Training Step: 11116... Training loss: 1.6932... 0.0467 sec/batch
Epoch: 3/20... Training Step: 11117... Training loss: 1.8177... 0.0552 sec/batch
Epoch: 3/20... Training Step: 11118... Training loss: 1.5679... 0.0494 sec/batch
Epoch: 3/20... Training Step: 11119... Training loss: 1.7309... 0.0607 sec/batch
Epoch: 3/20... Training Step: 11120... Training loss: 1.6086... 0.0458 sec/batch
Epoch: 3/20... Training Step: 11121... Training loss: 1.6778... 0.0564 sec/batch
Epoch: 3/20... Training Step: 11122... Training loss: 1.5976... 0.0586 sec/batch
Epoch: 3/20... Training Step: 11123... Training loss: 1.6728... 0.0668 sec/batch
Epoch: 3/20... Training Step: 11124... Training loss: 1.6395... 0.0679 sec/batch
Epoch: 3/20... Training Step: 11125... Training loss: 1.7956... 0.0648 sec/batch
Epoch: 3/20... Training Step: 11126... Training loss: 1.7577... 0.0530 sec/batch
Epoch: 3/20... Training Step: 11127... Training loss: 1.5381... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11128... Training loss: 1.7461... 0.0897 sec/batch
Epoch: 3/20... Training Step: 11129... Training loss: 1.6591... 0.0721 sec/batch
Epoch: 3/20... Training Step: 11130... Training loss: 1.6074... 0.0761 sec/batch
Epoch: 3/20... Training Step: 11131... Training loss: 1.5077... 0.0902 sec/batch
Epoch: 3/20... Training Step: 11132... Training loss: 1.5984... 0.0603 sec/batch
Epoch: 3/20... Training Step: 11133... Training loss: 1.7591... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11134... Training loss: 1.6515... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11135... Training loss: 1.6582... 0.0474 sec/batch
Epoch: 3/20... Training Step: 11136... Training loss: 1.6509... 0.0496 sec/batch
Epoch: 3/20... Training Step: 11137... Training loss: 1.6584... 0.0474 sec/batch
Epoch: 3/20... Training Step: 11138... Training loss: 1.7264... 0.0607 sec/batch
Epoch: 3/20... Training Step: 11139... Training loss: 1.6423... 0.0566 sec/batch
Epoch: 3/20... Training Step: 11140... Training loss: 1.7218... 0.0608 sec/batch
Epoch: 3/20... Training Step: 11141... Training loss: 1.7276... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11142... Training loss: 1.6567... 0.0500 sec/batch
Epoch: 3/20... Training Step: 11143... Training loss: 1.5900... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11144... Training loss: 1.8254... 0.0514 sec/batch
Epoch: 3/20... Training Step: 11145... Training loss: 1.6760... 0.0543 sec/batch
Epoch: 3/20... Training Step: 11146... Training loss: 1.7359... 0.0597 sec/batch
Epoch: 3/20... Training Step: 11147... Training loss: 1.6839... 0.0463 sec/batch
Epoch: 3/20... Training Step: 11148... Training loss: 1.7633... 0.0452 sec/batch
Epoch: 3/20... Training Step: 11149... Training loss: 1.5626... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11150... Training loss: 1.5507... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11151... Training loss: 1.8168... 0.0493 sec/batch
Epoch: 3/20... Training Step: 11152... Training loss: 1.8053... 0.0628 sec/batch
Epoch: 3/20... Training Step: 11153... Training loss: 1.7472... 0.0721 sec/batch
Epoch: 3/20... Training Step: 11154... Training loss: 1.7387... 0.0629 sec/batch
Epoch: 3/20... Training Step: 11155... Training loss: 1.6487... 0.0440 sec/batch
Epoch: 3/20... Training Step: 11156... Training loss: 1.7210... 0.0508 sec/batch
Epoch: 3/20... Training Step: 11157... Training loss: 1.5728... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11158... Training loss: 1.5662... 0.0596 sec/batch
Epoch: 3/20... Training Step: 11159... Training loss: 1.6281... 0.0467 sec/batch
Epoch: 3/20... Training Step: 11160... Training loss: 1.7319... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11161... Training loss: 1.6064... 0.0540 sec/batch
Epoch: 3/20... Training Step: 11162... Training loss: 1.7700... 0.0574 sec/batch
Epoch: 3/20... Training Step: 11163... Training loss: 1.6947... 0.0533 sec/batch
Epoch: 3/20... Training Step: 11164... Training loss: 1.6686... 0.0486 sec/batch
Epoch: 3/20... Training Step: 11165... Training loss: 1.6793... 0.0426 sec/batch
Epoch: 3/20... Training Step: 11166... Training loss: 1.5607... 0.0556 sec/batch
Epoch: 3/20... Training Step: 11167... Training loss: 1.6631... 0.0487 sec/batch
Epoch: 3/20... Training Step: 11168... Training loss: 1.6927... 0.0446 sec/batch
Epoch: 3/20... Training Step: 11169... Training loss: 1.6445... 0.0434 sec/batch
Epoch: 3/20... Training Step: 11170... Training loss: 1.7350... 0.0515 sec/batch
Epoch: 3/20... Training Step: 11171... Training loss: 1.7978... 0.0556 sec/batch
Epoch: 3/20... Training Step: 11172... Training loss: 1.6136... 0.0448 sec/batch
Epoch: 3/20... Training Step: 11173... Training loss: 1.7087... 0.0429 sec/batch
Epoch: 3/20... Training Step: 11174... Training loss: 1.5804... 0.0445 sec/batch
Epoch: 3/20... Training Step: 11175... Training loss: 1.6532... 0.0527 sec/batch
Epoch: 3/20... Training Step: 11176... Training loss: 1.6938... 0.0476 sec/batch
Epoch: 3/20... Training Step: 11177... Training loss: 1.6490... 0.0522 sec/batch
Epoch: 3/20... Training Step: 11178... Training loss: 1.6457... 0.0556 sec/batch
Epoch: 3/20... Training Step: 11179... Training loss: 1.6807... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11180... Training loss: 1.5887... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11181... Training loss: 1.7040... 0.0530 sec/batch
Epoch: 3/20... Training Step: 11182... Training loss: 1.5811... 0.0476 sec/batch
Epoch: 3/20... Training Step: 11183... Training loss: 1.7154... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11184... Training loss: 1.7207... 0.0458 sec/batch
Epoch: 3/20... Training Step: 11185... Training loss: 1.6905... 0.0508 sec/batch
Epoch: 3/20... Training Step: 11186... Training loss: 1.7414... 0.0529 sec/batch
Epoch: 3/20... Training Step: 11187... Training loss: 1.7171... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11188... Training loss: 1.7523... 0.0491 sec/batch
Epoch: 3/20... Training Step: 11189... Training loss: 1.8044... 0.0576 sec/batch
Epoch: 3/20... Training Step: 11190... Training loss: 1.7086... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11191... Training loss: 1.8608... 0.0532 sec/batch
Epoch: 3/20... Training Step: 11192... Training loss: 1.7917... 0.0581 sec/batch
Epoch: 3/20... Training Step: 11193... Training loss: 1.9480... 0.0489 sec/batch
Epoch: 3/20... Training Step: 11194... Training loss: 1.6583... 0.0544 sec/batch
Epoch: 3/20... Training Step: 11195... Training loss: 1.6635... 0.0445 sec/batch
Epoch: 3/20... Training Step: 11196... Training loss: 1.7328... 0.0579 sec/batch
Epoch: 3/20... Training Step: 11197... Training loss: 1.5544... 0.0491 sec/batch
Epoch: 3/20... Training Step: 11198... Training loss: 1.6529... 0.0605 sec/batch
Epoch: 3/20... Training Step: 11199... Training loss: 1.7934... 0.0659 sec/batch
Epoch: 3/20... Training Step: 11200... Training loss: 1.6920... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11201... Training loss: 1.7666... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11202... Training loss: 1.6283... 0.0527 sec/batch
Epoch: 3/20... Training Step: 11203... Training loss: 1.6980... 0.0494 sec/batch
Epoch: 3/20... Training Step: 11204... Training loss: 1.5757... 0.0651 sec/batch
Epoch: 3/20... Training Step: 11205... Training loss: 1.6341... 0.0582 sec/batch
Epoch: 3/20... Training Step: 11206... Training loss: 1.6968... 0.0489 sec/batch
Epoch: 3/20... Training Step: 11207... Training loss: 1.5780... 0.0604 sec/batch
Epoch: 3/20... Training Step: 11208... Training loss: 1.7085... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11209... Training loss: 1.5321... 0.0493 sec/batch
Epoch: 3/20... Training Step: 11210... Training loss: 1.7533... 0.0526 sec/batch
Epoch: 3/20... Training Step: 11211... Training loss: 1.7199... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11212... Training loss: 1.6374... 0.0654 sec/batch
Epoch: 3/20... Training Step: 11213... Training loss: 1.7437... 0.0652 sec/batch
Epoch: 3/20... Training Step: 11214... Training loss: 1.5716... 0.0458 sec/batch
Epoch: 3/20... Training Step: 11215... Training loss: 1.6278... 0.0738 sec/batch
Epoch: 3/20... Training Step: 11216... Training loss: 1.6669... 0.0792 sec/batch
Epoch: 3/20... Training Step: 11217... Training loss: 1.6981... 0.0803 sec/batch
Epoch: 3/20... Training Step: 11218... Training loss: 1.6792... 0.0580 sec/batch
Epoch: 3/20... Training Step: 11219... Training loss: 1.6434... 0.0616 sec/batch
Epoch: 3/20... Training Step: 11220... Training loss: 1.5305... 0.0481 sec/batch
Epoch: 3/20... Training Step: 11221... Training loss: 1.6600... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11222... Training loss: 1.6153... 0.0471 sec/batch
Epoch: 3/20... Training Step: 11223... Training loss: 1.6885... 0.0793 sec/batch
Epoch: 3/20... Training Step: 11224... Training loss: 1.6953... 0.0573 sec/batch
Epoch: 3/20... Training Step: 11225... Training loss: 1.5605... 0.0634 sec/batch
Epoch: 3/20... Training Step: 11226... Training loss: 1.6483... 0.0573 sec/batch
Epoch: 3/20... Training Step: 11227... Training loss: 1.5783... 0.0586 sec/batch
Epoch: 3/20... Training Step: 11228... Training loss: 1.6554... 0.0470 sec/batch
Epoch: 3/20... Training Step: 11229... Training loss: 1.7126... 0.0902 sec/batch
Epoch: 3/20... Training Step: 11230... Training loss: 1.8378... 0.0981 sec/batch
Epoch: 3/20... Training Step: 11231... Training loss: 1.8407... 0.0590 sec/batch
Epoch: 3/20... Training Step: 11232... Training loss: 1.6712... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11233... Training loss: 1.6385... 0.0539 sec/batch
Epoch: 3/20... Training Step: 11234... Training loss: 1.7331... 0.1001 sec/batch
Epoch: 3/20... Training Step: 11235... Training loss: 1.7403... 0.0649 sec/batch
Epoch: 3/20... Training Step: 11236... Training loss: 1.7707... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11237... Training loss: 1.7523... 0.0551 sec/batch
Epoch: 3/20... Training Step: 11238... Training loss: 1.7301... 0.0508 sec/batch
Epoch: 3/20... Training Step: 11239... Training loss: 1.6947... 0.0501 sec/batch
Epoch: 3/20... Training Step: 11240... Training loss: 1.7591... 0.0453 sec/batch
Epoch: 3/20... Training Step: 11241... Training loss: 1.7383... 0.0593 sec/batch
Epoch: 3/20... Training Step: 11242... Training loss: 1.7515... 0.0636 sec/batch
Epoch: 3/20... Training Step: 11243... Training loss: 1.6900... 0.0615 sec/batch
Epoch: 3/20... Training Step: 11244... Training loss: 1.7280... 0.0455 sec/batch
Epoch: 3/20... Training Step: 11245... Training loss: 1.8518... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11246... Training loss: 1.7490... 0.0565 sec/batch
Epoch: 3/20... Training Step: 11247... Training loss: 1.5787... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11248... Training loss: 1.5976... 0.0438 sec/batch
Epoch: 3/20... Training Step: 11249... Training loss: 1.6623... 0.0471 sec/batch
Epoch: 3/20... Training Step: 11250... Training loss: 1.7723... 0.0470 sec/batch
Epoch: 3/20... Training Step: 11251... Training loss: 1.7024... 0.0427 sec/batch
Epoch: 3/20... Training Step: 11252... Training loss: 1.6115... 0.0482 sec/batch
Epoch: 3/20... Training Step: 11253... Training loss: 1.6268... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11254... Training loss: 1.7769... 0.0603 sec/batch
Epoch: 3/20... Training Step: 11255... Training loss: 1.7529... 0.0554 sec/batch
Epoch: 3/20... Training Step: 11256... Training loss: 1.8205... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11257... Training loss: 1.7344... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11258... Training loss: 1.7414... 0.0473 sec/batch
Epoch: 3/20... Training Step: 11259... Training loss: 1.7259... 0.0554 sec/batch
Epoch: 3/20... Training Step: 11260... Training loss: 1.7054... 0.0659 sec/batch
Epoch: 3/20... Training Step: 11261... Training loss: 1.7064... 0.0589 sec/batch
Epoch: 3/20... Training Step: 11262... Training loss: 1.6152... 0.0533 sec/batch
Epoch: 3/20... Training Step: 11263... Training loss: 1.7104... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11264... Training loss: 1.5538... 0.0487 sec/batch
Epoch: 3/20... Training Step: 11265... Training loss: 1.8256... 0.0617 sec/batch
Epoch: 3/20... Training Step: 11266... Training loss: 1.8215... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11267... Training loss: 1.7704... 0.0424 sec/batch
Epoch: 3/20... Training Step: 11268... Training loss: 1.6992... 0.0407 sec/batch
Epoch: 3/20... Training Step: 11269... Training loss: 1.9079... 0.0414 sec/batch
Epoch: 3/20... Training Step: 11270... Training loss: 1.8112... 0.0455 sec/batch
Epoch: 3/20... Training Step: 11271... Training loss: 1.5243... 0.0427 sec/batch
Epoch: 3/20... Training Step: 11272... Training loss: 1.8349... 0.0404 sec/batch
Epoch: 3/20... Training Step: 11273... Training loss: 1.6810... 0.0423 sec/batch
Epoch: 3/20... Training Step: 11274... Training loss: 1.7191... 0.0454 sec/batch
Epoch: 3/20... Training Step: 11275... Training loss: 1.8208... 0.0437 sec/batch
Epoch: 3/20... Training Step: 11276... Training loss: 1.8664... 0.0408 sec/batch
Epoch: 3/20... Training Step: 11277... Training loss: 1.7554... 0.0415 sec/batch
Epoch: 3/20... Training Step: 11278... Training loss: 1.7584... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11279... Training loss: 1.5981... 0.0428 sec/batch
Epoch: 3/20... Training Step: 11280... Training loss: 1.6943... 0.0545 sec/batch
Epoch: 3/20... Training Step: 11281... Training loss: 1.7773... 0.0624 sec/batch
Epoch: 3/20... Training Step: 11282... Training loss: 1.6598... 0.0451 sec/batch
Epoch: 3/20... Training Step: 11283... Training loss: 1.6403... 0.0457 sec/batch
Epoch: 3/20... Training Step: 11284... Training loss: 1.6089... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11285... Training loss: 1.6313... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11286... Training loss: 1.8074... 0.0425 sec/batch
Epoch: 3/20... Training Step: 11287... Training loss: 1.8584... 0.0628 sec/batch
Epoch: 3/20... Training Step: 11288... Training loss: 1.6985... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11289... Training loss: 1.8006... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11290... Training loss: 1.5870... 0.0454 sec/batch
Epoch: 3/20... Training Step: 11291... Training loss: 1.7740... 0.0464 sec/batch
Epoch: 3/20... Training Step: 11292... Training loss: 1.5493... 0.0467 sec/batch
Epoch: 3/20... Training Step: 11293... Training loss: 1.6310... 0.0523 sec/batch
Epoch: 3/20... Training Step: 11294... Training loss: 1.6207... 0.0474 sec/batch
Epoch: 3/20... Training Step: 11295... Training loss: 1.6536... 0.0523 sec/batch
Epoch: 3/20... Training Step: 11296... Training loss: 1.6785... 0.0443 sec/batch
Epoch: 3/20... Training Step: 11297... Training loss: 1.7212... 0.0485 sec/batch
Epoch: 3/20... Training Step: 11298... Training loss: 1.7777... 0.0410 sec/batch
Epoch: 3/20... Training Step: 11299... Training loss: 1.7371... 0.0401 sec/batch
Epoch: 3/20... Training Step: 11300... Training loss: 1.6512... 0.0531 sec/batch
Epoch: 3/20... Training Step: 11301... Training loss: 1.6043... 0.0487 sec/batch
Epoch: 3/20... Training Step: 11302... Training loss: 1.5860... 0.0578 sec/batch
Epoch: 3/20... Training Step: 11303... Training loss: 1.6816... 0.0482 sec/batch
Epoch: 3/20... Training Step: 11304... Training loss: 1.6043... 0.0514 sec/batch
Epoch: 3/20... Training Step: 11305... Training loss: 1.6412... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11306... Training loss: 1.5936... 0.0476 sec/batch
Epoch: 3/20... Training Step: 11307... Training loss: 1.6652... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11308... Training loss: 1.7973... 0.0482 sec/batch
Epoch: 3/20... Training Step: 11309... Training loss: 1.7138... 0.0565 sec/batch
Epoch: 3/20... Training Step: 11310... Training loss: 1.6274... 0.0445 sec/batch
Epoch: 3/20... Training Step: 11311... Training loss: 1.7852... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11312... Training loss: 1.7919... 0.0577 sec/batch
Epoch: 3/20... Training Step: 11313... Training loss: 1.6735... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11314... Training loss: 1.6834... 0.0453 sec/batch
Epoch: 3/20... Training Step: 11315... Training loss: 1.6954... 0.0562 sec/batch
Epoch: 3/20... Training Step: 11316... Training loss: 1.6509... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11317... Training loss: 1.8077... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11318... Training loss: 1.5952... 0.0518 sec/batch
Epoch: 3/20... Training Step: 11319... Training loss: 1.4779... 0.0498 sec/batch
Epoch: 3/20... Training Step: 11320... Training loss: 1.5838... 0.0424 sec/batch
Epoch: 3/20... Training Step: 11321... Training loss: 1.6048... 0.0412 sec/batch
Epoch: 3/20... Training Step: 11322... Training loss: 1.6262... 0.0549 sec/batch
Epoch: 3/20... Training Step: 11323... Training loss: 1.6214... 0.0583 sec/batch
Epoch: 3/20... Training Step: 11324... Training loss: 1.6520... 0.0580 sec/batch
Epoch: 3/20... Training Step: 11325... Training loss: 1.6721... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11326... Training loss: 1.5415... 0.0575 sec/batch
Epoch: 3/20... Training Step: 11327... Training loss: 1.7299... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11328... Training loss: 1.6992... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11329... Training loss: 1.6583... 0.0528 sec/batch
Epoch: 3/20... Training Step: 11330... Training loss: 1.6821... 0.0471 sec/batch
Epoch: 3/20... Training Step: 11331... Training loss: 1.7057... 0.0414 sec/batch
Epoch: 3/20... Training Step: 11332... Training loss: 1.6753... 0.0416 sec/batch
Epoch: 3/20... Training Step: 11333... Training loss: 1.6326... 0.0584 sec/batch
Epoch: 3/20... Training Step: 11334... Training loss: 1.6844... 0.0568 sec/batch
Epoch: 3/20... Training Step: 11335... Training loss: 1.5512... 0.0528 sec/batch
Epoch: 3/20... Training Step: 11336... Training loss: 1.8081... 0.0490 sec/batch
Epoch: 3/20... Training Step: 11337... Training loss: 1.6781... 0.0435 sec/batch
Epoch: 3/20... Training Step: 11338... Training loss: 1.6246... 0.0433 sec/batch
Epoch: 3/20... Training Step: 11339... Training loss: 1.7455... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11340... Training loss: 1.6825... 0.0609 sec/batch
Epoch: 3/20... Training Step: 11341... Training loss: 1.6654... 0.0506 sec/batch
Epoch: 3/20... Training Step: 11342... Training loss: 1.6679... 0.0559 sec/batch
Epoch: 3/20... Training Step: 11343... Training loss: 1.5943... 0.0462 sec/batch
Epoch: 3/20... Training Step: 11344... Training loss: 1.5786... 0.0592 sec/batch
Epoch: 3/20... Training Step: 11345... Training loss: 1.6958... 0.0629 sec/batch
Epoch: 3/20... Training Step: 11346... Training loss: 1.6737... 0.0545 sec/batch
Epoch: 3/20... Training Step: 11347... Training loss: 1.4366... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11348... Training loss: 1.6093... 0.0516 sec/batch
Epoch: 3/20... Training Step: 11349... Training loss: 1.6882... 0.0456 sec/batch
Epoch: 3/20... Training Step: 11350... Training loss: 1.7272... 0.0506 sec/batch
Epoch: 3/20... Training Step: 11351... Training loss: 1.6811... 0.0545 sec/batch
Epoch: 3/20... Training Step: 11352... Training loss: 1.5627... 0.0549 sec/batch
Epoch: 3/20... Training Step: 11353... Training loss: 1.5216... 0.0592 sec/batch
Epoch: 3/20... Training Step: 11354... Training loss: 1.5946... 0.0457 sec/batch
Epoch: 3/20... Training Step: 11355... Training loss: 1.6499... 0.0490 sec/batch
Epoch: 3/20... Training Step: 11356... Training loss: 1.5926... 0.0477 sec/batch
Epoch: 3/20... Training Step: 11357... Training loss: 1.7888... 0.0450 sec/batch
Epoch: 3/20... Training Step: 11358... Training loss: 1.8320... 0.0515 sec/batch
Epoch: 3/20... Training Step: 11359... Training loss: 1.7076... 0.0437 sec/batch
Epoch: 3/20... Training Step: 11360... Training loss: 1.8417... 0.0507 sec/batch
Epoch: 3/20... Training Step: 11361... Training loss: 1.7083... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11362... Training loss: 1.7730... 0.0551 sec/batch
Epoch: 3/20... Training Step: 11363... Training loss: 1.7488... 0.0574 sec/batch
Epoch: 3/20... Training Step: 11364... Training loss: 1.6813... 0.0526 sec/batch
Epoch: 3/20... Training Step: 11365... Training loss: 1.8344... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11366... Training loss: 1.6437... 0.0578 sec/batch
Epoch: 3/20... Training Step: 11367... Training loss: 1.6297... 0.0509 sec/batch
Epoch: 3/20... Training Step: 11368... Training loss: 1.8389... 0.0566 sec/batch
Epoch: 3/20... Training Step: 11369... Training loss: 1.7011... 0.0521 sec/batch
Epoch: 3/20... Training Step: 11370... Training loss: 1.5848... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11371... Training loss: 1.6663... 0.0432 sec/batch
Epoch: 3/20... Training Step: 11372... Training loss: 1.7200... 0.0448 sec/batch
Epoch: 3/20... Training Step: 11373... Training loss: 1.7308... 0.0628 sec/batch
Epoch: 3/20... Training Step: 11374... Training loss: 1.7792... 0.0448 sec/batch
Epoch: 3/20... Training Step: 11375... Training loss: 1.6109... 0.0611 sec/batch
Epoch: 3/20... Training Step: 11376... Training loss: 1.7466... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11377... Training loss: 1.6974... 0.0593 sec/batch
Epoch: 3/20... Training Step: 11378... Training loss: 1.6029... 0.0577 sec/batch
Epoch: 3/20... Training Step: 11379... Training loss: 1.7733... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11380... Training loss: 1.7195... 0.0519 sec/batch
Epoch: 3/20... Training Step: 11381... Training loss: 1.6653... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11382... Training loss: 1.6053... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11383... Training loss: 1.7043... 0.0540 sec/batch
Epoch: 3/20... Training Step: 11384... Training loss: 1.7315... 0.0480 sec/batch
Epoch: 3/20... Training Step: 11385... Training loss: 1.6595... 0.0671 sec/batch
Epoch: 3/20... Training Step: 11386... Training loss: 1.6828... 0.0494 sec/batch
Epoch: 3/20... Training Step: 11387... Training loss: 1.7110... 0.0504 sec/batch
Epoch: 3/20... Training Step: 11388... Training loss: 1.5721... 0.0502 sec/batch
Epoch: 3/20... Training Step: 11389... Training loss: 1.6966... 0.0477 sec/batch
Epoch: 3/20... Training Step: 11390... Training loss: 1.5964... 0.0489 sec/batch
Epoch: 3/20... Training Step: 11391... Training loss: 1.7729... 0.0507 sec/batch
Epoch: 3/20... Training Step: 11392... Training loss: 1.6767... 0.0490 sec/batch
Epoch: 3/20... Training Step: 11393... Training loss: 1.7124... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11394... Training loss: 1.5673... 0.0487 sec/batch
Epoch: 3/20... Training Step: 11395... Training loss: 1.5588... 0.0486 sec/batch
Epoch: 3/20... Training Step: 11396... Training loss: 1.7895... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11397... Training loss: 1.6644... 0.0685 sec/batch
Epoch: 3/20... Training Step: 11398... Training loss: 1.5631... 0.0585 sec/batch
Epoch: 3/20... Training Step: 11399... Training loss: 1.6359... 0.0471 sec/batch
Epoch: 3/20... Training Step: 11400... Training loss: 1.6307... 0.0511 sec/batch
Epoch: 3/20... Training Step: 11401... Training loss: 1.6616... 0.0616 sec/batch
Epoch: 3/20... Training Step: 11402... Training loss: 1.5781... 0.0553 sec/batch
Epoch: 3/20... Training Step: 11403... Training loss: 1.7642... 0.0470 sec/batch
Epoch: 3/20... Training Step: 11404... Training loss: 1.5569... 0.0450 sec/batch
Epoch: 3/20... Training Step: 11405... Training loss: 1.6188... 0.0556 sec/batch
Epoch: 3/20... Training Step: 11406... Training loss: 1.5135... 0.0644 sec/batch
Epoch: 3/20... Training Step: 11407... Training loss: 1.5114... 0.0730 sec/batch
Epoch: 3/20... Training Step: 11408... Training loss: 1.6155... 0.0665 sec/batch
Epoch: 3/20... Training Step: 11409... Training loss: 1.7456... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11410... Training loss: 1.7172... 0.0550 sec/batch
Epoch: 3/20... Training Step: 11411... Training loss: 1.6024... 0.0516 sec/batch
Epoch: 3/20... Training Step: 11412... Training loss: 1.6389... 0.0483 sec/batch
Epoch: 3/20... Training Step: 11413... Training loss: 1.7810... 0.0650 sec/batch
Epoch: 3/20... Training Step: 11414... Training loss: 1.6658... 0.0581 sec/batch
Epoch: 3/20... Training Step: 11415... Training loss: 1.7220... 0.0677 sec/batch
Epoch: 3/20... Training Step: 11416... Training loss: 1.5250... 0.0714 sec/batch
Epoch: 3/20... Training Step: 11417... Training loss: 1.8289... 0.0821 sec/batch
Epoch: 3/20... Training Step: 11418... Training loss: 1.7100... 0.0891 sec/batch
Epoch: 3/20... Training Step: 11419... Training loss: 1.7535... 0.1030 sec/batch
Epoch: 3/20... Training Step: 11420... Training loss: 1.4457... 0.0964 sec/batch
Epoch: 3/20... Training Step: 11421... Training loss: 1.6577... 0.0681 sec/batch
Epoch: 3/20... Training Step: 11422... Training loss: 1.6706... 0.0794 sec/batch
Epoch: 3/20... Training Step: 11423... Training loss: 1.7868... 0.0657 sec/batch
Epoch: 3/20... Training Step: 11424... Training loss: 1.8223... 0.0730 sec/batch
Epoch: 3/20... Training Step: 11425... Training loss: 1.7175... 0.0682 sec/batch
Epoch: 3/20... Training Step: 11426... Training loss: 1.8952... 0.1182 sec/batch
Epoch: 3/20... Training Step: 11427... Training loss: 1.6957... 0.0985 sec/batch
Epoch: 3/20... Training Step: 11428... Training loss: 1.8115... 0.0768 sec/batch
Epoch: 3/20... Training Step: 11429... Training loss: 1.6651... 0.0753 sec/batch
Epoch: 3/20... Training Step: 11430... Training loss: 1.8637... 0.0702 sec/batch
Epoch: 3/20... Training Step: 11431... Training loss: 1.8725... 0.1071 sec/batch
Epoch: 3/20... Training Step: 11432... Training loss: 1.6605... 0.0772 sec/batch
Epoch: 3/20... Training Step: 11433... Training loss: 1.5613... 0.1235 sec/batch
Epoch: 3/20... Training Step: 11434... Training loss: 1.7623... 0.0893 sec/batch
Epoch: 3/20... Training Step: 11435... Training loss: 1.7254... 0.1366 sec/batch
Epoch: 3/20... Training Step: 11436... Training loss: 1.6170... 0.1309 sec/batch
Epoch: 3/20... Training Step: 11437... Training loss: 1.7042... 0.1494 sec/batch
Epoch: 3/20... Training Step: 11438... Training loss: 1.7342... 0.1508 sec/batch
Epoch: 3/20... Training Step: 11439... Training loss: 1.6529... 0.1223 sec/batch
Epoch: 3/20... Training Step: 11440... Training loss: 1.6691... 0.1436 sec/batch
Epoch: 3/20... Training Step: 11441... Training loss: 1.5965... 0.1028 sec/batch
Epoch: 3/20... Training Step: 11442... Training loss: 1.6463... 0.1134 sec/batch
Epoch: 3/20... Training Step: 11443... Training loss: 1.6507... 0.1183 sec/batch
Epoch: 3/20... Training Step: 11444... Training loss: 1.6524... 0.0923 sec/batch
Epoch: 3/20... Training Step: 11445... Training loss: 1.6983... 0.0672 sec/batch
Epoch: 3/20... Training Step: 11446... Training loss: 1.6777... 0.0746 sec/batch
Epoch: 3/20... Training Step: 11447... Training loss: 1.6373... 0.0709 sec/batch
Epoch: 3/20... Training Step: 11448... Training loss: 1.6488... 0.0927 sec/batch
Epoch: 3/20... Training Step: 11449... Training loss: 1.6118... 0.0859 sec/batch
Epoch: 3/20... Training Step: 11450... Training loss: 1.6678... 0.1585 sec/batch
Epoch: 3/20... Training Step: 11451... Training loss: 1.7506... 0.1233 sec/batch
Epoch: 3/20... Training Step: 11452... Training loss: 1.5926... 0.1333 sec/batch
Epoch: 3/20... Training Step: 11453... Training loss: 1.6765... 0.0711 sec/batch
Epoch: 3/20... Training Step: 11454... Training loss: 1.5163... 0.0834 sec/batch
Epoch: 3/20... Training Step: 11455... Training loss: 1.5258... 0.0881 sec/batch
Epoch: 3/20... Training Step: 11456... Training loss: 1.6491... 0.0808 sec/batch
Epoch: 3/20... Training Step: 11457... Training loss: 1.6237... 0.0746 sec/batch
Epoch: 3/20... Training Step: 11458... Training loss: 1.6894... 0.0970 sec/batch
Epoch: 3/20... Training Step: 11459... Training loss: 1.8160... 0.0562 sec/batch
Epoch: 3/20... Training Step: 11460... Training loss: 1.6499... 0.0654 sec/batch
Epoch: 3/20... Training Step: 11461... Training loss: 1.7126... 0.0510 sec/batch
Epoch: 3/20... Training Step: 11462... Training loss: 1.7905... 0.0671 sec/batch
Epoch: 3/20... Training Step: 11463... Training loss: 1.6720... 0.0690 sec/batch
Epoch: 3/20... Training Step: 11464... Training loss: 1.7049... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11465... Training loss: 1.4417... 0.0744 sec/batch
Epoch: 3/20... Training Step: 11466... Training loss: 1.7790... 0.0777 sec/batch
Epoch: 3/20... Training Step: 11467... Training loss: 1.6574... 0.0788 sec/batch
Epoch: 3/20... Training Step: 11468... Training loss: 1.7038... 0.0806 sec/batch
Epoch: 3/20... Training Step: 11469... Training loss: 1.5514... 0.0612 sec/batch
Epoch: 3/20... Training Step: 11470... Training loss: 1.5615... 0.0574 sec/batch
Epoch: 3/20... Training Step: 11471... Training loss: 1.5317... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11472... Training loss: 1.6874... 0.0595 sec/batch
Epoch: 3/20... Training Step: 11473... Training loss: 1.7001... 0.0550 sec/batch
Epoch: 3/20... Training Step: 11474... Training loss: 1.7184... 0.0568 sec/batch
Epoch: 3/20... Training Step: 11475... Training loss: 1.7955... 0.0517 sec/batch
Epoch: 3/20... Training Step: 11476... Training loss: 1.6357... 0.0605 sec/batch
Epoch: 3/20... Training Step: 11477... Training loss: 1.6317... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11478... Training loss: 1.7377... 0.0608 sec/batch
Epoch: 3/20... Training Step: 11479... Training loss: 1.7772... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11480... Training loss: 1.6772... 0.0547 sec/batch
Epoch: 3/20... Training Step: 11481... Training loss: 1.7744... 0.0603 sec/batch
Epoch: 3/20... Training Step: 11482... Training loss: 1.6842... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11483... Training loss: 1.5201... 0.0670 sec/batch
Epoch: 3/20... Training Step: 11484... Training loss: 1.6064... 0.0483 sec/batch
Epoch: 3/20... Training Step: 11485... Training loss: 1.6545... 0.0569 sec/batch
Epoch: 3/20... Training Step: 11486... Training loss: 1.5350... 0.0703 sec/batch
Epoch: 3/20... Training Step: 11487... Training loss: 1.6622... 0.0871 sec/batch
Epoch: 3/20... Training Step: 11488... Training loss: 1.5301... 0.0783 sec/batch
Epoch: 3/20... Training Step: 11489... Training loss: 1.6162... 0.0769 sec/batch
Epoch: 3/20... Training Step: 11490... Training loss: 1.8225... 0.0701 sec/batch
Epoch: 3/20... Training Step: 11491... Training loss: 1.6685... 0.0644 sec/batch
Epoch: 3/20... Training Step: 11492... Training loss: 1.6667... 0.0643 sec/batch
Epoch: 3/20... Training Step: 11493... Training loss: 1.7550... 0.0532 sec/batch
Epoch: 3/20... Training Step: 11494... Training loss: 1.7744... 0.0662 sec/batch
Epoch: 3/20... Training Step: 11495... Training loss: 1.5215... 0.0592 sec/batch
Epoch: 3/20... Training Step: 11496... Training loss: 1.7925... 0.0554 sec/batch
Epoch: 3/20... Training Step: 11497... Training loss: 1.5987... 0.0579 sec/batch
Epoch: 3/20... Training Step: 11498... Training loss: 1.4871... 0.0676 sec/batch
Epoch: 3/20... Training Step: 11499... Training loss: 1.7761... 0.0564 sec/batch
Epoch: 3/20... Training Step: 11500... Training loss: 1.6807... 0.0554 sec/batch
Epoch: 3/20... Training Step: 11501... Training loss: 1.6158... 0.0533 sec/batch
Epoch: 3/20... Training Step: 11502... Training loss: 1.6204... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11503... Training loss: 1.6856... 0.0502 sec/batch
Epoch: 3/20... Training Step: 11504... Training loss: 1.5965... 0.0660 sec/batch
Epoch: 3/20... Training Step: 11505... Training loss: 1.7331... 0.0715 sec/batch
Epoch: 3/20... Training Step: 11506... Training loss: 1.7170... 0.1034 sec/batch
Epoch: 3/20... Training Step: 11507... Training loss: 1.6987... 0.1172 sec/batch
Epoch: 3/20... Training Step: 11508... Training loss: 1.6662... 0.0761 sec/batch
Epoch: 3/20... Training Step: 11509... Training loss: 1.6861... 0.0985 sec/batch
Epoch: 3/20... Training Step: 11510... Training loss: 1.4763... 0.0793 sec/batch
Epoch: 3/20... Training Step: 11511... Training loss: 1.6742... 0.1004 sec/batch
Epoch: 3/20... Training Step: 11512... Training loss: 1.8277... 0.0945 sec/batch
Epoch: 3/20... Training Step: 11513... Training loss: 1.6568... 0.0610 sec/batch
Epoch: 3/20... Training Step: 11514... Training loss: 1.5824... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11515... Training loss: 1.5394... 0.0437 sec/batch
Epoch: 3/20... Training Step: 11516... Training loss: 1.5689... 0.0606 sec/batch
Epoch: 3/20... Training Step: 11517... Training loss: 1.6465... 0.1065 sec/batch
Epoch: 3/20... Training Step: 11518... Training loss: 1.5571... 0.0917 sec/batch
Epoch: 3/20... Training Step: 11519... Training loss: 1.6018... 0.0906 sec/batch
Epoch: 3/20... Training Step: 11520... Training loss: 1.5650... 0.0771 sec/batch
Epoch: 3/20... Training Step: 11521... Training loss: 1.6314... 0.0856 sec/batch
Epoch: 3/20... Training Step: 11522... Training loss: 1.6816... 0.0877 sec/batch
Epoch: 3/20... Training Step: 11523... Training loss: 1.5672... 0.0689 sec/batch
Epoch: 3/20... Training Step: 11524... Training loss: 1.7542... 0.0733 sec/batch
Epoch: 3/20... Training Step: 11525... Training loss: 1.6390... 0.1078 sec/batch
Epoch: 3/20... Training Step: 11526... Training loss: 1.7666... 0.1017 sec/batch
Epoch: 3/20... Training Step: 11527... Training loss: 1.7631... 0.0857 sec/batch
Epoch: 3/20... Training Step: 11528... Training loss: 1.6116... 0.0880 sec/batch
Epoch: 3/20... Training Step: 11529... Training loss: 1.5774... 0.0902 sec/batch
Epoch: 3/20... Training Step: 11530... Training loss: 1.6766... 0.1122 sec/batch
Epoch: 3/20... Training Step: 11531... Training loss: 1.6151... 0.0954 sec/batch
Epoch: 3/20... Training Step: 11532... Training loss: 1.5866... 0.0885 sec/batch
Epoch: 3/20... Training Step: 11533... Training loss: 1.6692... 0.0810 sec/batch
Epoch: 3/20... Training Step: 11534... Training loss: 2.1704... 0.0491 sec/batch
Epoch: 3/20... Training Step: 11535... Training loss: 1.8014... 0.0526 sec/batch
Epoch: 3/20... Training Step: 11536... Training loss: 1.7216... 0.0569 sec/batch
Epoch: 3/20... Training Step: 11537... Training loss: 1.8126... 0.0527 sec/batch
Epoch: 3/20... Training Step: 11538... Training loss: 1.7530... 0.0515 sec/batch
Epoch: 3/20... Training Step: 11539... Training loss: 1.7729... 0.0530 sec/batch
Epoch: 3/20... Training Step: 11540... Training loss: 1.7261... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11541... Training loss: 1.7882... 0.0779 sec/batch
Epoch: 3/20... Training Step: 11542... Training loss: 1.6707... 0.1161 sec/batch
Epoch: 3/20... Training Step: 11543... Training loss: 1.7620... 0.0918 sec/batch
Epoch: 3/20... Training Step: 11544... Training loss: 1.8247... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11545... Training loss: 1.6946... 0.0573 sec/batch
Epoch: 3/20... Training Step: 11546... Training loss: 1.6479... 0.0583 sec/batch
Epoch: 3/20... Training Step: 11547... Training loss: 1.7024... 0.0529 sec/batch
Epoch: 3/20... Training Step: 11548... Training loss: 1.7148... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11549... Training loss: 1.7996... 0.0516 sec/batch
Epoch: 3/20... Training Step: 11550... Training loss: 1.7642... 0.0555 sec/batch
Epoch: 3/20... Training Step: 11551... Training loss: 1.7249... 0.0454 sec/batch
Epoch: 3/20... Training Step: 11552... Training loss: 1.6662... 0.0528 sec/batch
Epoch: 3/20... Training Step: 11553... Training loss: 1.7046... 0.0640 sec/batch
Epoch: 3/20... Training Step: 11554... Training loss: 1.6395... 0.0486 sec/batch
Epoch: 3/20... Training Step: 11555... Training loss: 1.7190... 0.0468 sec/batch
Epoch: 3/20... Training Step: 11556... Training loss: 1.6886... 0.0440 sec/batch
Epoch: 3/20... Training Step: 11557... Training loss: 1.7332... 0.0432 sec/batch
Epoch: 3/20... Training Step: 11558... Training loss: 1.8030... 0.0435 sec/batch
Epoch: 3/20... Training Step: 11559... Training loss: 1.7100... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11560... Training loss: 1.6605... 0.0793 sec/batch
Epoch: 3/20... Training Step: 11561... Training loss: 1.6053... 0.0783 sec/batch
Epoch: 3/20... Training Step: 11562... Training loss: 1.7930... 0.0838 sec/batch
Epoch: 3/20... Training Step: 11563... Training loss: 1.6859... 0.0702 sec/batch
Epoch: 3/20... Training Step: 11564... Training loss: 1.8368... 0.0844 sec/batch
Epoch: 3/20... Training Step: 11565... Training loss: 1.6435... 0.0582 sec/batch
Epoch: 3/20... Training Step: 11566... Training loss: 1.7016... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11567... Training loss: 1.9359... 0.0557 sec/batch
Epoch: 3/20... Training Step: 11568... Training loss: 1.7131... 0.0483 sec/batch
Epoch: 3/20... Training Step: 11569... Training loss: 1.7616... 0.0472 sec/batch
Epoch: 3/20... Training Step: 11570... Training loss: 1.6525... 0.0650 sec/batch
Epoch: 3/20... Training Step: 11571... Training loss: 1.7200... 0.0879 sec/batch
Epoch: 3/20... Training Step: 11572... Training loss: 1.8058... 0.0887 sec/batch
Epoch: 3/20... Training Step: 11573... Training loss: 1.7725... 0.0999 sec/batch
Epoch: 3/20... Training Step: 11574... Training loss: 1.7441... 0.0606 sec/batch
Epoch: 3/20... Training Step: 11575... Training loss: 1.7004... 0.0572 sec/batch
Epoch: 3/20... Training Step: 11576... Training loss: 1.8352... 0.0606 sec/batch
Epoch: 3/20... Training Step: 11577... Training loss: 1.6692... 0.0895 sec/batch
Epoch: 3/20... Training Step: 11578... Training loss: 1.7750... 0.1535 sec/batch
Epoch: 3/20... Training Step: 11579... Training loss: 1.7684... 0.0519 sec/batch
Epoch: 3/20... Training Step: 11580... Training loss: 1.6160... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11581... Training loss: 1.7998... 0.0513 sec/batch
Epoch: 3/20... Training Step: 11582... Training loss: 1.7591... 0.0597 sec/batch
Epoch: 3/20... Training Step: 11583... Training loss: 1.8574... 0.0526 sec/batch
Epoch: 3/20... Training Step: 11584... Training loss: 1.8293... 0.0604 sec/batch
Epoch: 3/20... Training Step: 11585... Training loss: 1.7421... 0.0526 sec/batch
Epoch: 3/20... Training Step: 11586... Training loss: 1.5690... 0.0529 sec/batch
Epoch: 3/20... Training Step: 11587... Training loss: 1.7143... 0.0624 sec/batch
Epoch: 3/20... Training Step: 11588... Training loss: 1.7711... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11589... Training loss: 1.6242... 0.0787 sec/batch
Epoch: 3/20... Training Step: 11590... Training loss: 1.8532... 0.0902 sec/batch
Epoch: 3/20... Training Step: 11591... Training loss: 1.7809... 0.0849 sec/batch
Epoch: 3/20... Training Step: 11592... Training loss: 1.6217... 0.0967 sec/batch
Epoch: 3/20... Training Step: 11593... Training loss: 1.7427... 0.0974 sec/batch
Epoch: 3/20... Training Step: 11594... Training loss: 1.6830... 0.0547 sec/batch
Epoch: 3/20... Training Step: 11595... Training loss: 1.9248... 0.0525 sec/batch
Epoch: 3/20... Training Step: 11596... Training loss: 1.9185... 0.0634 sec/batch
Epoch: 3/20... Training Step: 11597... Training loss: 1.7209... 0.0480 sec/batch
Epoch: 3/20... Training Step: 11598... Training loss: 1.7910... 0.0490 sec/batch
Epoch: 3/20... Training Step: 11599... Training loss: 1.6773... 0.0660 sec/batch
Epoch: 3/20... Training Step: 11600... Training loss: 1.7937... 0.0562 sec/batch
Epoch: 3/20... Training Step: 11601... Training loss: 1.7692... 0.0583 sec/batch
Epoch: 3/20... Training Step: 11602... Training loss: 1.6648... 0.0710 sec/batch
Epoch: 3/20... Training Step: 11603... Training loss: 1.7800... 0.0538 sec/batch
Epoch: 3/20... Training Step: 11604... Training loss: 1.8809... 0.0550 sec/batch
Epoch: 3/20... Training Step: 11605... Training loss: 1.6480... 0.0863 sec/batch
Epoch: 3/20... Training Step: 11606... Training loss: 1.8832... 0.1064 sec/batch
Epoch: 3/20... Training Step: 11607... Training loss: 1.5674... 0.1008 sec/batch
Epoch: 3/20... Training Step: 11608... Training loss: 1.8472... 0.0803 sec/batch
Epoch: 3/20... Training Step: 11609... Training loss: 1.7753... 0.0561 sec/batch
Epoch: 3/20... Training Step: 11610... Training loss: 1.7301... 0.0989 sec/batch
Epoch: 3/20... Training Step: 11611... Training loss: 1.6411... 0.1064 sec/batch
Epoch: 3/20... Training Step: 11612... Training loss: 1.7540... 0.0893 sec/batch
Epoch: 3/20... Training Step: 11613... Training loss: 1.7074... 0.0765 sec/batch
Epoch: 3/20... Training Step: 11614... Training loss: 1.7663... 0.0716 sec/batch
Epoch: 3/20... Training Step: 11615... Training loss: 1.6253... 0.0977 sec/batch
Epoch: 3/20... Training Step: 11616... Training loss: 1.9386... 0.1014 sec/batch
Epoch: 3/20... Training Step: 11617... Training loss: 1.7652... 0.0538 sec/batch
Epoch: 3/20... Training Step: 11618... Training loss: 1.7079... 0.0485 sec/batch
Epoch: 3/20... Training Step: 11619... Training loss: 1.7791... 0.0553 sec/batch
Epoch: 3/20... Training Step: 11620... Training loss: 1.8440... 0.0531 sec/batch
Epoch: 3/20... Training Step: 11621... Training loss: 1.7078... 0.0565 sec/batch
Epoch: 3/20... Training Step: 11622... Training loss: 1.7914... 0.1046 sec/batch
Epoch: 3/20... Training Step: 11623... Training loss: 1.6717... 0.0983 sec/batch
Epoch: 3/20... Training Step: 11624... Training loss: 1.6453... 0.0881 sec/batch
Epoch: 3/20... Training Step: 11625... Training loss: 1.7198... 0.0833 sec/batch
Epoch: 3/20... Training Step: 11626... Training loss: 1.8362... 0.0577 sec/batch
Epoch: 3/20... Training Step: 11627... Training loss: 1.7637... 0.0471 sec/batch
Epoch: 3/20... Training Step: 11628... Training loss: 1.6359... 0.0596 sec/batch
Epoch: 3/20... Training Step: 11629... Training loss: 1.6526... 0.0527 sec/batch
Epoch: 3/20... Training Step: 11630... Training loss: 1.6442... 0.0488 sec/batch
Epoch: 3/20... Training Step: 11631... Training loss: 1.8134... 0.0587 sec/batch
Epoch: 3/20... Training Step: 11632... Training loss: 1.7183... 0.0580 sec/batch
Epoch: 3/20... Training Step: 11633... Training loss: 1.8519... 0.0619 sec/batch
Epoch: 3/20... Training Step: 11634... Training loss: 1.7743... 0.0480 sec/batch
Epoch: 3/20... Training Step: 11635... Training loss: 1.7791... 0.0561 sec/batch
Epoch: 3/20... Training Step: 11636... Training loss: 1.8166... 0.0563 sec/batch
Epoch: 3/20... Training Step: 11637... Training loss: 1.6719... 0.0590 sec/batch
Epoch: 3/20... Training Step: 11638... Training loss: 1.7767... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11639... Training loss: 1.6754... 0.0601 sec/batch
Epoch: 3/20... Training Step: 11640... Training loss: 1.6378... 0.0544 sec/batch
Epoch: 3/20... Training Step: 11641... Training loss: 1.8496... 0.0485 sec/batch
Epoch: 3/20... Training Step: 11642... Training loss: 1.7475... 0.0554 sec/batch
Epoch: 3/20... Training Step: 11643... Training loss: 1.6352... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11644... Training loss: 1.8642... 0.0561 sec/batch
Epoch: 3/20... Training Step: 11645... Training loss: 1.6254... 0.0474 sec/batch
Epoch: 3/20... Training Step: 11646... Training loss: 1.5945... 0.0599 sec/batch
Epoch: 3/20... Training Step: 11647... Training loss: 1.6386... 0.0955 sec/batch
Epoch: 3/20... Training Step: 11648... Training loss: 1.7071... 0.0927 sec/batch
Epoch: 3/20... Training Step: 11649... Training loss: 1.7312... 0.0943 sec/batch
Epoch: 3/20... Training Step: 11650... Training loss: 1.7638... 0.0853 sec/batch
Epoch: 3/20... Training Step: 11651... Training loss: 1.7470... 0.0926 sec/batch
Epoch: 3/20... Training Step: 11652... Training loss: 1.8321... 0.0956 sec/batch
Epoch: 3/20... Training Step: 11653... Training loss: 1.6775... 0.0840 sec/batch
Epoch: 3/20... Training Step: 11654... Training loss: 1.8285... 0.0793 sec/batch
Epoch: 3/20... Training Step: 11655... Training loss: 1.9309... 0.0679 sec/batch
Epoch: 3/20... Training Step: 11656... Training loss: 1.6447... 0.0503 sec/batch
Epoch: 3/20... Training Step: 11657... Training loss: 1.6949... 0.0550 sec/batch
Epoch: 3/20... Training Step: 11658... Training loss: 1.6487... 0.0676 sec/batch
Epoch: 3/20... Training Step: 11659... Training loss: 1.6805... 0.0855 sec/batch
Epoch: 3/20... Training Step: 11660... Training loss: 1.6397... 0.0998 sec/batch
Epoch: 3/20... Training Step: 11661... Training loss: 1.6132... 0.0979 sec/batch
Epoch: 3/20... Training Step: 11662... Training loss: 1.5561... 0.0578 sec/batch
Epoch: 3/20... Training Step: 11663... Training loss: 1.8655... 0.0622 sec/batch
Epoch: 3/20... Training Step: 11664... Training loss: 1.8445... 0.1142 sec/batch
Epoch: 3/20... Training Step: 11665... Training loss: 1.7496... 0.1176 sec/batch
Epoch: 3/20... Training Step: 11666... Training loss: 1.8042... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11667... Training loss: 1.8306... 0.0581 sec/batch
Epoch: 3/20... Training Step: 11668... Training loss: 1.8407... 0.0488 sec/batch
Epoch: 3/20... Training Step: 11669... Training loss: 1.9344... 0.0582 sec/batch
Epoch: 3/20... Training Step: 11670... Training loss: 1.8590... 0.0600 sec/batch
Epoch: 3/20... Training Step: 11671... Training loss: 1.6963... 0.0452 sec/batch
Epoch: 3/20... Training Step: 11672... Training loss: 1.8315... 0.0645 sec/batch
Epoch: 3/20... Training Step: 11673... Training loss: 1.6881... 0.0813 sec/batch
Epoch: 3/20... Training Step: 11674... Training loss: 1.6354... 0.0662 sec/batch
Epoch: 3/20... Training Step: 11675... Training loss: 1.8080... 0.0888 sec/batch
Epoch: 3/20... Training Step: 11676... Training loss: 1.7304... 0.0495 sec/batch
Epoch: 3/20... Training Step: 11677... Training loss: 1.6693... 0.0753 sec/batch
Epoch: 3/20... Training Step: 11678... Training loss: 1.9258... 0.0720 sec/batch
Epoch: 3/20... Training Step: 11679... Training loss: 1.9101... 0.0857 sec/batch
Epoch: 3/20... Training Step: 11680... Training loss: 1.7065... 0.0726 sec/batch
Epoch: 3/20... Training Step: 11681... Training loss: 1.7424... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11682... Training loss: 1.8894... 0.0714 sec/batch
Epoch: 3/20... Training Step: 11683... Training loss: 1.8021... 0.0958 sec/batch
Epoch: 3/20... Training Step: 11684... Training loss: 1.8614... 0.0879 sec/batch
Epoch: 3/20... Training Step: 11685... Training loss: 1.6309... 0.0803 sec/batch
Epoch: 3/20... Training Step: 11686... Training loss: 1.9230... 0.0789 sec/batch
Epoch: 3/20... Training Step: 11687... Training loss: 1.7544... 0.0917 sec/batch
Epoch: 3/20... Training Step: 11688... Training loss: 2.0493... 0.1017 sec/batch
Epoch: 3/20... Training Step: 11689... Training loss: 1.6734... 0.0919 sec/batch
Epoch: 3/20... Training Step: 11690... Training loss: 1.7897... 0.0877 sec/batch
Epoch: 3/20... Training Step: 11691... Training loss: 1.6643... 0.0797 sec/batch
Epoch: 3/20... Training Step: 11692... Training loss: 1.9617... 0.0680 sec/batch
Epoch: 3/20... Training Step: 11693... Training loss: 1.6935... 0.0987 sec/batch
Epoch: 3/20... Training Step: 11694... Training loss: 1.7278... 0.0914 sec/batch
Epoch: 3/20... Training Step: 11695... Training loss: 1.7488... 0.0959 sec/batch
Epoch: 3/20... Training Step: 11696... Training loss: 1.7572... 0.0758 sec/batch
Epoch: 3/20... Training Step: 11697... Training loss: 1.7270... 0.0589 sec/batch
Epoch: 3/20... Training Step: 11698... Training loss: 1.8933... 0.0760 sec/batch
Epoch: 3/20... Training Step: 11699... Training loss: 1.7392... 0.1062 sec/batch
Epoch: 3/20... Training Step: 11700... Training loss: 1.6598... 0.0858 sec/batch
Epoch: 3/20... Training Step: 11701... Training loss: 1.7614... 0.0801 sec/batch
Epoch: 3/20... Training Step: 11702... Training loss: 1.6233... 0.0837 sec/batch
Epoch: 3/20... Training Step: 11703... Training loss: 1.6883... 0.0530 sec/batch
Epoch: 3/20... Training Step: 11704... Training loss: 1.7836... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11705... Training loss: 1.7095... 0.0571 sec/batch
Epoch: 3/20... Training Step: 11706... Training loss: 1.7507... 0.0776 sec/batch
Epoch: 3/20... Training Step: 11707... Training loss: 1.6164... 0.1026 sec/batch
Epoch: 3/20... Training Step: 11708... Training loss: 1.7688... 0.1297 sec/batch
Epoch: 3/20... Training Step: 11709... Training loss: 1.8486... 0.0764 sec/batch
Epoch: 3/20... Training Step: 11710... Training loss: 1.6630... 0.0613 sec/batch
Epoch: 3/20... Training Step: 11711... Training loss: 1.6293... 0.0558 sec/batch
Epoch: 3/20... Training Step: 11712... Training loss: 1.6409... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11713... Training loss: 1.7821... 0.0634 sec/batch
Epoch: 3/20... Training Step: 11714... Training loss: 1.7256... 0.0574 sec/batch
Epoch: 3/20... Training Step: 11715... Training loss: 1.7699... 0.0693 sec/batch
Epoch: 3/20... Training Step: 11716... Training loss: 1.7548... 0.0490 sec/batch
Epoch: 3/20... Training Step: 11717... Training loss: 1.7173... 0.0647 sec/batch
Epoch: 3/20... Training Step: 11718... Training loss: 1.7095... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11719... Training loss: 1.7007... 0.0595 sec/batch
Epoch: 3/20... Training Step: 11720... Training loss: 1.9291... 0.0595 sec/batch
Epoch: 3/20... Training Step: 11721... Training loss: 1.6795... 0.0668 sec/batch
Epoch: 3/20... Training Step: 11722... Training loss: 1.6115... 0.0655 sec/batch
Epoch: 3/20... Training Step: 11723... Training loss: 1.6638... 0.0573 sec/batch
Epoch: 3/20... Training Step: 11724... Training loss: 1.6873... 0.0674 sec/batch
Epoch: 3/20... Training Step: 11725... Training loss: 1.6062... 0.0509 sec/batch
Epoch: 3/20... Training Step: 11726... Training loss: 1.5735... 0.0607 sec/batch
Epoch: 3/20... Training Step: 11727... Training loss: 1.7120... 0.0564 sec/batch
Epoch: 3/20... Training Step: 11728... Training loss: 1.6627... 0.0635 sec/batch
Epoch: 3/20... Training Step: 11729... Training loss: 1.5529... 0.0566 sec/batch
Epoch: 3/20... Training Step: 11730... Training loss: 1.6406... 0.0510 sec/batch
Epoch: 3/20... Training Step: 11731... Training loss: 1.5453... 0.0553 sec/batch
Epoch: 3/20... Training Step: 11732... Training loss: 1.7470... 0.0553 sec/batch
Epoch: 3/20... Training Step: 11733... Training loss: 1.7890... 0.0511 sec/batch
Epoch: 3/20... Training Step: 11734... Training loss: 1.9194... 0.0461 sec/batch
Epoch: 3/20... Training Step: 11735... Training loss: 1.8889... 0.0532 sec/batch
Epoch: 3/20... Training Step: 11736... Training loss: 1.7980... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11737... Training loss: 1.8750... 0.0459 sec/batch
Epoch: 3/20... Training Step: 11738... Training loss: 1.9760... 0.0565 sec/batch
Epoch: 3/20... Training Step: 11739... Training loss: 1.7172... 0.0486 sec/batch
Epoch: 3/20... Training Step: 11740... Training loss: 1.6932... 0.0537 sec/batch
Epoch: 3/20... Training Step: 11741... Training loss: 1.8206... 0.0516 sec/batch
Epoch: 3/20... Training Step: 11742... Training loss: 1.5887... 0.0582 sec/batch
Epoch: 3/20... Training Step: 11743... Training loss: 1.6594... 0.0606 sec/batch
Epoch: 3/20... Training Step: 11744... Training loss: 1.6399... 0.0545 sec/batch
Epoch: 3/20... Training Step: 11745... Training loss: 1.7693... 0.0564 sec/batch
Epoch: 3/20... Training Step: 11746... Training loss: 1.5839... 0.0480 sec/batch
Epoch: 3/20... Training Step: 11747... Training loss: 1.7997... 0.0940 sec/batch
Epoch: 3/20... Training Step: 11748... Training loss: 2.1081... 0.0992 sec/batch
Epoch: 3/20... Training Step: 11749... Training loss: 1.7367... 0.0920 sec/batch
Epoch: 3/20... Training Step: 11750... Training loss: 1.7733... 0.0801 sec/batch
Epoch: 3/20... Training Step: 11751... Training loss: 1.5833... 0.0815 sec/batch
Epoch: 3/20... Training Step: 11752... Training loss: 1.8353... 0.0886 sec/batch
Epoch: 3/20... Training Step: 11753... Training loss: 1.5193... 0.0878 sec/batch
Epoch: 3/20... Training Step: 11754... Training loss: 1.5991... 0.0646 sec/batch
Epoch: 3/20... Training Step: 11755... Training loss: 1.8600... 0.0667 sec/batch
Epoch: 3/20... Training Step: 11756... Training loss: 2.0785... 0.0674 sec/batch
Epoch: 3/20... Training Step: 11757... Training loss: 2.0869... 0.0682 sec/batch
Epoch: 3/20... Training Step: 11758... Training loss: 2.0109... 0.0627 sec/batch
Epoch: 3/20... Training Step: 11759... Training loss: 1.9927... 0.0594 sec/batch
Epoch: 3/20... Training Step: 11760... Training loss: 2.0846... 0.0763 sec/batch
Epoch: 3/20... Training Step: 11761... Training loss: 1.9132... 0.0869 sec/batch
Epoch: 3/20... Training Step: 11762... Training loss: 1.8367... 0.0743 sec/batch
Epoch: 3/20... Training Step: 11763... Training loss: 1.9104... 0.0751 sec/batch
Epoch: 3/20... Training Step: 11764... Training loss: 1.8138... 0.0931 sec/batch
Epoch: 3/20... Training Step: 11765... Training loss: 1.6896... 0.0897 sec/batch
Epoch: 3/20... Training Step: 11766... Training loss: 1.7022... 0.0719 sec/batch
Epoch: 3/20... Training Step: 11767... Training loss: 1.6742... 0.1098 sec/batch
Epoch: 3/20... Training Step: 11768... Training loss: 1.7439... 0.0765 sec/batch
Epoch: 3/20... Training Step: 11769... Training loss: 1.6258... 0.0897 sec/batch
Epoch: 3/20... Training Step: 11770... Training loss: 1.7839... 0.0722 sec/batch
Epoch: 3/20... Training Step: 11771... Training loss: 1.6588... 0.0638 sec/batch
Epoch: 3/20... Training Step: 11772... Training loss: 1.5564... 0.0699 sec/batch
Epoch: 3/20... Training Step: 11773... Training loss: 1.6643... 0.0854 sec/batch
Epoch: 3/20... Training Step: 11774... Training loss: 1.6019... 0.0724 sec/batch
Epoch: 3/20... Training Step: 11775... Training loss: 1.8073... 0.0972 sec/batch
Epoch: 3/20... Training Step: 11776... Training loss: 1.6557... 0.1088 sec/batch
Epoch: 3/20... Training Step: 11777... Training loss: 1.8082... 0.0762 sec/batch
Epoch: 3/20... Training Step: 11778... Training loss: 1.7748... 0.0949 sec/batch
Epoch: 3/20... Training Step: 11779... Training loss: 1.8020... 0.0960 sec/batch
Epoch: 3/20... Training Step: 11780... Training loss: 1.8912... 0.0711 sec/batch
Epoch: 3/20... Training Step: 11781... Training loss: 1.7689... 0.1005 sec/batch
Epoch: 3/20... Training Step: 11782... Training loss: 1.8799... 0.0650 sec/batch
Epoch: 3/20... Training Step: 11783... Training loss: 1.9611... 0.0745 sec/batch
Epoch: 3/20... Training Step: 11784... Training loss: 1.8462... 0.0931 sec/batch
Epoch: 3/20... Training Step: 11785... Training loss: 1.7812... 0.0780 sec/batch
Epoch: 3/20... Training Step: 11786... Training loss: 1.7408... 0.0685 sec/batch
Epoch: 3/20... Training Step: 11787... Training loss: 1.6387... 0.0662 sec/batch
Epoch: 3/20... Training Step: 11788... Training loss: 1.8218... 0.0520 sec/batch
Epoch: 3/20... Training Step: 11789... Training loss: 1.6233... 0.0719 sec/batch
Epoch: 3/20... Training Step: 11790... Training loss: 1.6272... 0.0610 sec/batch
Epoch: 3/20... Training Step: 11791... Training loss: 1.9303... 0.0859 sec/batch
Epoch: 3/20... Training Step: 11792... Training loss: 1.8460... 0.0788 sec/batch
Epoch: 3/20... Training Step: 11793... Training loss: 1.8941... 0.0705 sec/batch
Epoch: 3/20... Training Step: 11794... Training loss: 1.6414... 0.0977 sec/batch
Epoch: 3/20... Training Step: 11795... Training loss: 1.8570... 0.0949 sec/batch
Epoch: 3/20... Training Step: 11796... Training loss: 1.6202... 0.0977 sec/batch
Epoch: 3/20... Training Step: 11797... Training loss: 1.8228... 0.1098 sec/batch
Epoch: 3/20... Training Step: 11798... Training loss: 1.6251... 0.0965 sec/batch
Epoch: 3/20... Training Step: 11799... Training loss: 1.8085... 0.0935 sec/batch
Epoch: 3/20... Training Step: 11800... Training loss: 1.7694... 0.0824 sec/batch
Epoch: 3/20... Training Step: 11801... Training loss: 1.7993... 0.1066 sec/batch
Epoch: 3/20... Training Step: 11802... Training loss: 1.6349... 0.1020 sec/batch
Epoch: 3/20... Training Step: 11803... Training loss: 1.6939... 0.1278 sec/batch
Epoch: 3/20... Training Step: 11804... Training loss: 1.7064... 0.0797 sec/batch
Epoch: 3/20... Training Step: 11805... Training loss: 1.8005... 0.0967 sec/batch
Epoch: 3/20... Training Step: 11806... Training loss: 1.6867... 0.0833 sec/batch
Epoch: 3/20... Training Step: 11807... Training loss: 1.7426... 0.0935 sec/batch
Epoch: 3/20... Training Step: 11808... Training loss: 1.6405... 0.0728 sec/batch
Epoch: 3/20... Training Step: 11809... Training loss: 1.7674... 0.0892 sec/batch
Epoch: 3/20... Training Step: 11810... Training loss: 1.8473... 0.0787 sec/batch
Epoch: 3/20... Training Step: 11811... Training loss: 1.7387... 0.0797 sec/batch
Epoch: 3/20... Training Step: 11812... Training loss: 1.7766... 0.0783 sec/batch
Epoch: 3/20... Training Step: 11813... Training loss: 1.7530... 0.0719 sec/batch
Epoch: 3/20... Training Step: 11814... Training loss: 1.6624... 0.0674 sec/batch
Epoch: 3/20... Training Step: 11815... Training loss: 1.5879... 0.0689 sec/batch
Epoch: 3/20... Training Step: 11816... Training loss: 1.7194... 0.0975 sec/batch
Epoch: 3/20... Training Step: 11817... Training loss: 1.6689... 0.0917 sec/batch
Epoch: 3/20... Training Step: 11818... Training loss: 1.6762... 0.0913 sec/batch
Epoch: 3/20... Training Step: 11819... Training loss: 1.6553... 0.0649 sec/batch
Epoch: 3/20... Training Step: 11820... Training loss: 1.6768... 0.1078 sec/batch
Epoch: 3/20... Training Step: 11821... Training loss: 1.7887... 0.0840 sec/batch
Epoch: 3/20... Training Step: 11822... Training loss: 1.8596... 0.0830 sec/batch
Epoch: 3/20... Training Step: 11823... Training loss: 1.7700... 0.0861 sec/batch
Epoch: 3/20... Training Step: 11824... Training loss: 1.6958... 0.0724 sec/batch
Epoch: 3/20... Training Step: 11825... Training loss: 1.7522... 0.0850 sec/batch
Epoch: 3/20... Training Step: 11826... Training loss: 1.6969... 0.0961 sec/batch
Epoch: 3/20... Training Step: 11827... Training loss: 1.7359... 0.0698 sec/batch
Epoch: 3/20... Training Step: 11828... Training loss: 1.8698... 0.0716 sec/batch
Epoch: 3/20... Training Step: 11829... Training loss: 1.6985... 0.0705 sec/batch
Epoch: 3/20... Training Step: 11830... Training loss: 1.8060... 0.0906 sec/batch
Epoch: 3/20... Training Step: 11831... Training loss: 1.8090... 0.0772 sec/batch
Epoch: 3/20... Training Step: 11832... Training loss: 1.7016... 0.0690 sec/batch
Epoch: 3/20... Training Step: 11833... Training loss: 1.7190... 0.0577 sec/batch
Epoch: 3/20... Training Step: 11834... Training loss: 1.7885... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11835... Training loss: 1.6949... 0.0693 sec/batch
Epoch: 3/20... Training Step: 11836... Training loss: 2.0271... 0.0498 sec/batch
Epoch: 3/20... Training Step: 11837... Training loss: 1.8827... 0.0514 sec/batch
Epoch: 3/20... Training Step: 11838... Training loss: 1.7663... 0.0507 sec/batch
Epoch: 3/20... Training Step: 11839... Training loss: 1.7326... 0.0743 sec/batch
Epoch: 3/20... Training Step: 11840... Training loss: 1.8207... 0.0961 sec/batch
Epoch: 3/20... Training Step: 11841... Training loss: 1.8103... 0.0821 sec/batch
Epoch: 3/20... Training Step: 11842... Training loss: 2.0066... 0.0829 sec/batch
Epoch: 3/20... Training Step: 11843... Training loss: 1.8830... 0.0669 sec/batch
Epoch: 3/20... Training Step: 11844... Training loss: 1.8330... 0.0773 sec/batch
Epoch: 3/20... Training Step: 11845... Training loss: 1.9530... 0.0467 sec/batch
Epoch: 3/20... Training Step: 11846... Training loss: 1.9966... 0.0790 sec/batch
Epoch: 3/20... Training Step: 11847... Training loss: 1.8605... 0.0875 sec/batch
Epoch: 3/20... Training Step: 11848... Training loss: 1.7718... 0.1119 sec/batch
Epoch: 3/20... Training Step: 11849... Training loss: 1.9844... 0.0932 sec/batch
Epoch: 3/20... Training Step: 11850... Training loss: 1.9147... 0.0952 sec/batch
Epoch: 3/20... Training Step: 11851... Training loss: 1.9236... 0.0816 sec/batch
Epoch: 3/20... Training Step: 11852... Training loss: 1.7371... 0.0744 sec/batch
Epoch: 3/20... Training Step: 11853... Training loss: 1.7070... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11854... Training loss: 1.7784... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11855... Training loss: 1.6784... 0.0489 sec/batch
Epoch: 3/20... Training Step: 11856... Training loss: 1.6590... 0.0610 sec/batch
Epoch: 3/20... Training Step: 11857... Training loss: 1.7710... 0.0810 sec/batch
Epoch: 3/20... Training Step: 11858... Training loss: 1.6852... 0.0714 sec/batch
Epoch: 3/20... Training Step: 11859... Training loss: 1.7355... 0.0687 sec/batch
Epoch: 3/20... Training Step: 11860... Training loss: 1.9897... 0.0549 sec/batch
Epoch: 3/20... Training Step: 11861... Training loss: 1.6405... 0.0505 sec/batch
Epoch: 3/20... Training Step: 11862... Training loss: 1.8014... 0.0464 sec/batch
Epoch: 3/20... Training Step: 11863... Training loss: 1.7003... 0.0497 sec/batch
Epoch: 3/20... Training Step: 11864... Training loss: 1.8159... 0.0483 sec/batch
Epoch: 3/20... Training Step: 11865... Training loss: 1.8732... 0.0612 sec/batch
Epoch: 3/20... Training Step: 11866... Training loss: 1.7766... 0.0485 sec/batch
Epoch: 3/20... Training Step: 11867... Training loss: 1.7449... 0.0487 sec/batch
Epoch: 3/20... Training Step: 11868... Training loss: 1.6803... 0.0605 sec/batch
Epoch: 3/20... Training Step: 11869... Training loss: 1.7051... 0.0512 sec/batch
Epoch: 3/20... Training Step: 11870... Training loss: 1.8074... 0.0602 sec/batch
Epoch: 3/20... Training Step: 11871... Training loss: 1.7569... 0.0704 sec/batch
Epoch: 3/20... Training Step: 11872... Training loss: 1.8862... 0.0509 sec/batch
Epoch: 3/20... Training Step: 11873... Training loss: 1.8711... 0.0496 sec/batch
Epoch: 3/20... Training Step: 11874... Training loss: 1.7332... 0.0800 sec/batch
Epoch: 3/20... Training Step: 11875... Training loss: 1.6499... 0.0765 sec/batch
Epoch: 3/20... Training Step: 11876... Training loss: 1.6480... 0.0640 sec/batch
Epoch: 3/20... Training Step: 11877... Training loss: 1.7827... 0.0802 sec/batch
Epoch: 3/20... Training Step: 11878... Training loss: 1.7000... 0.0583 sec/batch
Epoch: 3/20... Training Step: 11879... Training loss: 1.9086... 0.0534 sec/batch
Epoch: 3/20... Training Step: 11880... Training loss: 1.6944... 0.0626 sec/batch
Epoch: 3/20... Training Step: 11881... Training loss: 1.7783... 0.0588 sec/batch
Epoch: 3/20... Training Step: 11882... Training loss: 1.7665... 0.0768 sec/batch
Epoch: 3/20... Training Step: 11883... Training loss: 1.7235... 0.0479 sec/batch
Epoch: 3/20... Training Step: 11884... Training loss: 1.6573... 0.0488 sec/batch
Epoch: 3/20... Training Step: 11885... Training loss: 1.7115... 0.0566 sec/batch
Epoch: 3/20... Training Step: 11886... Training loss: 1.9421... 0.0525 sec/batch
Epoch: 3/20... Training Step: 11887... Training loss: 1.8007... 0.0534 sec/batch
Epoch: 3/20... Training Step: 11888... Training loss: 1.7692... 0.0511 sec/batch
Epoch: 3/20... Training Step: 11889... Training loss: 1.7710... 0.0596 sec/batch
Epoch: 3/20... Training Step: 11890... Training loss: 1.6254... 0.0682 sec/batch
Epoch: 3/20... Training Step: 11891... Training loss: 1.6698... 0.0475 sec/batch
Epoch: 3/20... Training Step: 11892... Training loss: 1.6109... 0.0465 sec/batch
Epoch: 3/20... Training Step: 11893... Training loss: 1.6675... 0.0483 sec/batch
Epoch: 3/20... Training Step: 11894... Training loss: 1.7639... 0.0493 sec/batch
Epoch: 3/20... Training Step: 11895... Training loss: 1.5717... 0.0492 sec/batch
Epoch: 3/20... Training Step: 11896... Training loss: 1.8538... 0.0508 sec/batch
Epoch: 3/20... Training Step: 11897... Training loss: 1.7759... 0.0467 sec/batch
Epoch: 3/20... Training Step: 11898... Training loss: 1.7537... 0.0598 sec/batch
Epoch: 3/20... Training Step: 11899... Training loss: 1.6473... 0.0469 sec/batch
Epoch: 3/20... Training Step: 11900... Training loss: 1.7711... 0.0581 sec/batch
Epoch: 3/20... Training Step: 11901... Training loss: 1.6552... 0.0499 sec/batch
Epoch: 3/20... Training Step: 11902... Training loss: 1.8105... 0.0507 sec/batch
Epoch: 3/20... Training Step: 11903... Training loss: 1.7959... 0.0463 sec/batch
Epoch: 3/20... Training Step: 11904... Training loss: 1.6373... 0.0514 sec/batch
Epoch: 3/20... Training Step: 11905... Training loss: 1.7207... 0.0478 sec/batch
Epoch: 3/20... Training Step: 11906... Training loss: 1.7604... 0.0419 sec/batch
Epoch: 3/20... Training Step: 11907... Training loss: 1.6539... 0.0439 sec/batch
Epoch: 3/20... Training Step: 11908... Training loss: 1.7644... 0.0560 sec/batch
Epoch: 3/20... Training Step: 11909... Training loss: 1.8049... 0.0477 sec/batch
Epoch: 3/20... Training Step: 11910... Training loss: 1.8160... 0.0455 sec/batch
Epoch: 4/20... Training Step: 11911... Training loss: 1.6984... 0.0405 sec/batch
Epoch: 4/20... Training Step: 11912... Training loss: 1.5992... 0.0554 sec/batch
Epoch: 4/20... Training Step: 11913... Training loss: 1.6323... 0.0505 sec/batch
Epoch: 4/20... Training Step: 11914... Training loss: 1.5925... 0.0438 sec/batch
Epoch: 4/20... Training Step: 11915... Training loss: 1.7358... 0.0450 sec/batch
Epoch: 4/20... Training Step: 11916... Training loss: 1.6187... 0.0429 sec/batch
Epoch: 4/20... Training Step: 11917... Training loss: 1.5820... 0.0448 sec/batch
Epoch: 4/20... Training Step: 11918... Training loss: 1.6456... 0.0580 sec/batch
Epoch: 4/20... Training Step: 11919... Training loss: 1.7282... 0.0640 sec/batch
Epoch: 4/20... Training Step: 11920... Training loss: 1.7233... 0.0455 sec/batch
Epoch: 4/20... Training Step: 11921... Training loss: 1.5920... 0.0453 sec/batch
Epoch: 4/20... Training Step: 11922... Training loss: 1.5383... 0.0527 sec/batch
Epoch: 4/20... Training Step: 11923... Training loss: 1.6358... 0.0427 sec/batch
Epoch: 4/20... Training Step: 11924... Training loss: 1.6372... 0.0635 sec/batch
Epoch: 4/20... Training Step: 11925... Training loss: 1.5354... 0.0456 sec/batch
Epoch: 4/20... Training Step: 11926... Training loss: 1.6494... 0.0434 sec/batch
Epoch: 4/20... Training Step: 11927... Training loss: 1.6162... 0.0477 sec/batch
Epoch: 4/20... Training Step: 11928... Training loss: 1.5464... 0.0440 sec/batch
Epoch: 4/20... Training Step: 11929... Training loss: 1.6749... 0.0441 sec/batch
Epoch: 4/20... Training Step: 11930... Training loss: 1.8152... 0.0429 sec/batch
Epoch: 4/20... Training Step: 11931... Training loss: 1.6634... 0.0495 sec/batch
Epoch: 4/20... Training Step: 11932... Training loss: 1.5933... 0.0462 sec/batch
Epoch: 4/20... Training Step: 11933... Training loss: 1.6019... 0.0446 sec/batch
Epoch: 4/20... Training Step: 11934... Training loss: 1.6525... 0.0494 sec/batch
Epoch: 4/20... Training Step: 11935... Training loss: 1.6214... 0.0461 sec/batch
Epoch: 4/20... Training Step: 11936... Training loss: 1.6897... 0.0515 sec/batch
Epoch: 4/20... Training Step: 11937... Training loss: 1.8526... 0.0480 sec/batch
Epoch: 4/20... Training Step: 11938... Training loss: 1.9426... 0.0579 sec/batch
Epoch: 4/20... Training Step: 11939... Training loss: 1.9317... 0.0777 sec/batch
Epoch: 4/20... Training Step: 11940... Training loss: 1.6802... 0.0648 sec/batch
Epoch: 4/20... Training Step: 11941... Training loss: 1.8510... 0.0756 sec/batch
Epoch: 4/20... Training Step: 11942... Training loss: 1.8229... 0.0880 sec/batch
Epoch: 4/20... Training Step: 11943... Training loss: 1.7308... 0.0798 sec/batch
Epoch: 4/20... Training Step: 11944... Training loss: 1.7421... 0.0605 sec/batch
Epoch: 4/20... Training Step: 11945... Training loss: 1.7380... 0.0455 sec/batch
Epoch: 4/20... Training Step: 11946... Training loss: 1.5647... 0.0424 sec/batch
Epoch: 4/20... Training Step: 11947... Training loss: 1.6617... 0.0550 sec/batch
Epoch: 4/20... Training Step: 11948... Training loss: 1.7996... 0.0508 sec/batch
Epoch: 4/20... Training Step: 11949... Training loss: 1.7100... 0.0454 sec/batch
Epoch: 4/20... Training Step: 11950... Training loss: 1.7226... 0.0496 sec/batch
Epoch: 4/20... Training Step: 11951... Training loss: 1.7051... 0.0630 sec/batch
Epoch: 4/20... Training Step: 11952... Training loss: 1.6732... 0.0928 sec/batch
Epoch: 4/20... Training Step: 11953... Training loss: 1.7105... 0.0897 sec/batch
Epoch: 4/20... Training Step: 11954... Training loss: 1.6478... 0.0824 sec/batch
Epoch: 4/20... Training Step: 11955... Training loss: 1.6697... 0.0727 sec/batch
Epoch: 4/20... Training Step: 11956... Training loss: 1.8287... 0.0498 sec/batch
Epoch: 4/20... Training Step: 11957... Training loss: 1.6543... 0.0402 sec/batch
Epoch: 4/20... Training Step: 11958... Training loss: 1.6581... 0.0565 sec/batch
Epoch: 4/20... Training Step: 11959... Training loss: 1.7922... 0.0498 sec/batch
Epoch: 4/20... Training Step: 11960... Training loss: 1.8829... 0.0422 sec/batch
Epoch: 4/20... Training Step: 11961... Training loss: 1.7494... 0.0423 sec/batch
Epoch: 4/20... Training Step: 11962... Training loss: 1.6622... 0.0527 sec/batch
Epoch: 4/20... Training Step: 11963... Training loss: 1.8364... 0.0517 sec/batch
Epoch: 4/20... Training Step: 11964... Training loss: 1.8782... 0.0571 sec/batch
Epoch: 4/20... Training Step: 11965... Training loss: 1.7625... 0.0441 sec/batch
Epoch: 4/20... Training Step: 11966... Training loss: 1.6548... 0.0617 sec/batch
Epoch: 4/20... Training Step: 11967... Training loss: 1.7302... 0.0618 sec/batch
Epoch: 4/20... Training Step: 11968... Training loss: 1.5959... 0.0497 sec/batch
Epoch: 4/20... Training Step: 11969... Training loss: 1.7240... 0.0641 sec/batch
Epoch: 4/20... Training Step: 11970... Training loss: 1.7085... 0.0563 sec/batch
Epoch: 4/20... Training Step: 11971... Training loss: 1.8832... 0.0535 sec/batch
Epoch: 4/20... Training Step: 11972... Training loss: 1.8541... 0.0474 sec/batch
Epoch: 4/20... Training Step: 11973... Training loss: 1.7297... 0.0574 sec/batch
Epoch: 4/20... Training Step: 11974... Training loss: 1.5447... 0.0476 sec/batch
Epoch: 4/20... Training Step: 11975... Training loss: 1.8690... 0.0479 sec/batch
Epoch: 4/20... Training Step: 11976... Training loss: 1.7630... 0.0448 sec/batch
Epoch: 4/20... Training Step: 11977... Training loss: 1.7390... 0.0427 sec/batch
Epoch: 4/20... Training Step: 11978... Training loss: 1.8599... 0.0452 sec/batch
Epoch: 4/20... Training Step: 11979... Training loss: 1.8996... 0.0460 sec/batch
Epoch: 4/20... Training Step: 11980... Training loss: 1.7122... 0.0594 sec/batch
Epoch: 4/20... Training Step: 11981... Training loss: 1.7589... 0.0637 sec/batch
Epoch: 4/20... Training Step: 11982... Training loss: 1.6936... 0.0569 sec/batch
Epoch: 4/20... Training Step: 11983... Training loss: 1.6272... 0.0475 sec/batch
Epoch: 4/20... Training Step: 11984... Training loss: 1.7061... 0.0475 sec/batch
Epoch: 4/20... Training Step: 11985... Training loss: 1.5464... 0.0435 sec/batch
Epoch: 4/20... Training Step: 11986... Training loss: 1.6301... 0.0544 sec/batch
Epoch: 4/20... Training Step: 11987... Training loss: 1.6703... 0.0537 sec/batch
Epoch: 4/20... Training Step: 11988... Training loss: 1.5578... 0.0512 sec/batch
Epoch: 4/20... Training Step: 11989... Training loss: 1.6203... 0.0463 sec/batch
Epoch: 4/20... Training Step: 11990... Training loss: 1.6955... 0.0537 sec/batch
Epoch: 4/20... Training Step: 11991... Training loss: 1.5802... 0.0521 sec/batch
Epoch: 4/20... Training Step: 11992... Training loss: 1.5863... 0.0645 sec/batch
Epoch: 4/20... Training Step: 11993... Training loss: 1.6249... 0.0617 sec/batch
Epoch: 4/20... Training Step: 11994... Training loss: 1.6366... 0.0541 sec/batch
Epoch: 4/20... Training Step: 11995... Training loss: 1.5385... 0.0543 sec/batch
Epoch: 4/20... Training Step: 11996... Training loss: 1.6558... 0.0512 sec/batch
Epoch: 4/20... Training Step: 11997... Training loss: 1.6877... 0.0483 sec/batch
Epoch: 4/20... Training Step: 11998... Training loss: 1.5836... 0.0493 sec/batch
Epoch: 4/20... Training Step: 11999... Training loss: 1.6086... 0.0481 sec/batch
Epoch: 4/20... Training Step: 12000... Training loss: 1.7411... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12001... Training loss: 1.5949... 0.0649 sec/batch
Epoch: 4/20... Training Step: 12002... Training loss: 1.6054... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12003... Training loss: 1.6851... 0.0596 sec/batch
Epoch: 4/20... Training Step: 12004... Training loss: 1.7255... 0.0490 sec/batch
Epoch: 4/20... Training Step: 12005... Training loss: 1.5976... 0.0568 sec/batch
Epoch: 4/20... Training Step: 12006... Training loss: 1.6184... 0.0477 sec/batch
Epoch: 4/20... Training Step: 12007... Training loss: 1.7435... 0.0489 sec/batch
Epoch: 4/20... Training Step: 12008... Training loss: 1.7449... 0.0577 sec/batch
Epoch: 4/20... Training Step: 12009... Training loss: 1.7423... 0.0514 sec/batch
Epoch: 4/20... Training Step: 12010... Training loss: 1.6722... 0.0443 sec/batch
Epoch: 4/20... Training Step: 12011... Training loss: 1.7670... 0.0547 sec/batch
Epoch: 4/20... Training Step: 12012... Training loss: 1.6433... 0.0466 sec/batch
Epoch: 4/20... Training Step: 12013... Training loss: 1.7230... 0.0668 sec/batch
Epoch: 4/20... Training Step: 12014... Training loss: 1.5847... 0.0581 sec/batch
Epoch: 4/20... Training Step: 12015... Training loss: 1.6963... 0.0496 sec/batch
Epoch: 4/20... Training Step: 12016... Training loss: 1.6063... 0.0544 sec/batch
Epoch: 4/20... Training Step: 12017... Training loss: 1.5796... 0.0564 sec/batch
Epoch: 4/20... Training Step: 12018... Training loss: 1.5890... 0.0570 sec/batch
Epoch: 4/20... Training Step: 12019... Training loss: 1.6190... 0.0448 sec/batch
Epoch: 4/20... Training Step: 12020... Training loss: 1.6575... 0.0474 sec/batch
Epoch: 4/20... Training Step: 12021... Training loss: 1.5543... 0.0438 sec/batch
Epoch: 4/20... Training Step: 12022... Training loss: 1.6852... 0.0548 sec/batch
Epoch: 4/20... Training Step: 12023... Training loss: 1.6063... 0.0541 sec/batch
Epoch: 4/20... Training Step: 12024... Training loss: 1.6343... 0.0502 sec/batch
Epoch: 4/20... Training Step: 12025... Training loss: 1.7122... 0.0486 sec/batch
Epoch: 4/20... Training Step: 12026... Training loss: 1.7288... 0.0551 sec/batch
Epoch: 4/20... Training Step: 12027... Training loss: 1.6795... 0.0469 sec/batch
Epoch: 4/20... Training Step: 12028... Training loss: 1.5896... 0.0470 sec/batch
Epoch: 4/20... Training Step: 12029... Training loss: 1.5714... 0.0540 sec/batch
Epoch: 4/20... Training Step: 12030... Training loss: 1.6212... 0.0492 sec/batch
Epoch: 4/20... Training Step: 12031... Training loss: 1.5407... 0.0481 sec/batch
Epoch: 4/20... Training Step: 12032... Training loss: 1.5615... 0.0557 sec/batch
Epoch: 4/20... Training Step: 12033... Training loss: 1.6644... 0.0465 sec/batch
Epoch: 4/20... Training Step: 12034... Training loss: 1.5952... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12035... Training loss: 1.6481... 0.0563 sec/batch
Epoch: 4/20... Training Step: 12036... Training loss: 1.6767... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12037... Training loss: 1.6466... 0.0488 sec/batch
Epoch: 4/20... Training Step: 12038... Training loss: 1.5520... 0.0500 sec/batch
Epoch: 4/20... Training Step: 12039... Training loss: 1.8067... 0.1036 sec/batch
Epoch: 4/20... Training Step: 12040... Training loss: 1.6310... 0.0676 sec/batch
Epoch: 4/20... Training Step: 12041... Training loss: 1.7485... 0.0590 sec/batch
Epoch: 4/20... Training Step: 12042... Training loss: 1.5725... 0.0682 sec/batch
Epoch: 4/20... Training Step: 12043... Training loss: 1.7102... 0.0613 sec/batch
Epoch: 4/20... Training Step: 12044... Training loss: 1.6681... 0.0616 sec/batch
Epoch: 4/20... Training Step: 12045... Training loss: 1.6968... 0.0486 sec/batch
Epoch: 4/20... Training Step: 12046... Training loss: 1.5948... 0.0485 sec/batch
Epoch: 4/20... Training Step: 12047... Training loss: 1.7393... 0.0550 sec/batch
Epoch: 4/20... Training Step: 12048... Training loss: 1.6147... 0.0523 sec/batch
Epoch: 4/20... Training Step: 12049... Training loss: 1.6376... 0.0550 sec/batch
Epoch: 4/20... Training Step: 12050... Training loss: 1.6219... 0.0652 sec/batch
Epoch: 4/20... Training Step: 12051... Training loss: 1.6971... 0.0529 sec/batch
Epoch: 4/20... Training Step: 12052... Training loss: 1.6458... 0.0546 sec/batch
Epoch: 4/20... Training Step: 12053... Training loss: 1.6608... 0.0503 sec/batch
Epoch: 4/20... Training Step: 12054... Training loss: 1.7595... 0.1006 sec/batch
Epoch: 4/20... Training Step: 12055... Training loss: 1.5207... 0.0915 sec/batch
Epoch: 4/20... Training Step: 12056... Training loss: 1.7678... 0.0878 sec/batch
Epoch: 4/20... Training Step: 12057... Training loss: 1.7094... 0.0913 sec/batch
Epoch: 4/20... Training Step: 12058... Training loss: 1.5373... 0.0813 sec/batch
Epoch: 4/20... Training Step: 12059... Training loss: 1.5225... 0.0638 sec/batch
Epoch: 4/20... Training Step: 12060... Training loss: 1.5378... 0.0616 sec/batch
Epoch: 4/20... Training Step: 12061... Training loss: 1.7522... 0.0773 sec/batch
Epoch: 4/20... Training Step: 12062... Training loss: 1.6280... 0.0494 sec/batch
Epoch: 4/20... Training Step: 12063... Training loss: 1.6182... 0.0529 sec/batch
Epoch: 4/20... Training Step: 12064... Training loss: 1.6324... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12065... Training loss: 1.8960... 0.0554 sec/batch
Epoch: 4/20... Training Step: 12066... Training loss: 1.6998... 0.0729 sec/batch
Epoch: 4/20... Training Step: 12067... Training loss: 1.5353... 0.0722 sec/batch
Epoch: 4/20... Training Step: 12068... Training loss: 1.7546... 0.0807 sec/batch
Epoch: 4/20... Training Step: 12069... Training loss: 1.8795... 0.0839 sec/batch
Epoch: 4/20... Training Step: 12070... Training loss: 1.8908... 0.0677 sec/batch
Epoch: 4/20... Training Step: 12071... Training loss: 1.7194... 0.0452 sec/batch
Epoch: 4/20... Training Step: 12072... Training loss: 1.6902... 0.0511 sec/batch
Epoch: 4/20... Training Step: 12073... Training loss: 1.6493... 0.0460 sec/batch
Epoch: 4/20... Training Step: 12074... Training loss: 1.7198... 0.0469 sec/batch
Epoch: 4/20... Training Step: 12075... Training loss: 1.7769... 0.0518 sec/batch
Epoch: 4/20... Training Step: 12076... Training loss: 1.6935... 0.0529 sec/batch
Epoch: 4/20... Training Step: 12077... Training loss: 1.6437... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12078... Training loss: 1.8464... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12079... Training loss: 1.6582... 0.0558 sec/batch
Epoch: 4/20... Training Step: 12080... Training loss: 1.6990... 0.0844 sec/batch
Epoch: 4/20... Training Step: 12081... Training loss: 1.6817... 0.0731 sec/batch
Epoch: 4/20... Training Step: 12082... Training loss: 1.7235... 0.0879 sec/batch
Epoch: 4/20... Training Step: 12083... Training loss: 1.7523... 0.0838 sec/batch
Epoch: 4/20... Training Step: 12084... Training loss: 1.6599... 0.0736 sec/batch
Epoch: 4/20... Training Step: 12085... Training loss: 1.7256... 0.0875 sec/batch
Epoch: 4/20... Training Step: 12086... Training loss: 1.6325... 0.1029 sec/batch
Epoch: 4/20... Training Step: 12087... Training loss: 1.5952... 0.0966 sec/batch
Epoch: 4/20... Training Step: 12088... Training loss: 1.6455... 0.0875 sec/batch
Epoch: 4/20... Training Step: 12089... Training loss: 1.6285... 0.0870 sec/batch
Epoch: 4/20... Training Step: 12090... Training loss: 1.5960... 0.0870 sec/batch
Epoch: 4/20... Training Step: 12091... Training loss: 1.5974... 0.0769 sec/batch
Epoch: 4/20... Training Step: 12092... Training loss: 1.7274... 0.0715 sec/batch
Epoch: 4/20... Training Step: 12093... Training loss: 1.6014... 0.0701 sec/batch
Epoch: 4/20... Training Step: 12094... Training loss: 1.6046... 0.0732 sec/batch
Epoch: 4/20... Training Step: 12095... Training loss: 1.5856... 0.0612 sec/batch
Epoch: 4/20... Training Step: 12096... Training loss: 1.6269... 0.0614 sec/batch
Epoch: 4/20... Training Step: 12097... Training loss: 1.5441... 0.0704 sec/batch
Epoch: 4/20... Training Step: 12098... Training loss: 1.5887... 0.0698 sec/batch
Epoch: 4/20... Training Step: 12099... Training loss: 1.6076... 0.0764 sec/batch
Epoch: 4/20... Training Step: 12100... Training loss: 1.7634... 0.0573 sec/batch
Epoch: 4/20... Training Step: 12101... Training loss: 1.6022... 0.0753 sec/batch
Epoch: 4/20... Training Step: 12102... Training loss: 1.5948... 0.0671 sec/batch
Epoch: 4/20... Training Step: 12103... Training loss: 1.6248... 0.0700 sec/batch
Epoch: 4/20... Training Step: 12104... Training loss: 1.7803... 0.0803 sec/batch
Epoch: 4/20... Training Step: 12105... Training loss: 1.6975... 0.0825 sec/batch
Epoch: 4/20... Training Step: 12106... Training loss: 1.6084... 0.0837 sec/batch
Epoch: 4/20... Training Step: 12107... Training loss: 1.5878... 0.0699 sec/batch
Epoch: 4/20... Training Step: 12108... Training loss: 1.7803... 0.0670 sec/batch
Epoch: 4/20... Training Step: 12109... Training loss: 1.6234... 0.0648 sec/batch
Epoch: 4/20... Training Step: 12110... Training loss: 1.6755... 0.0676 sec/batch
Epoch: 4/20... Training Step: 12111... Training loss: 1.5617... 0.0641 sec/batch
Epoch: 4/20... Training Step: 12112... Training loss: 1.7602... 0.0674 sec/batch
Epoch: 4/20... Training Step: 12113... Training loss: 1.6932... 0.0652 sec/batch
Epoch: 4/20... Training Step: 12114... Training loss: 1.7807... 0.0910 sec/batch
Epoch: 4/20... Training Step: 12115... Training loss: 1.6922... 0.1050 sec/batch
Epoch: 4/20... Training Step: 12116... Training loss: 1.7053... 0.0931 sec/batch
Epoch: 4/20... Training Step: 12117... Training loss: 1.7021... 0.0916 sec/batch
Epoch: 4/20... Training Step: 12118... Training loss: 1.6956... 0.0733 sec/batch
Epoch: 4/20... Training Step: 12119... Training loss: 1.6972... 0.0832 sec/batch
Epoch: 4/20... Training Step: 12120... Training loss: 1.6134... 0.0929 sec/batch
Epoch: 4/20... Training Step: 12121... Training loss: 1.5889... 0.0970 sec/batch
Epoch: 4/20... Training Step: 12122... Training loss: 1.7308... 0.1025 sec/batch
Epoch: 4/20... Training Step: 12123... Training loss: 1.6657... 0.1057 sec/batch
Epoch: 4/20... Training Step: 12124... Training loss: 1.7512... 0.0924 sec/batch
Epoch: 4/20... Training Step: 12125... Training loss: 1.6636... 0.0869 sec/batch
Epoch: 4/20... Training Step: 12126... Training loss: 1.7442... 0.0837 sec/batch
Epoch: 4/20... Training Step: 12127... Training loss: 1.7647... 0.0780 sec/batch
Epoch: 4/20... Training Step: 12128... Training loss: 1.7457... 0.0848 sec/batch
Epoch: 4/20... Training Step: 12129... Training loss: 1.6918... 0.1039 sec/batch
Epoch: 4/20... Training Step: 12130... Training loss: 1.6134... 0.0684 sec/batch
Epoch: 4/20... Training Step: 12131... Training loss: 1.7061... 0.0945 sec/batch
Epoch: 4/20... Training Step: 12132... Training loss: 1.7164... 0.0954 sec/batch
Epoch: 4/20... Training Step: 12133... Training loss: 1.7576... 0.1081 sec/batch
Epoch: 4/20... Training Step: 12134... Training loss: 1.7566... 0.0879 sec/batch
Epoch: 4/20... Training Step: 12135... Training loss: 1.7070... 0.0963 sec/batch
Epoch: 4/20... Training Step: 12136... Training loss: 1.7573... 0.0997 sec/batch
Epoch: 4/20... Training Step: 12137... Training loss: 1.8309... 0.1072 sec/batch
Epoch: 4/20... Training Step: 12138... Training loss: 1.8341... 0.0654 sec/batch
Epoch: 4/20... Training Step: 12139... Training loss: 1.6624... 0.0790 sec/batch
Epoch: 4/20... Training Step: 12140... Training loss: 1.6115... 0.0610 sec/batch
Epoch: 4/20... Training Step: 12141... Training loss: 1.6364... 0.0721 sec/batch
Epoch: 4/20... Training Step: 12142... Training loss: 1.6082... 0.0787 sec/batch
Epoch: 4/20... Training Step: 12143... Training loss: 1.6666... 0.0770 sec/batch
Epoch: 4/20... Training Step: 12144... Training loss: 1.6928... 0.0647 sec/batch
Epoch: 4/20... Training Step: 12145... Training loss: 1.6829... 0.0672 sec/batch
Epoch: 4/20... Training Step: 12146... Training loss: 1.8514... 0.0673 sec/batch
Epoch: 4/20... Training Step: 12147... Training loss: 1.7589... 0.0731 sec/batch
Epoch: 4/20... Training Step: 12148... Training loss: 1.8194... 0.0751 sec/batch
Epoch: 4/20... Training Step: 12149... Training loss: 1.7205... 0.0764 sec/batch
Epoch: 4/20... Training Step: 12150... Training loss: 1.6834... 0.0780 sec/batch
Epoch: 4/20... Training Step: 12151... Training loss: 1.7973... 0.0856 sec/batch
Epoch: 4/20... Training Step: 12152... Training loss: 1.8107... 0.0726 sec/batch
Epoch: 4/20... Training Step: 12153... Training loss: 1.6475... 0.0755 sec/batch
Epoch: 4/20... Training Step: 12154... Training loss: 1.7523... 0.0680 sec/batch
Epoch: 4/20... Training Step: 12155... Training loss: 1.7957... 0.0724 sec/batch
Epoch: 4/20... Training Step: 12156... Training loss: 1.8266... 0.0776 sec/batch
Epoch: 4/20... Training Step: 12157... Training loss: 1.7497... 0.0729 sec/batch
Epoch: 4/20... Training Step: 12158... Training loss: 1.7243... 0.0685 sec/batch
Epoch: 4/20... Training Step: 12159... Training loss: 1.5727... 0.0616 sec/batch
Epoch: 4/20... Training Step: 12160... Training loss: 1.6474... 0.0589 sec/batch
Epoch: 4/20... Training Step: 12161... Training loss: 1.6993... 0.0619 sec/batch
Epoch: 4/20... Training Step: 12162... Training loss: 1.7901... 0.0541 sec/batch
Epoch: 4/20... Training Step: 12163... Training loss: 1.7441... 0.0676 sec/batch
Epoch: 4/20... Training Step: 12164... Training loss: 1.7382... 0.0617 sec/batch
Epoch: 4/20... Training Step: 12165... Training loss: 1.8120... 0.0618 sec/batch
Epoch: 4/20... Training Step: 12166... Training loss: 1.6899... 0.0568 sec/batch
Epoch: 4/20... Training Step: 12167... Training loss: 1.6142... 0.0574 sec/batch
Epoch: 4/20... Training Step: 12168... Training loss: 1.6344... 0.0550 sec/batch
Epoch: 4/20... Training Step: 12169... Training loss: 1.6475... 0.0617 sec/batch
Epoch: 4/20... Training Step: 12170... Training loss: 1.6721... 0.0702 sec/batch
Epoch: 4/20... Training Step: 12171... Training loss: 1.6849... 0.0839 sec/batch
Epoch: 4/20... Training Step: 12172... Training loss: 1.5760... 0.0723 sec/batch
Epoch: 4/20... Training Step: 12173... Training loss: 1.7312... 0.0782 sec/batch
Epoch: 4/20... Training Step: 12174... Training loss: 1.7161... 0.0735 sec/batch
Epoch: 4/20... Training Step: 12175... Training loss: 1.5661... 0.0574 sec/batch
Epoch: 4/20... Training Step: 12176... Training loss: 1.6328... 0.0594 sec/batch
Epoch: 4/20... Training Step: 12177... Training loss: 1.6549... 0.0624 sec/batch
Epoch: 4/20... Training Step: 12178... Training loss: 1.6758... 0.0654 sec/batch
Epoch: 4/20... Training Step: 12179... Training loss: 1.7441... 0.0605 sec/batch
Epoch: 4/20... Training Step: 12180... Training loss: 1.7433... 0.0599 sec/batch
Epoch: 4/20... Training Step: 12181... Training loss: 1.6667... 0.0582 sec/batch
Epoch: 4/20... Training Step: 12182... Training loss: 1.6687... 0.0831 sec/batch
Epoch: 4/20... Training Step: 12183... Training loss: 1.8414... 0.0749 sec/batch
Epoch: 4/20... Training Step: 12184... Training loss: 1.7402... 0.0609 sec/batch
Epoch: 4/20... Training Step: 12185... Training loss: 1.7745... 0.0668 sec/batch
Epoch: 4/20... Training Step: 12186... Training loss: 1.6688... 0.0578 sec/batch
Epoch: 4/20... Training Step: 12187... Training loss: 1.6460... 0.0749 sec/batch
Epoch: 4/20... Training Step: 12188... Training loss: 1.6952... 0.0618 sec/batch
Epoch: 4/20... Training Step: 12189... Training loss: 1.6543... 0.0714 sec/batch
Epoch: 4/20... Training Step: 12190... Training loss: 1.6949... 0.0670 sec/batch
Epoch: 4/20... Training Step: 12191... Training loss: 1.5544... 0.0713 sec/batch
Epoch: 4/20... Training Step: 12192... Training loss: 1.6545... 0.0664 sec/batch
Epoch: 4/20... Training Step: 12193... Training loss: 1.6230... 0.0625 sec/batch
Epoch: 4/20... Training Step: 12194... Training loss: 1.7609... 0.0716 sec/batch
Epoch: 4/20... Training Step: 12195... Training loss: 1.6369... 0.0629 sec/batch
Epoch: 4/20... Training Step: 12196... Training loss: 1.6761... 0.0626 sec/batch
Epoch: 4/20... Training Step: 12197... Training loss: 1.5589... 0.0658 sec/batch
Epoch: 4/20... Training Step: 12198... Training loss: 1.7125... 0.0622 sec/batch
Epoch: 4/20... Training Step: 12199... Training loss: 1.6364... 0.0530 sec/batch
Epoch: 4/20... Training Step: 12200... Training loss: 1.6514... 0.0629 sec/batch
Epoch: 4/20... Training Step: 12201... Training loss: 1.6740... 0.0609 sec/batch
Epoch: 4/20... Training Step: 12202... Training loss: 1.5952... 0.0830 sec/batch
Epoch: 4/20... Training Step: 12203... Training loss: 1.6318... 0.0627 sec/batch
Epoch: 4/20... Training Step: 12204... Training loss: 1.5759... 0.0813 sec/batch
Epoch: 4/20... Training Step: 12205... Training loss: 1.7465... 0.0595 sec/batch
Epoch: 4/20... Training Step: 12206... Training loss: 1.7055... 0.0674 sec/batch
Epoch: 4/20... Training Step: 12207... Training loss: 1.5121... 0.0617 sec/batch
Epoch: 4/20... Training Step: 12208... Training loss: 1.6898... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12209... Training loss: 1.5722... 0.0626 sec/batch
Epoch: 4/20... Training Step: 12210... Training loss: 1.5937... 0.0580 sec/batch
Epoch: 4/20... Training Step: 12211... Training loss: 1.7259... 0.0736 sec/batch
Epoch: 4/20... Training Step: 12212... Training loss: 1.5918... 0.0626 sec/batch
Epoch: 4/20... Training Step: 12213... Training loss: 1.6532... 0.0806 sec/batch
Epoch: 4/20... Training Step: 12214... Training loss: 1.7024... 0.0610 sec/batch
Epoch: 4/20... Training Step: 12215... Training loss: 1.7450... 0.0668 sec/batch
Epoch: 4/20... Training Step: 12216... Training loss: 1.5489... 0.0682 sec/batch
Epoch: 4/20... Training Step: 12217... Training loss: 1.8034... 0.0736 sec/batch
Epoch: 4/20... Training Step: 12218... Training loss: 1.7003... 0.0716 sec/batch
Epoch: 4/20... Training Step: 12219... Training loss: 1.7067... 0.0664 sec/batch
Epoch: 4/20... Training Step: 12220... Training loss: 1.7907... 0.0755 sec/batch
Epoch: 4/20... Training Step: 12221... Training loss: 1.6931... 0.0713 sec/batch
Epoch: 4/20... Training Step: 12222... Training loss: 1.6761... 0.0871 sec/batch
Epoch: 4/20... Training Step: 12223... Training loss: 1.7086... 0.0756 sec/batch
Epoch: 4/20... Training Step: 12224... Training loss: 1.7270... 0.0807 sec/batch
Epoch: 4/20... Training Step: 12225... Training loss: 1.7775... 0.0814 sec/batch
Epoch: 4/20... Training Step: 12226... Training loss: 1.7593... 0.0586 sec/batch
Epoch: 4/20... Training Step: 12227... Training loss: 1.6091... 0.0800 sec/batch
Epoch: 4/20... Training Step: 12228... Training loss: 1.6431... 0.0746 sec/batch
Epoch: 4/20... Training Step: 12229... Training loss: 1.7454... 0.0637 sec/batch
Epoch: 4/20... Training Step: 12230... Training loss: 1.7542... 0.0831 sec/batch
Epoch: 4/20... Training Step: 12231... Training loss: 1.7420... 0.0706 sec/batch
Epoch: 4/20... Training Step: 12232... Training loss: 1.5065... 0.0648 sec/batch
Epoch: 4/20... Training Step: 12233... Training loss: 1.7559... 0.0589 sec/batch
Epoch: 4/20... Training Step: 12234... Training loss: 1.6490... 0.0760 sec/batch
Epoch: 4/20... Training Step: 12235... Training loss: 1.6499... 0.0620 sec/batch
Epoch: 4/20... Training Step: 12236... Training loss: 1.7760... 0.0755 sec/batch
Epoch: 4/20... Training Step: 12237... Training loss: 1.8135... 0.0772 sec/batch
Epoch: 4/20... Training Step: 12238... Training loss: 1.7412... 0.0844 sec/batch
Epoch: 4/20... Training Step: 12239... Training loss: 1.8087... 0.0813 sec/batch
Epoch: 4/20... Training Step: 12240... Training loss: 1.7064... 0.0736 sec/batch
Epoch: 4/20... Training Step: 12241... Training loss: 1.8304... 0.0715 sec/batch
Epoch: 4/20... Training Step: 12242... Training loss: 1.7180... 0.0806 sec/batch
Epoch: 4/20... Training Step: 12243... Training loss: 1.6428... 0.0687 sec/batch
Epoch: 4/20... Training Step: 12244... Training loss: 1.7903... 0.0793 sec/batch
Epoch: 4/20... Training Step: 12245... Training loss: 1.7906... 0.0813 sec/batch
Epoch: 4/20... Training Step: 12246... Training loss: 1.7795... 0.0698 sec/batch
Epoch: 4/20... Training Step: 12247... Training loss: 1.6173... 0.0658 sec/batch
Epoch: 4/20... Training Step: 12248... Training loss: 1.6265... 0.0921 sec/batch
Epoch: 4/20... Training Step: 12249... Training loss: 1.6905... 0.0891 sec/batch
Epoch: 4/20... Training Step: 12250... Training loss: 1.6510... 0.0770 sec/batch
Epoch: 4/20... Training Step: 12251... Training loss: 1.9092... 0.0980 sec/batch
Epoch: 4/20... Training Step: 12252... Training loss: 1.9629... 0.0962 sec/batch
Epoch: 4/20... Training Step: 12253... Training loss: 1.9020... 0.0663 sec/batch
Epoch: 4/20... Training Step: 12254... Training loss: 1.6884... 0.0852 sec/batch
Epoch: 4/20... Training Step: 12255... Training loss: 1.7229... 0.0753 sec/batch
Epoch: 4/20... Training Step: 12256... Training loss: 1.8390... 0.0745 sec/batch
Epoch: 4/20... Training Step: 12257... Training loss: 1.8684... 0.0751 sec/batch
Epoch: 4/20... Training Step: 12258... Training loss: 1.7624... 0.0624 sec/batch
Epoch: 4/20... Training Step: 12259... Training loss: 1.7436... 0.0932 sec/batch
Epoch: 4/20... Training Step: 12260... Training loss: 1.7009... 0.0823 sec/batch
Epoch: 4/20... Training Step: 12261... Training loss: 1.6871... 0.0902 sec/batch
Epoch: 4/20... Training Step: 12262... Training loss: 1.5075... 0.0931 sec/batch
Epoch: 4/20... Training Step: 12263... Training loss: 1.7226... 0.0806 sec/batch
Epoch: 4/20... Training Step: 12264... Training loss: 1.7452... 0.0973 sec/batch
Epoch: 4/20... Training Step: 12265... Training loss: 1.7352... 0.0693 sec/batch
Epoch: 4/20... Training Step: 12266... Training loss: 1.6133... 0.0755 sec/batch
Epoch: 4/20... Training Step: 12267... Training loss: 1.7801... 0.0715 sec/batch
Epoch: 4/20... Training Step: 12268... Training loss: 1.7910... 0.0706 sec/batch
Epoch: 4/20... Training Step: 12269... Training loss: 1.6239... 0.0721 sec/batch
Epoch: 4/20... Training Step: 12270... Training loss: 1.8104... 0.0834 sec/batch
Epoch: 4/20... Training Step: 12271... Training loss: 1.7785... 0.0836 sec/batch
Epoch: 4/20... Training Step: 12272... Training loss: 1.6656... 0.0854 sec/batch
Epoch: 4/20... Training Step: 12273... Training loss: 1.8027... 0.0762 sec/batch
Epoch: 4/20... Training Step: 12274... Training loss: 1.6084... 0.0650 sec/batch
Epoch: 4/20... Training Step: 12275... Training loss: 1.8371... 0.0556 sec/batch
Epoch: 4/20... Training Step: 12276... Training loss: 1.6900... 0.0552 sec/batch
Epoch: 4/20... Training Step: 12277... Training loss: 1.8005... 0.0528 sec/batch
Epoch: 4/20... Training Step: 12278... Training loss: 1.7573... 0.0573 sec/batch
Epoch: 4/20... Training Step: 12279... Training loss: 1.7574... 0.0486 sec/batch
Epoch: 4/20... Training Step: 12280... Training loss: 1.8177... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12281... Training loss: 1.6731... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12282... Training loss: 1.6934... 0.0412 sec/batch
Epoch: 4/20... Training Step: 12283... Training loss: 1.6933... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12284... Training loss: 1.6207... 0.0477 sec/batch
Epoch: 4/20... Training Step: 12285... Training loss: 1.7453... 0.0506 sec/batch
Epoch: 4/20... Training Step: 12286... Training loss: 1.6017... 0.0596 sec/batch
Epoch: 4/20... Training Step: 12287... Training loss: 1.7596... 0.0490 sec/batch
Epoch: 4/20... Training Step: 12288... Training loss: 1.7112... 0.0520 sec/batch
Epoch: 4/20... Training Step: 12289... Training loss: 1.7350... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12290... Training loss: 1.6846... 0.0511 sec/batch
Epoch: 4/20... Training Step: 12291... Training loss: 1.7537... 0.0470 sec/batch
Epoch: 4/20... Training Step: 12292... Training loss: 1.7224... 0.0465 sec/batch
Epoch: 4/20... Training Step: 12293... Training loss: 1.8398... 0.0580 sec/batch
Epoch: 4/20... Training Step: 12294... Training loss: 1.8397... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12295... Training loss: 1.7409... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12296... Training loss: 1.7042... 0.0493 sec/batch
Epoch: 4/20... Training Step: 12297... Training loss: 1.6695... 0.0601 sec/batch
Epoch: 4/20... Training Step: 12298... Training loss: 1.7447... 0.0548 sec/batch
Epoch: 4/20... Training Step: 12299... Training loss: 1.6385... 0.0585 sec/batch
Epoch: 4/20... Training Step: 12300... Training loss: 1.6686... 0.0538 sec/batch
Epoch: 4/20... Training Step: 12301... Training loss: 1.6612... 0.0496 sec/batch
Epoch: 4/20... Training Step: 12302... Training loss: 1.8188... 0.0484 sec/batch
Epoch: 4/20... Training Step: 12303... Training loss: 1.6700... 0.0461 sec/batch
Epoch: 4/20... Training Step: 12304... Training loss: 1.6196... 0.0532 sec/batch
Epoch: 4/20... Training Step: 12305... Training loss: 1.7334... 0.0513 sec/batch
Epoch: 4/20... Training Step: 12306... Training loss: 1.7181... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12307... Training loss: 1.6398... 0.0631 sec/batch
Epoch: 4/20... Training Step: 12308... Training loss: 1.7889... 0.0487 sec/batch
Epoch: 4/20... Training Step: 12309... Training loss: 1.7749... 0.0473 sec/batch
Epoch: 4/20... Training Step: 12310... Training loss: 1.7281... 0.0448 sec/batch
Epoch: 4/20... Training Step: 12311... Training loss: 1.6288... 0.0553 sec/batch
Epoch: 4/20... Training Step: 12312... Training loss: 1.6323... 0.0478 sec/batch
Epoch: 4/20... Training Step: 12313... Training loss: 1.6811... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12314... Training loss: 1.6088... 0.0630 sec/batch
Epoch: 4/20... Training Step: 12315... Training loss: 1.7383... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12316... Training loss: 1.7007... 0.0567 sec/batch
Epoch: 4/20... Training Step: 12317... Training loss: 1.7066... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12318... Training loss: 1.7606... 0.0477 sec/batch
Epoch: 4/20... Training Step: 12319... Training loss: 1.7035... 0.0657 sec/batch
Epoch: 4/20... Training Step: 12320... Training loss: 1.6941... 0.0489 sec/batch
Epoch: 4/20... Training Step: 12321... Training loss: 1.7366... 0.0530 sec/batch
Epoch: 4/20... Training Step: 12322... Training loss: 1.6581... 0.0632 sec/batch
Epoch: 4/20... Training Step: 12323... Training loss: 1.8650... 0.0521 sec/batch
Epoch: 4/20... Training Step: 12324... Training loss: 1.6608... 0.0488 sec/batch
Epoch: 4/20... Training Step: 12325... Training loss: 1.7288... 0.0547 sec/batch
Epoch: 4/20... Training Step: 12326... Training loss: 1.7965... 0.0510 sec/batch
Epoch: 4/20... Training Step: 12327... Training loss: 1.8382... 0.0483 sec/batch
Epoch: 4/20... Training Step: 12328... Training loss: 1.5952... 0.0584 sec/batch
Epoch: 4/20... Training Step: 12329... Training loss: 1.6539... 0.0555 sec/batch
Epoch: 4/20... Training Step: 12330... Training loss: 1.6852... 0.0478 sec/batch
Epoch: 4/20... Training Step: 12331... Training loss: 1.7213... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12332... Training loss: 1.7551... 0.0577 sec/batch
Epoch: 4/20... Training Step: 12333... Training loss: 1.6251... 0.0511 sec/batch
Epoch: 4/20... Training Step: 12334... Training loss: 1.6303... 0.0551 sec/batch
Epoch: 4/20... Training Step: 12335... Training loss: 1.7181... 0.0582 sec/batch
Epoch: 4/20... Training Step: 12336... Training loss: 1.7026... 0.0426 sec/batch
Epoch: 4/20... Training Step: 12337... Training loss: 1.7459... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12338... Training loss: 1.7167... 0.0547 sec/batch
Epoch: 4/20... Training Step: 12339... Training loss: 1.7048... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12340... Training loss: 1.6569... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12341... Training loss: 1.5600... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12342... Training loss: 1.6921... 0.0522 sec/batch
Epoch: 4/20... Training Step: 12343... Training loss: 1.5895... 0.0481 sec/batch
Epoch: 4/20... Training Step: 12344... Training loss: 1.6565... 0.0465 sec/batch
Epoch: 4/20... Training Step: 12345... Training loss: 1.6242... 0.0514 sec/batch
Epoch: 4/20... Training Step: 12346... Training loss: 1.6360... 0.0497 sec/batch
Epoch: 4/20... Training Step: 12347... Training loss: 1.7276... 0.0484 sec/batch
Epoch: 4/20... Training Step: 12348... Training loss: 1.6291... 0.0540 sec/batch
Epoch: 4/20... Training Step: 12349... Training loss: 1.6191... 0.0467 sec/batch
Epoch: 4/20... Training Step: 12350... Training loss: 1.7066... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12351... Training loss: 1.8244... 0.0650 sec/batch
Epoch: 4/20... Training Step: 12352... Training loss: 1.5869... 0.0709 sec/batch
Epoch: 4/20... Training Step: 12353... Training loss: 1.7697... 0.0459 sec/batch
Epoch: 4/20... Training Step: 12354... Training loss: 1.6252... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12355... Training loss: 1.7285... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12356... Training loss: 1.7693... 0.0454 sec/batch
Epoch: 4/20... Training Step: 12357... Training loss: 1.7975... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12358... Training loss: 1.5558... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12359... Training loss: 1.5664... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12360... Training loss: 1.7619... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12361... Training loss: 1.7006... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12362... Training loss: 1.6582... 0.0482 sec/batch
Epoch: 4/20... Training Step: 12363... Training loss: 1.6122... 0.0415 sec/batch
Epoch: 4/20... Training Step: 12364... Training loss: 1.7150... 0.0450 sec/batch
Epoch: 4/20... Training Step: 12365... Training loss: 1.5597... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12366... Training loss: 1.6544... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12367... Training loss: 1.7423... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12368... Training loss: 1.7283... 0.0432 sec/batch
Epoch: 4/20... Training Step: 12369... Training loss: 1.8502... 0.0430 sec/batch
Epoch: 4/20... Training Step: 12370... Training loss: 1.8218... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12371... Training loss: 1.6836... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12372... Training loss: 1.7323... 0.0497 sec/batch
Epoch: 4/20... Training Step: 12373... Training loss: 1.6720... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12374... Training loss: 1.5975... 0.0512 sec/batch
Epoch: 4/20... Training Step: 12375... Training loss: 1.7067... 0.0616 sec/batch
Epoch: 4/20... Training Step: 12376... Training loss: 1.6928... 0.0559 sec/batch
Epoch: 4/20... Training Step: 12377... Training loss: 1.6901... 0.0618 sec/batch
Epoch: 4/20... Training Step: 12378... Training loss: 1.6589... 0.0607 sec/batch
Epoch: 4/20... Training Step: 12379... Training loss: 1.5815... 0.0502 sec/batch
Epoch: 4/20... Training Step: 12380... Training loss: 1.7685... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12381... Training loss: 1.6220... 0.0467 sec/batch
Epoch: 4/20... Training Step: 12382... Training loss: 1.6423... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12383... Training loss: 1.7310... 0.0487 sec/batch
Epoch: 4/20... Training Step: 12384... Training loss: 1.7419... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12385... Training loss: 1.7604... 0.0481 sec/batch
Epoch: 4/20... Training Step: 12386... Training loss: 1.5505... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12387... Training loss: 1.5921... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12388... Training loss: 1.5570... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12389... Training loss: 1.6474... 0.0448 sec/batch
Epoch: 4/20... Training Step: 12390... Training loss: 1.8028... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12391... Training loss: 1.5982... 0.0461 sec/batch
Epoch: 4/20... Training Step: 12392... Training loss: 1.6274... 0.0430 sec/batch
Epoch: 4/20... Training Step: 12393... Training loss: 1.6851... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12394... Training loss: 1.6563... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12395... Training loss: 1.7196... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12396... Training loss: 1.7846... 0.0478 sec/batch
Epoch: 4/20... Training Step: 12397... Training loss: 1.8393... 0.0482 sec/batch
Epoch: 4/20... Training Step: 12398... Training loss: 1.6690... 0.0466 sec/batch
Epoch: 4/20... Training Step: 12399... Training loss: 1.7973... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12400... Training loss: 1.7200... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12401... Training loss: 1.5860... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12402... Training loss: 1.6955... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12403... Training loss: 1.7655... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12404... Training loss: 1.7408... 0.0475 sec/batch
Epoch: 4/20... Training Step: 12405... Training loss: 1.5536... 0.0659 sec/batch
Epoch: 4/20... Training Step: 12406... Training loss: 1.6035... 0.0685 sec/batch
Epoch: 4/20... Training Step: 12407... Training loss: 1.5185... 0.0485 sec/batch
Epoch: 4/20... Training Step: 12408... Training loss: 1.6546... 0.0566 sec/batch
Epoch: 4/20... Training Step: 12409... Training loss: 1.5670... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12410... Training loss: 1.8381... 0.0626 sec/batch
Epoch: 4/20... Training Step: 12411... Training loss: 1.6348... 0.0627 sec/batch
Epoch: 4/20... Training Step: 12412... Training loss: 1.7077... 0.0638 sec/batch
Epoch: 4/20... Training Step: 12413... Training loss: 1.6902... 0.0535 sec/batch
Epoch: 4/20... Training Step: 12414... Training loss: 1.6605... 0.0454 sec/batch
Epoch: 4/20... Training Step: 12415... Training loss: 1.6878... 0.0513 sec/batch
Epoch: 4/20... Training Step: 12416... Training loss: 1.7190... 0.0645 sec/batch
Epoch: 4/20... Training Step: 12417... Training loss: 1.7659... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12418... Training loss: 1.7902... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12419... Training loss: 1.8069... 0.0422 sec/batch
Epoch: 4/20... Training Step: 12420... Training loss: 1.7017... 0.0449 sec/batch
Epoch: 4/20... Training Step: 12421... Training loss: 1.6727... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12422... Training loss: 1.8729... 0.0436 sec/batch
Epoch: 4/20... Training Step: 12423... Training loss: 1.6143... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12424... Training loss: 1.6740... 0.0617 sec/batch
Epoch: 4/20... Training Step: 12425... Training loss: 1.8413... 0.0473 sec/batch
Epoch: 4/20... Training Step: 12426... Training loss: 1.8239... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12427... Training loss: 1.8038... 0.0409 sec/batch
Epoch: 4/20... Training Step: 12428... Training loss: 1.7451... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12429... Training loss: 1.6817... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12430... Training loss: 1.6297... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12431... Training loss: 1.6680... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12432... Training loss: 1.7390... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12433... Training loss: 1.6843... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12434... Training loss: 1.6144... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12435... Training loss: 1.6605... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12436... Training loss: 1.6696... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12437... Training loss: 1.6173... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12438... Training loss: 1.6943... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12439... Training loss: 1.7500... 0.0464 sec/batch
Epoch: 4/20... Training Step: 12440... Training loss: 1.6640... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12441... Training loss: 1.7598... 0.0447 sec/batch
Epoch: 4/20... Training Step: 12442... Training loss: 1.7291... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12443... Training loss: 1.6304... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12444... Training loss: 1.6636... 0.0399 sec/batch
Epoch: 4/20... Training Step: 12445... Training loss: 1.6927... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12446... Training loss: 1.6192... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12447... Training loss: 1.5158... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12448... Training loss: 1.5789... 0.0407 sec/batch
Epoch: 4/20... Training Step: 12449... Training loss: 1.5186... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12450... Training loss: 1.5729... 0.0413 sec/batch
Epoch: 4/20... Training Step: 12451... Training loss: 1.5493... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12452... Training loss: 1.6148... 0.0408 sec/batch
Epoch: 4/20... Training Step: 12453... Training loss: 1.6803... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12454... Training loss: 1.6382... 0.0449 sec/batch
Epoch: 4/20... Training Step: 12455... Training loss: 1.6452... 0.0481 sec/batch
Epoch: 4/20... Training Step: 12456... Training loss: 1.5792... 0.0422 sec/batch
Epoch: 4/20... Training Step: 12457... Training loss: 1.6829... 0.0426 sec/batch
Epoch: 4/20... Training Step: 12458... Training loss: 1.6279... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12459... Training loss: 1.5883... 0.0422 sec/batch
Epoch: 4/20... Training Step: 12460... Training loss: 1.4936... 0.0486 sec/batch
Epoch: 4/20... Training Step: 12461... Training loss: 1.5839... 0.0468 sec/batch
Epoch: 4/20... Training Step: 12462... Training loss: 1.7442... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12463... Training loss: 1.5697... 0.0417 sec/batch
Epoch: 4/20... Training Step: 12464... Training loss: 1.6167... 0.0465 sec/batch
Epoch: 4/20... Training Step: 12465... Training loss: 1.6616... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12466... Training loss: 1.7520... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12467... Training loss: 1.6229... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12468... Training loss: 1.5985... 0.0404 sec/batch
Epoch: 4/20... Training Step: 12469... Training loss: 1.6513... 0.0426 sec/batch
Epoch: 4/20... Training Step: 12470... Training loss: 1.7259... 0.0447 sec/batch
Epoch: 4/20... Training Step: 12471... Training loss: 1.6355... 0.0558 sec/batch
Epoch: 4/20... Training Step: 12472... Training loss: 1.6952... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12473... Training loss: 1.5740... 0.0430 sec/batch
Epoch: 4/20... Training Step: 12474... Training loss: 1.5927... 0.0506 sec/batch
Epoch: 4/20... Training Step: 12475... Training loss: 1.5799... 0.0448 sec/batch
Epoch: 4/20... Training Step: 12476... Training loss: 1.6277... 0.0412 sec/batch
Epoch: 4/20... Training Step: 12477... Training loss: 1.5774... 0.0537 sec/batch
Epoch: 4/20... Training Step: 12478... Training loss: 1.6955... 0.0519 sec/batch
Epoch: 4/20... Training Step: 12479... Training loss: 1.7030... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12480... Training loss: 1.6877... 0.0451 sec/batch
Epoch: 4/20... Training Step: 12481... Training loss: 1.6920... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12482... Training loss: 1.7138... 0.0451 sec/batch
Epoch: 4/20... Training Step: 12483... Training loss: 1.5384... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12484... Training loss: 1.5492... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12485... Training loss: 1.5882... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12486... Training loss: 1.5941... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12487... Training loss: 1.7564... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12488... Training loss: 1.7391... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12489... Training loss: 1.7042... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12490... Training loss: 1.8845... 0.0457 sec/batch
Epoch: 4/20... Training Step: 12491... Training loss: 1.7517... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12492... Training loss: 1.5773... 0.0411 sec/batch
Epoch: 4/20... Training Step: 12493... Training loss: 1.6582... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12494... Training loss: 1.6924... 0.0407 sec/batch
Epoch: 4/20... Training Step: 12495... Training loss: 1.5421... 0.0586 sec/batch
Epoch: 4/20... Training Step: 12496... Training loss: 1.6924... 0.0903 sec/batch
Epoch: 4/20... Training Step: 12497... Training loss: 1.5439... 0.0694 sec/batch
Epoch: 4/20... Training Step: 12498... Training loss: 1.7132... 0.0557 sec/batch
Epoch: 4/20... Training Step: 12499... Training loss: 2.0289... 0.0527 sec/batch
Epoch: 4/20... Training Step: 12500... Training loss: 1.8083... 0.0545 sec/batch
Epoch: 4/20... Training Step: 12501... Training loss: 1.7160... 0.0413 sec/batch
Epoch: 4/20... Training Step: 12502... Training loss: 1.6608... 0.0399 sec/batch
Epoch: 4/20... Training Step: 12503... Training loss: 1.6371... 0.0408 sec/batch
Epoch: 4/20... Training Step: 12504... Training loss: 1.8039... 0.0438 sec/batch
Epoch: 4/20... Training Step: 12505... Training loss: 1.8716... 0.0405 sec/batch
Epoch: 4/20... Training Step: 12506... Training loss: 1.7443... 0.0466 sec/batch
Epoch: 4/20... Training Step: 12507... Training loss: 1.6150... 0.0786 sec/batch
Epoch: 4/20... Training Step: 12508... Training loss: 1.7803... 0.0677 sec/batch
Epoch: 4/20... Training Step: 12509... Training loss: 1.7107... 0.0612 sec/batch
Epoch: 4/20... Training Step: 12510... Training loss: 1.6383... 0.0581 sec/batch
Epoch: 4/20... Training Step: 12511... Training loss: 1.9161... 0.0602 sec/batch
Epoch: 4/20... Training Step: 12512... Training loss: 1.6200... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12513... Training loss: 1.6571... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12514... Training loss: 1.5523... 0.0415 sec/batch
Epoch: 4/20... Training Step: 12515... Training loss: 1.6565... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12516... Training loss: 1.5977... 0.0458 sec/batch
Epoch: 4/20... Training Step: 12517... Training loss: 1.5380... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12518... Training loss: 1.7551... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12519... Training loss: 1.8435... 0.0449 sec/batch
Epoch: 4/20... Training Step: 12520... Training loss: 1.5980... 0.0458 sec/batch
Epoch: 4/20... Training Step: 12521... Training loss: 1.7540... 0.0458 sec/batch
Epoch: 4/20... Training Step: 12522... Training loss: 1.6524... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12523... Training loss: 1.6726... 0.0393 sec/batch
Epoch: 4/20... Training Step: 12524... Training loss: 1.7839... 0.0444 sec/batch
Epoch: 4/20... Training Step: 12525... Training loss: 1.6093... 0.0470 sec/batch
Epoch: 4/20... Training Step: 12526... Training loss: 1.6593... 0.0518 sec/batch
Epoch: 4/20... Training Step: 12527... Training loss: 1.6349... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12528... Training loss: 1.5943... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12529... Training loss: 1.5814... 0.0432 sec/batch
Epoch: 4/20... Training Step: 12530... Training loss: 1.5090... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12531... Training loss: 1.5579... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12532... Training loss: 1.6528... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12533... Training loss: 1.7452... 0.0457 sec/batch
Epoch: 4/20... Training Step: 12534... Training loss: 1.6755... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12535... Training loss: 1.5293... 0.0828 sec/batch
Epoch: 4/20... Training Step: 12536... Training loss: 1.6036... 0.0581 sec/batch
Epoch: 4/20... Training Step: 12537... Training loss: 1.6194... 0.0574 sec/batch
Epoch: 4/20... Training Step: 12538... Training loss: 1.5066... 0.0597 sec/batch
Epoch: 4/20... Training Step: 12539... Training loss: 1.6205... 0.0648 sec/batch
Epoch: 4/20... Training Step: 12540... Training loss: 1.7615... 0.0578 sec/batch
Epoch: 4/20... Training Step: 12541... Training loss: 1.6877... 0.0596 sec/batch
Epoch: 4/20... Training Step: 12542... Training loss: 1.6595... 0.0848 sec/batch
Epoch: 4/20... Training Step: 12543... Training loss: 1.6382... 0.0515 sec/batch
Epoch: 4/20... Training Step: 12544... Training loss: 1.6965... 0.0571 sec/batch
Epoch: 4/20... Training Step: 12545... Training loss: 1.8782... 0.0468 sec/batch
Epoch: 4/20... Training Step: 12546... Training loss: 1.7602... 0.0469 sec/batch
Epoch: 4/20... Training Step: 12547... Training loss: 1.7408... 0.0480 sec/batch
Epoch: 4/20... Training Step: 12548... Training loss: 1.7110... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12549... Training loss: 1.7200... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12550... Training loss: 1.5304... 0.0637 sec/batch
Epoch: 4/20... Training Step: 12551... Training loss: 1.6870... 0.0464 sec/batch
Epoch: 4/20... Training Step: 12552... Training loss: 1.6554... 0.0496 sec/batch
Epoch: 4/20... Training Step: 12553... Training loss: 1.6425... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12554... Training loss: 1.5568... 0.0474 sec/batch
Epoch: 4/20... Training Step: 12555... Training loss: 1.6060... 0.0557 sec/batch
Epoch: 4/20... Training Step: 12556... Training loss: 1.9173... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12557... Training loss: 1.6470... 0.0485 sec/batch
Epoch: 4/20... Training Step: 12558... Training loss: 1.7439... 0.0444 sec/batch
Epoch: 4/20... Training Step: 12559... Training loss: 1.6827... 0.0450 sec/batch
Epoch: 4/20... Training Step: 12560... Training loss: 1.5046... 0.0514 sec/batch
Epoch: 4/20... Training Step: 12561... Training loss: 1.7329... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12562... Training loss: 1.8093... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12563... Training loss: 1.7911... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12564... Training loss: 1.5967... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12565... Training loss: 1.7219... 0.0493 sec/batch
Epoch: 4/20... Training Step: 12566... Training loss: 1.7351... 0.0415 sec/batch
Epoch: 4/20... Training Step: 12567... Training loss: 1.6024... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12568... Training loss: 1.7587... 0.0417 sec/batch
Epoch: 4/20... Training Step: 12569... Training loss: 1.7525... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12570... Training loss: 1.7047... 0.0480 sec/batch
Epoch: 4/20... Training Step: 12571... Training loss: 1.7130... 0.0438 sec/batch
Epoch: 4/20... Training Step: 12572... Training loss: 1.6610... 0.0414 sec/batch
Epoch: 4/20... Training Step: 12573... Training loss: 1.7033... 0.0436 sec/batch
Epoch: 4/20... Training Step: 12574... Training loss: 1.8119... 0.0461 sec/batch
Epoch: 4/20... Training Step: 12575... Training loss: 1.7307... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12576... Training loss: 1.7052... 0.0459 sec/batch
Epoch: 4/20... Training Step: 12577... Training loss: 1.7434... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12578... Training loss: 1.5670... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12579... Training loss: 1.5728... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12580... Training loss: 1.5980... 0.0525 sec/batch
Epoch: 4/20... Training Step: 12581... Training loss: 1.6586... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12582... Training loss: 1.6611... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12583... Training loss: 1.8046... 0.0454 sec/batch
Epoch: 4/20... Training Step: 12584... Training loss: 1.6996... 0.0482 sec/batch
Epoch: 4/20... Training Step: 12585... Training loss: 1.6234... 0.0508 sec/batch
Epoch: 4/20... Training Step: 12586... Training loss: 1.6451... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12587... Training loss: 1.9131... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12588... Training loss: 1.7295... 0.0470 sec/batch
Epoch: 4/20... Training Step: 12589... Training loss: 1.6512... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12590... Training loss: 1.7826... 0.0483 sec/batch
Epoch: 4/20... Training Step: 12591... Training loss: 1.7027... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12592... Training loss: 1.5852... 0.0397 sec/batch
Epoch: 4/20... Training Step: 12593... Training loss: 1.7272... 0.0395 sec/batch
Epoch: 4/20... Training Step: 12594... Training loss: 1.6097... 0.0430 sec/batch
Epoch: 4/20... Training Step: 12595... Training loss: 1.6485... 0.0393 sec/batch
Epoch: 4/20... Training Step: 12596... Training loss: 1.6845... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12597... Training loss: 1.7995... 0.0409 sec/batch
Epoch: 4/20... Training Step: 12598... Training loss: 1.6133... 0.0388 sec/batch
Epoch: 4/20... Training Step: 12599... Training loss: 1.6122... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12600... Training loss: 1.6657... 0.0414 sec/batch
Epoch: 4/20... Training Step: 12601... Training loss: 1.7912... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12602... Training loss: 1.7243... 0.0436 sec/batch
Epoch: 4/20... Training Step: 12603... Training loss: 1.7248... 0.0414 sec/batch
Epoch: 4/20... Training Step: 12604... Training loss: 1.7764... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12605... Training loss: 1.6354... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12606... Training loss: 1.7684... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12607... Training loss: 1.6947... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12608... Training loss: 1.6978... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12609... Training loss: 1.7221... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12610... Training loss: 1.7090... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12611... Training loss: 1.7525... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12612... Training loss: 1.6519... 0.0438 sec/batch
Epoch: 4/20... Training Step: 12613... Training loss: 1.6681... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12614... Training loss: 1.6954... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12615... Training loss: 1.6205... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12616... Training loss: 1.7118... 0.0394 sec/batch
Epoch: 4/20... Training Step: 12617... Training loss: 1.7500... 0.0401 sec/batch
Epoch: 4/20... Training Step: 12618... Training loss: 1.6286... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12619... Training loss: 1.6632... 0.0411 sec/batch
Epoch: 4/20... Training Step: 12620... Training loss: 1.7581... 0.0398 sec/batch
Epoch: 4/20... Training Step: 12621... Training loss: 1.5741... 0.0412 sec/batch
Epoch: 4/20... Training Step: 12622... Training loss: 1.7021... 0.0486 sec/batch
Epoch: 4/20... Training Step: 12623... Training loss: 1.7740... 0.0544 sec/batch
Epoch: 4/20... Training Step: 12624... Training loss: 1.5631... 0.0678 sec/batch
Epoch: 4/20... Training Step: 12625... Training loss: 1.6398... 0.0484 sec/batch
Epoch: 4/20... Training Step: 12626... Training loss: 1.5997... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12627... Training loss: 1.6843... 0.0569 sec/batch
Epoch: 4/20... Training Step: 12628... Training loss: 1.5660... 0.0593 sec/batch
Epoch: 4/20... Training Step: 12629... Training loss: 1.6459... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12630... Training loss: 1.7183... 0.0474 sec/batch
Epoch: 4/20... Training Step: 12631... Training loss: 1.6683... 0.0404 sec/batch
Epoch: 4/20... Training Step: 12632... Training loss: 1.5246... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12633... Training loss: 1.6819... 0.0462 sec/batch
Epoch: 4/20... Training Step: 12634... Training loss: 1.7110... 0.0405 sec/batch
Epoch: 4/20... Training Step: 12635... Training loss: 1.5944... 0.0417 sec/batch
Epoch: 4/20... Training Step: 12636... Training loss: 1.5755... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12637... Training loss: 1.6369... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12638... Training loss: 1.5882... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12639... Training loss: 1.6271... 0.0397 sec/batch
Epoch: 4/20... Training Step: 12640... Training loss: 1.5573... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12641... Training loss: 1.6880... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12642... Training loss: 1.5483... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12643... Training loss: 1.7215... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12644... Training loss: 1.5535... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12645... Training loss: 1.5663... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12646... Training loss: 1.8036... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12647... Training loss: 1.6369... 0.0414 sec/batch
Epoch: 4/20... Training Step: 12648... Training loss: 1.7520... 0.0467 sec/batch
Epoch: 4/20... Training Step: 12649... Training loss: 1.6952... 0.0461 sec/batch
Epoch: 4/20... Training Step: 12650... Training loss: 1.6331... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12651... Training loss: 1.6084... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12652... Training loss: 1.5389... 0.0411 sec/batch
Epoch: 4/20... Training Step: 12653... Training loss: 1.5297... 0.0494 sec/batch
Epoch: 4/20... Training Step: 12654... Training loss: 1.6162... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12655... Training loss: 1.6218... 0.0444 sec/batch
Epoch: 4/20... Training Step: 12656... Training loss: 1.7532... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12657... Training loss: 1.7622... 0.0401 sec/batch
Epoch: 4/20... Training Step: 12658... Training loss: 1.6186... 0.0415 sec/batch
Epoch: 4/20... Training Step: 12659... Training loss: 1.5872... 0.0408 sec/batch
Epoch: 4/20... Training Step: 12660... Training loss: 1.6684... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12661... Training loss: 1.6632... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12662... Training loss: 1.9556... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12663... Training loss: 1.7023... 0.0426 sec/batch
Epoch: 4/20... Training Step: 12664... Training loss: 1.5856... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12665... Training loss: 1.6573... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12666... Training loss: 1.5692... 0.0397 sec/batch
Epoch: 4/20... Training Step: 12667... Training loss: 1.6834... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12668... Training loss: 1.6347... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12669... Training loss: 1.7504... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12670... Training loss: 1.6254... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12671... Training loss: 1.5561... 0.0404 sec/batch
Epoch: 4/20... Training Step: 12672... Training loss: 1.7956... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12673... Training loss: 1.6975... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12674... Training loss: 1.6863... 0.0426 sec/batch
Epoch: 4/20... Training Step: 12675... Training loss: 1.7350... 0.0436 sec/batch
Epoch: 4/20... Training Step: 12676... Training loss: 1.7507... 0.0402 sec/batch
Epoch: 4/20... Training Step: 12677... Training loss: 1.7181... 0.0411 sec/batch
Epoch: 4/20... Training Step: 12678... Training loss: 1.6913... 0.0475 sec/batch
Epoch: 4/20... Training Step: 12679... Training loss: 1.7767... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12680... Training loss: 1.7121... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12681... Training loss: 1.6508... 0.0387 sec/batch
Epoch: 4/20... Training Step: 12682... Training loss: 1.6235... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12683... Training loss: 1.6806... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12684... Training loss: 1.8108... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12685... Training loss: 1.6826... 0.0424 sec/batch
Epoch: 4/20... Training Step: 12686... Training loss: 1.7373... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12687... Training loss: 1.7721... 0.0402 sec/batch
Epoch: 4/20... Training Step: 12688... Training loss: 1.8082... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12689... Training loss: 1.7319... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12690... Training loss: 1.8040... 0.0399 sec/batch
Epoch: 4/20... Training Step: 12691... Training loss: 1.6300... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12692... Training loss: 1.6858... 0.0431 sec/batch
Epoch: 4/20... Training Step: 12693... Training loss: 1.6930... 0.0493 sec/batch
Epoch: 4/20... Training Step: 12694... Training loss: 1.5927... 0.0443 sec/batch
Epoch: 4/20... Training Step: 12695... Training loss: 1.6278... 0.0413 sec/batch
Epoch: 4/20... Training Step: 12696... Training loss: 1.7967... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12697... Training loss: 1.6830... 0.0405 sec/batch
Epoch: 4/20... Training Step: 12698... Training loss: 1.5059... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12699... Training loss: 1.7707... 0.0395 sec/batch
Epoch: 4/20... Training Step: 12700... Training loss: 1.6897... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12701... Training loss: 1.6886... 0.0441 sec/batch
Epoch: 4/20... Training Step: 12702... Training loss: 1.5201... 0.0444 sec/batch
Epoch: 4/20... Training Step: 12703... Training loss: 1.7291... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12704... Training loss: 1.5966... 0.0398 sec/batch
Epoch: 4/20... Training Step: 12705... Training loss: 1.5901... 0.0411 sec/batch
Epoch: 4/20... Training Step: 12706... Training loss: 1.5288... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12707... Training loss: 1.6885... 0.0400 sec/batch
Epoch: 4/20... Training Step: 12708... Training loss: 1.5476... 0.0402 sec/batch
Epoch: 4/20... Training Step: 12709... Training loss: 1.7416... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12710... Training loss: 1.6585... 0.0404 sec/batch
Epoch: 4/20... Training Step: 12711... Training loss: 1.7038... 0.0387 sec/batch
Epoch: 4/20... Training Step: 12712... Training loss: 1.5874... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12713... Training loss: 1.7784... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12714... Training loss: 1.8814... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12715... Training loss: 1.7702... 0.0444 sec/batch
Epoch: 4/20... Training Step: 12716... Training loss: 1.7134... 0.0448 sec/batch
Epoch: 4/20... Training Step: 12717... Training loss: 1.7659... 0.0432 sec/batch
Epoch: 4/20... Training Step: 12718... Training loss: 1.6737... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12719... Training loss: 1.6088... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12720... Training loss: 1.6516... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12721... Training loss: 1.6517... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12722... Training loss: 1.6482... 0.0389 sec/batch
Epoch: 4/20... Training Step: 12723... Training loss: 1.6734... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12724... Training loss: 1.7005... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12725... Training loss: 1.6971... 0.0452 sec/batch
Epoch: 4/20... Training Step: 12726... Training loss: 1.6288... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12727... Training loss: 1.4877... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12728... Training loss: 1.5773... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12729... Training loss: 1.6281... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12730... Training loss: 1.6903... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12731... Training loss: 1.6737... 0.0453 sec/batch
Epoch: 4/20... Training Step: 12732... Training loss: 1.7179... 0.0462 sec/batch
Epoch: 4/20... Training Step: 12733... Training loss: 1.6443... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12734... Training loss: 1.6365... 0.0483 sec/batch
Epoch: 4/20... Training Step: 12735... Training loss: 1.5233... 0.0424 sec/batch
Epoch: 4/20... Training Step: 12736... Training loss: 1.6182... 0.0417 sec/batch
Epoch: 4/20... Training Step: 12737... Training loss: 1.6728... 0.0450 sec/batch
Epoch: 4/20... Training Step: 12738... Training loss: 1.6376... 0.0413 sec/batch
Epoch: 4/20... Training Step: 12739... Training loss: 1.6568... 0.0395 sec/batch
Epoch: 4/20... Training Step: 12740... Training loss: 1.5559... 0.0477 sec/batch
Epoch: 4/20... Training Step: 12741... Training loss: 1.7793... 0.0496 sec/batch
Epoch: 4/20... Training Step: 12742... Training loss: 1.5963... 0.0695 sec/batch
Epoch: 4/20... Training Step: 12743... Training loss: 1.6121... 0.0605 sec/batch
Epoch: 4/20... Training Step: 12744... Training loss: 1.6123... 0.0654 sec/batch
Epoch: 4/20... Training Step: 12745... Training loss: 1.7037... 0.0473 sec/batch
Epoch: 4/20... Training Step: 12746... Training loss: 1.5046... 0.0724 sec/batch
Epoch: 4/20... Training Step: 12747... Training loss: 1.6226... 0.0482 sec/batch
Epoch: 4/20... Training Step: 12748... Training loss: 1.5532... 0.0582 sec/batch
Epoch: 4/20... Training Step: 12749... Training loss: 1.7008... 0.0501 sec/batch
Epoch: 4/20... Training Step: 12750... Training loss: 1.5983... 0.0679 sec/batch
Epoch: 4/20... Training Step: 12751... Training loss: 1.6621... 0.0582 sec/batch
Epoch: 4/20... Training Step: 12752... Training loss: 1.6518... 0.0489 sec/batch
Epoch: 4/20... Training Step: 12753... Training loss: 1.6939... 0.0517 sec/batch
Epoch: 4/20... Training Step: 12754... Training loss: 1.7131... 0.0505 sec/batch
Epoch: 4/20... Training Step: 12755... Training loss: 1.6383... 0.0478 sec/batch
Epoch: 4/20... Training Step: 12756... Training loss: 1.6364... 0.0624 sec/batch
Epoch: 4/20... Training Step: 12757... Training loss: 1.7500... 0.0434 sec/batch
Epoch: 4/20... Training Step: 12758... Training loss: 1.6822... 0.0590 sec/batch
Epoch: 4/20... Training Step: 12759... Training loss: 1.6022... 0.0533 sec/batch
Epoch: 4/20... Training Step: 12760... Training loss: 1.7713... 0.0538 sec/batch
Epoch: 4/20... Training Step: 12761... Training loss: 1.7771... 0.0491 sec/batch
Epoch: 4/20... Training Step: 12762... Training loss: 1.4899... 0.0533 sec/batch
Epoch: 4/20... Training Step: 12763... Training loss: 1.6846... 0.0625 sec/batch
Epoch: 4/20... Training Step: 12764... Training loss: 1.5470... 0.0678 sec/batch
Epoch: 4/20... Training Step: 12765... Training loss: 1.7724... 0.0581 sec/batch
Epoch: 4/20... Training Step: 12766... Training loss: 1.6895... 0.0508 sec/batch
Epoch: 4/20... Training Step: 12767... Training loss: 1.7501... 0.0609 sec/batch
Epoch: 4/20... Training Step: 12768... Training loss: 1.6649... 0.0608 sec/batch
Epoch: 4/20... Training Step: 12769... Training loss: 1.6976... 0.0513 sec/batch
Epoch: 4/20... Training Step: 12770... Training loss: 1.7838... 0.0527 sec/batch
Epoch: 4/20... Training Step: 12771... Training loss: 1.6860... 0.0669 sec/batch
Epoch: 4/20... Training Step: 12772... Training loss: 1.7509... 0.0594 sec/batch
Epoch: 4/20... Training Step: 12773... Training loss: 1.8078... 0.0585 sec/batch
Epoch: 4/20... Training Step: 12774... Training loss: 1.6710... 0.0597 sec/batch
Epoch: 4/20... Training Step: 12775... Training loss: 1.7930... 0.0482 sec/batch
Epoch: 4/20... Training Step: 12776... Training loss: 1.9885... 0.0663 sec/batch
Epoch: 4/20... Training Step: 12777... Training loss: 1.7768... 0.0624 sec/batch
Epoch: 4/20... Training Step: 12778... Training loss: 1.7195... 0.0532 sec/batch
Epoch: 4/20... Training Step: 12779... Training loss: 1.7244... 0.0518 sec/batch
Epoch: 4/20... Training Step: 12780... Training loss: 1.7551... 0.0492 sec/batch
Epoch: 4/20... Training Step: 12781... Training loss: 1.6014... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12782... Training loss: 1.6217... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12783... Training loss: 1.7119... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12784... Training loss: 1.6536... 0.0567 sec/batch
Epoch: 4/20... Training Step: 12785... Training loss: 1.7745... 0.0629 sec/batch
Epoch: 4/20... Training Step: 12786... Training loss: 1.5960... 0.0880 sec/batch
Epoch: 4/20... Training Step: 12787... Training loss: 1.6920... 0.0823 sec/batch
Epoch: 4/20... Training Step: 12788... Training loss: 1.7860... 0.0861 sec/batch
Epoch: 4/20... Training Step: 12789... Training loss: 1.5990... 0.1191 sec/batch
Epoch: 4/20... Training Step: 12790... Training loss: 1.5680... 0.0725 sec/batch
Epoch: 4/20... Training Step: 12791... Training loss: 1.9317... 0.0894 sec/batch
Epoch: 4/20... Training Step: 12792... Training loss: 1.8901... 0.0559 sec/batch
Epoch: 4/20... Training Step: 12793... Training loss: 1.7349... 0.1052 sec/batch
Epoch: 4/20... Training Step: 12794... Training loss: 1.7449... 0.0844 sec/batch
Epoch: 4/20... Training Step: 12795... Training loss: 1.5523... 0.0538 sec/batch
Epoch: 4/20... Training Step: 12796... Training loss: 1.6341... 0.0528 sec/batch
Epoch: 4/20... Training Step: 12797... Training loss: 1.7185... 0.0523 sec/batch
Epoch: 4/20... Training Step: 12798... Training loss: 1.6140... 0.0529 sec/batch
Epoch: 4/20... Training Step: 12799... Training loss: 1.5712... 0.0591 sec/batch
Epoch: 4/20... Training Step: 12800... Training loss: 1.7099... 0.0525 sec/batch
Epoch: 4/20... Training Step: 12801... Training loss: 1.6700... 0.0642 sec/batch
Epoch: 4/20... Training Step: 12802... Training loss: 1.7022... 0.0588 sec/batch
Epoch: 4/20... Training Step: 12803... Training loss: 1.8202... 0.0483 sec/batch
Epoch: 4/20... Training Step: 12804... Training loss: 1.7200... 0.0724 sec/batch
Epoch: 4/20... Training Step: 12805... Training loss: 1.7459... 0.0636 sec/batch
Epoch: 4/20... Training Step: 12806... Training loss: 1.6287... 0.0656 sec/batch
Epoch: 4/20... Training Step: 12807... Training loss: 1.8178... 0.0795 sec/batch
Epoch: 4/20... Training Step: 12808... Training loss: 1.5868... 0.0550 sec/batch
Epoch: 4/20... Training Step: 12809... Training loss: 1.6805... 0.0609 sec/batch
Epoch: 4/20... Training Step: 12810... Training loss: 1.7198... 0.0570 sec/batch
Epoch: 4/20... Training Step: 12811... Training loss: 1.5047... 0.0449 sec/batch
Epoch: 4/20... Training Step: 12812... Training loss: 1.5580... 0.0538 sec/batch
Epoch: 4/20... Training Step: 12813... Training loss: 1.6160... 0.0563 sec/batch
Epoch: 4/20... Training Step: 12814... Training loss: 1.6113... 0.0464 sec/batch
Epoch: 4/20... Training Step: 12815... Training loss: 1.6303... 0.0707 sec/batch
Epoch: 4/20... Training Step: 12816... Training loss: 1.7806... 0.0689 sec/batch
Epoch: 4/20... Training Step: 12817... Training loss: 1.7316... 0.0566 sec/batch
Epoch: 4/20... Training Step: 12818... Training loss: 1.7346... 0.0585 sec/batch
Epoch: 4/20... Training Step: 12819... Training loss: 1.8278... 0.0586 sec/batch
Epoch: 4/20... Training Step: 12820... Training loss: 1.7389... 0.0537 sec/batch
Epoch: 4/20... Training Step: 12821... Training loss: 1.6671... 0.0672 sec/batch
Epoch: 4/20... Training Step: 12822... Training loss: 1.6964... 0.0674 sec/batch
Epoch: 4/20... Training Step: 12823... Training loss: 1.8464... 0.0555 sec/batch
Epoch: 4/20... Training Step: 12824... Training loss: 1.6629... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12825... Training loss: 1.6526... 0.0462 sec/batch
Epoch: 4/20... Training Step: 12826... Training loss: 1.6834... 0.0581 sec/batch
Epoch: 4/20... Training Step: 12827... Training loss: 1.6721... 0.0591 sec/batch
Epoch: 4/20... Training Step: 12828... Training loss: 1.5118... 0.0549 sec/batch
Epoch: 4/20... Training Step: 12829... Training loss: 1.5934... 0.0598 sec/batch
Epoch: 4/20... Training Step: 12830... Training loss: 1.6365... 0.0507 sec/batch
Epoch: 4/20... Training Step: 12831... Training loss: 1.5356... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12832... Training loss: 1.7344... 0.0616 sec/batch
Epoch: 4/20... Training Step: 12833... Training loss: 1.7405... 0.0467 sec/batch
Epoch: 4/20... Training Step: 12834... Training loss: 1.7081... 0.0650 sec/batch
Epoch: 4/20... Training Step: 12835... Training loss: 1.6847... 0.0671 sec/batch
Epoch: 4/20... Training Step: 12836... Training loss: 1.7319... 0.0527 sec/batch
Epoch: 4/20... Training Step: 12837... Training loss: 1.6702... 0.0539 sec/batch
Epoch: 4/20... Training Step: 12838... Training loss: 1.5809... 0.0644 sec/batch
Epoch: 4/20... Training Step: 12839... Training loss: 1.5455... 0.0480 sec/batch
Epoch: 4/20... Training Step: 12840... Training loss: 1.7093... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12841... Training loss: 1.7200... 0.0518 sec/batch
Epoch: 4/20... Training Step: 12842... Training loss: 1.5849... 0.0682 sec/batch
Epoch: 4/20... Training Step: 12843... Training loss: 1.6559... 0.0599 sec/batch
Epoch: 4/20... Training Step: 12844... Training loss: 1.6351... 0.0565 sec/batch
Epoch: 4/20... Training Step: 12845... Training loss: 1.6296... 0.0456 sec/batch
Epoch: 4/20... Training Step: 12846... Training loss: 1.6777... 0.0513 sec/batch
Epoch: 4/20... Training Step: 12847... Training loss: 1.6247... 0.0432 sec/batch
Epoch: 4/20... Training Step: 12848... Training loss: 1.6159... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12849... Training loss: 1.5591... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12850... Training loss: 1.6216... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12851... Training loss: 1.5890... 0.0398 sec/batch
Epoch: 4/20... Training Step: 12852... Training loss: 1.5554... 0.0405 sec/batch
Epoch: 4/20... Training Step: 12853... Training loss: 1.6582... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12854... Training loss: 1.7444... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12855... Training loss: 1.6863... 0.0394 sec/batch
Epoch: 4/20... Training Step: 12856... Training loss: 1.5715... 0.0396 sec/batch
Epoch: 4/20... Training Step: 12857... Training loss: 1.4899... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12858... Training loss: 1.6836... 0.0422 sec/batch
Epoch: 4/20... Training Step: 12859... Training loss: 1.6696... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12860... Training loss: 1.7138... 0.0401 sec/batch
Epoch: 4/20... Training Step: 12861... Training loss: 1.6516... 0.0534 sec/batch
Epoch: 4/20... Training Step: 12862... Training loss: 1.7206... 0.0518 sec/batch
Epoch: 4/20... Training Step: 12863... Training loss: 1.7278... 0.0510 sec/batch
Epoch: 4/20... Training Step: 12864... Training loss: 1.6898... 0.0409 sec/batch
Epoch: 4/20... Training Step: 12865... Training loss: 1.6936... 0.0407 sec/batch
Epoch: 4/20... Training Step: 12866... Training loss: 1.5777... 0.0491 sec/batch
Epoch: 4/20... Training Step: 12867... Training loss: 1.7941... 0.0402 sec/batch
Epoch: 4/20... Training Step: 12868... Training loss: 1.7230... 0.0437 sec/batch
Epoch: 4/20... Training Step: 12869... Training loss: 1.5973... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12870... Training loss: 1.8167... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12871... Training loss: 1.5202... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12872... Training loss: 1.6097... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12873... Training loss: 1.5546... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12874... Training loss: 1.5288... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12875... Training loss: 1.7118... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12876... Training loss: 1.7004... 0.0422 sec/batch
Epoch: 4/20... Training Step: 12877... Training loss: 1.5986... 0.0455 sec/batch
Epoch: 4/20... Training Step: 12878... Training loss: 1.6179... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12879... Training loss: 1.6640... 0.0615 sec/batch
Epoch: 4/20... Training Step: 12880... Training loss: 1.5457... 0.0583 sec/batch
Epoch: 4/20... Training Step: 12881... Training loss: 1.5887... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12882... Training loss: 1.7619... 0.0611 sec/batch
Epoch: 4/20... Training Step: 12883... Training loss: 1.6425... 0.0429 sec/batch
Epoch: 4/20... Training Step: 12884... Training loss: 1.5820... 0.0406 sec/batch
Epoch: 4/20... Training Step: 12885... Training loss: 1.5850... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12886... Training loss: 1.7018... 0.0397 sec/batch
Epoch: 4/20... Training Step: 12887... Training loss: 1.8008... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12888... Training loss: 1.6385... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12889... Training loss: 1.7457... 0.0412 sec/batch
Epoch: 4/20... Training Step: 12890... Training loss: 1.5687... 0.0398 sec/batch
Epoch: 4/20... Training Step: 12891... Training loss: 1.7573... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12892... Training loss: 1.5704... 0.0394 sec/batch
Epoch: 4/20... Training Step: 12893... Training loss: 1.8272... 0.0504 sec/batch
Epoch: 4/20... Training Step: 12894... Training loss: 1.6836... 0.0497 sec/batch
Epoch: 4/20... Training Step: 12895... Training loss: 1.5859... 0.0594 sec/batch
Epoch: 4/20... Training Step: 12896... Training loss: 1.6424... 0.0608 sec/batch
Epoch: 4/20... Training Step: 12897... Training loss: 1.5704... 0.0569 sec/batch
Epoch: 4/20... Training Step: 12898... Training loss: 1.6096... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12899... Training loss: 1.6264... 0.0498 sec/batch
Epoch: 4/20... Training Step: 12900... Training loss: 1.6214... 0.0407 sec/batch
Epoch: 4/20... Training Step: 12901... Training loss: 1.4912... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12902... Training loss: 1.7031... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12903... Training loss: 1.7263... 0.0390 sec/batch
Epoch: 4/20... Training Step: 12904... Training loss: 1.7165... 0.0401 sec/batch
Epoch: 4/20... Training Step: 12905... Training loss: 1.6635... 0.0513 sec/batch
Epoch: 4/20... Training Step: 12906... Training loss: 1.7478... 0.0706 sec/batch
Epoch: 4/20... Training Step: 12907... Training loss: 1.6191... 0.0525 sec/batch
Epoch: 4/20... Training Step: 12908... Training loss: 1.7686... 0.0752 sec/batch
Epoch: 4/20... Training Step: 12909... Training loss: 1.6161... 0.0599 sec/batch
Epoch: 4/20... Training Step: 12910... Training loss: 1.6859... 0.0449 sec/batch
Epoch: 4/20... Training Step: 12911... Training loss: 1.7154... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12912... Training loss: 1.5477... 0.0535 sec/batch
Epoch: 4/20... Training Step: 12913... Training loss: 1.6425... 0.0605 sec/batch
Epoch: 4/20... Training Step: 12914... Training loss: 1.5161... 0.0541 sec/batch
Epoch: 4/20... Training Step: 12915... Training loss: 1.5806... 0.0489 sec/batch
Epoch: 4/20... Training Step: 12916... Training loss: 1.5615... 0.0644 sec/batch
Epoch: 4/20... Training Step: 12917... Training loss: 1.6626... 0.0527 sec/batch
Epoch: 4/20... Training Step: 12918... Training loss: 1.7218... 0.0782 sec/batch
Epoch: 4/20... Training Step: 12919... Training loss: 1.6997... 0.0760 sec/batch
Epoch: 4/20... Training Step: 12920... Training loss: 1.8095... 0.0949 sec/batch
Epoch: 4/20... Training Step: 12921... Training loss: 1.6279... 0.0568 sec/batch
Epoch: 4/20... Training Step: 12922... Training loss: 1.6708... 0.0769 sec/batch
Epoch: 4/20... Training Step: 12923... Training loss: 1.6712... 0.0633 sec/batch
Epoch: 4/20... Training Step: 12924... Training loss: 1.7678... 0.0540 sec/batch
Epoch: 4/20... Training Step: 12925... Training loss: 1.5203... 0.0623 sec/batch
Epoch: 4/20... Training Step: 12926... Training loss: 1.7788... 0.0653 sec/batch
Epoch: 4/20... Training Step: 12927... Training loss: 1.7234... 0.0638 sec/batch
Epoch: 4/20... Training Step: 12928... Training loss: 1.6511... 0.0578 sec/batch
Epoch: 4/20... Training Step: 12929... Training loss: 1.6216... 0.0647 sec/batch
Epoch: 4/20... Training Step: 12930... Training loss: 1.4863... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12931... Training loss: 1.5559... 0.0588 sec/batch
Epoch: 4/20... Training Step: 12932... Training loss: 1.6643... 0.0632 sec/batch
Epoch: 4/20... Training Step: 12933... Training loss: 1.5219... 0.0552 sec/batch
Epoch: 4/20... Training Step: 12934... Training loss: 1.5980... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12935... Training loss: 1.6451... 0.0526 sec/batch
Epoch: 4/20... Training Step: 12936... Training loss: 1.7001... 0.0551 sec/batch
Epoch: 4/20... Training Step: 12937... Training loss: 1.6934... 0.0578 sec/batch
Epoch: 4/20... Training Step: 12938... Training loss: 1.6548... 0.0559 sec/batch
Epoch: 4/20... Training Step: 12939... Training loss: 1.7245... 0.0598 sec/batch
Epoch: 4/20... Training Step: 12940... Training loss: 1.6190... 0.0447 sec/batch
Epoch: 4/20... Training Step: 12941... Training loss: 1.5857... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12942... Training loss: 1.6545... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12943... Training loss: 1.6462... 0.0479 sec/batch
Epoch: 4/20... Training Step: 12944... Training loss: 1.6618... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12945... Training loss: 1.7711... 0.0459 sec/batch
Epoch: 4/20... Training Step: 12946... Training loss: 1.8352... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12947... Training loss: 1.5997... 0.0439 sec/batch
Epoch: 4/20... Training Step: 12948... Training loss: 1.6334... 0.0440 sec/batch
Epoch: 4/20... Training Step: 12949... Training loss: 1.6684... 0.0451 sec/batch
Epoch: 4/20... Training Step: 12950... Training loss: 1.7075... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12951... Training loss: 1.6313... 0.0418 sec/batch
Epoch: 4/20... Training Step: 12952... Training loss: 1.5930... 0.0414 sec/batch
Epoch: 4/20... Training Step: 12953... Training loss: 1.6939... 0.0505 sec/batch
Epoch: 4/20... Training Step: 12954... Training loss: 1.7091... 0.0579 sec/batch
Epoch: 4/20... Training Step: 12955... Training loss: 1.6849... 0.0464 sec/batch
Epoch: 4/20... Training Step: 12956... Training loss: 1.6785... 0.0483 sec/batch
Epoch: 4/20... Training Step: 12957... Training loss: 1.5378... 0.0530 sec/batch
Epoch: 4/20... Training Step: 12958... Training loss: 1.5792... 0.0478 sec/batch
Epoch: 4/20... Training Step: 12959... Training loss: 1.5760... 0.0459 sec/batch
Epoch: 4/20... Training Step: 12960... Training loss: 1.7072... 0.0572 sec/batch
Epoch: 4/20... Training Step: 12961... Training loss: 1.6750... 0.0506 sec/batch
Epoch: 4/20... Training Step: 12962... Training loss: 1.6718... 0.0530 sec/batch
Epoch: 4/20... Training Step: 12963... Training loss: 1.5697... 0.0452 sec/batch
Epoch: 4/20... Training Step: 12964... Training loss: 1.6215... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12965... Training loss: 1.6471... 0.0476 sec/batch
Epoch: 4/20... Training Step: 12966... Training loss: 1.5713... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12967... Training loss: 1.5187... 0.0416 sec/batch
Epoch: 4/20... Training Step: 12968... Training loss: 1.6541... 0.0516 sec/batch
Epoch: 4/20... Training Step: 12969... Training loss: 1.6861... 0.0427 sec/batch
Epoch: 4/20... Training Step: 12970... Training loss: 1.6756... 0.0467 sec/batch
Epoch: 4/20... Training Step: 12971... Training loss: 1.7156... 0.0433 sec/batch
Epoch: 4/20... Training Step: 12972... Training loss: 1.6635... 0.0445 sec/batch
Epoch: 4/20... Training Step: 12973... Training loss: 1.6624... 0.0493 sec/batch
Epoch: 4/20... Training Step: 12974... Training loss: 1.6165... 0.0446 sec/batch
Epoch: 4/20... Training Step: 12975... Training loss: 1.6109... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12976... Training loss: 1.7299... 0.0472 sec/batch
Epoch: 4/20... Training Step: 12977... Training loss: 1.5973... 0.0442 sec/batch
Epoch: 4/20... Training Step: 12978... Training loss: 1.7698... 0.0421 sec/batch
Epoch: 4/20... Training Step: 12979... Training loss: 1.5771... 0.0424 sec/batch
Epoch: 4/20... Training Step: 12980... Training loss: 1.6560... 0.0413 sec/batch
Epoch: 4/20... Training Step: 12981... Training loss: 1.6562... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12982... Training loss: 1.5228... 0.0435 sec/batch
Epoch: 4/20... Training Step: 12983... Training loss: 1.4746... 0.0410 sec/batch
Epoch: 4/20... Training Step: 12984... Training loss: 1.5525... 0.0424 sec/batch
Epoch: 4/20... Training Step: 12985... Training loss: 1.6220... 0.0428 sec/batch
Epoch: 4/20... Training Step: 12986... Training loss: 1.5498... 0.0463 sec/batch
Epoch: 4/20... Training Step: 12987... Training loss: 1.6557... 0.0425 sec/batch
Epoch: 4/20... Training Step: 12988... Training loss: 1.7375... 0.0423 sec/batch
Epoch: 4/20... Training Step: 12989... Training loss: 1.6280... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12990... Training loss: 1.5110... 0.0432 sec/batch
Epoch: 4/20... Training Step: 12991... Training loss: 1.6383... 0.0403 sec/batch
Epoch: 4/20... Training Step: 12992... Training loss: 1.6219... 0.0407 sec/batch
Epoch: 4/20... Training Step: 12993... Training loss: 1.8046... 0.0419 sec/batch
Epoch: 4/20... Training Step: 12994... Training loss: 1.4026... 0.0404 sec/batch
Epoch: 4/20... Training Step: 12995... Training loss: 1.5517... 0.0420 sec/batch
Epoch: 4/20... Training Step: 12996... Training loss: 1.6388... 0.0495 sec/batch
Epoch: 4/20... Training Step: 12997... Training loss: 1.6490... 0.0614 sec/batch
Epoch: 4/20... Training Step: 12998... Training loss: 1.4539... 0.0499 sec/batch
Epoch: 4/20... Training Step: 12999... Training loss: 1.6341... 0.0476 sec/batch
Epoch: 4/20... Training Step: 13000... Training loss: 1.8220... 0.0425 sec/batch
Epoch: 4/20... Training Step: 13001... Training loss: 1.7084... 0.0408 sec/batch
Epoch: 4/20... Training Step: 13002... Training loss: 1.7842... 0.0437 sec/batch
Epoch: 4/20... Training Step: 13003... Training loss: 1.6368... 0.0396 sec/batch
Epoch: 4/20... Training Step: 13004... Training loss: 1.6540... 0.0463 sec/batch
Epoch: 4/20... Training Step: 13005... Training loss: 1.5689... 0.0395 sec/batch
Epoch: 4/20... Training Step: 13006... Training loss: 1.6935... 0.0410 sec/batch
Epoch: 4/20... Training Step: 13007... Training loss: 1.5518... 0.0400 sec/batch
Epoch: 4/20... Training Step: 13008... Training loss: 1.4687... 0.0398 sec/batch
Epoch: 4/20... Training Step: 13009... Training loss: 1.7718... 0.0402 sec/batch
Epoch: 4/20... Training Step: 13010... Training loss: 1.6523... 0.0444 sec/batch
Epoch: 4/20... Training Step: 13011... Training loss: 1.5511... 0.0399 sec/batch
Epoch: 4/20... Training Step: 13012... Training loss: 1.5988... 0.0420 sec/batch
Epoch: 4/20... Training Step: 13013... Training loss: 1.5848... 0.0406 sec/batch
Epoch: 4/20... Training Step: 13014... Training loss: 1.7336... 0.0431 sec/batch
Epoch: 4/20... Training Step: 13015... Training loss: 1.6402... 0.0424 sec/batch
Epoch: 4/20... Training Step: 13016... Training loss: 1.5896... 0.0406 sec/batch
Epoch: 4/20... Training Step: 13017... Training loss: 1.7012... 0.0446 sec/batch
Epoch: 4/20... Training Step: 13018... Training loss: 1.7055... 0.0396 sec/batch
Epoch: 4/20... Training Step: 13019... Training loss: 1.5559... 0.0393 sec/batch
Epoch: 4/20... Training Step: 13020... Training loss: 1.5905... 0.0435 sec/batch
Epoch: 4/20... Training Step: 13021... Training loss: 1.6567... 0.0393 sec/batch
Epoch: 4/20... Training Step: 13022... Training loss: 1.6614... 0.0389 sec/batch
Epoch: 4/20... Training Step: 13023... Training loss: 1.5125... 0.0501 sec/batch
Epoch: 4/20... Training Step: 13024... Training loss: 1.6835... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13025... Training loss: 1.7401... 0.0569 sec/batch
Epoch: 4/20... Training Step: 13026... Training loss: 1.7999... 0.0513 sec/batch
Epoch: 4/20... Training Step: 13027... Training loss: 1.6467... 0.0450 sec/batch
Epoch: 4/20... Training Step: 13028... Training loss: 1.6964... 0.0407 sec/batch
Epoch: 4/20... Training Step: 13029... Training loss: 1.5877... 0.0413 sec/batch
Epoch: 4/20... Training Step: 13030... Training loss: 1.4867... 0.0430 sec/batch
Epoch: 4/20... Training Step: 13031... Training loss: 1.7154... 0.0419 sec/batch
Epoch: 4/20... Training Step: 13032... Training loss: 1.7033... 0.0411 sec/batch
Epoch: 4/20... Training Step: 13033... Training loss: 1.5927... 0.0392 sec/batch
Epoch: 4/20... Training Step: 13034... Training loss: 1.6070... 0.0422 sec/batch
Epoch: 4/20... Training Step: 13035... Training loss: 1.6597... 0.0414 sec/batch
Epoch: 4/20... Training Step: 13036... Training loss: 1.6403... 0.0496 sec/batch
Epoch: 4/20... Training Step: 13037... Training loss: 1.7046... 0.0414 sec/batch
Epoch: 4/20... Training Step: 13038... Training loss: 1.5363... 0.0476 sec/batch
Epoch: 4/20... Training Step: 13039... Training loss: 1.5276... 0.0518 sec/batch
Epoch: 4/20... Training Step: 13040... Training loss: 1.5823... 0.0569 sec/batch
Epoch: 4/20... Training Step: 13041... Training loss: 1.7041... 0.0546 sec/batch
Epoch: 4/20... Training Step: 13042... Training loss: 1.4958... 0.0402 sec/batch
Epoch: 4/20... Training Step: 13043... Training loss: 1.6082... 0.0420 sec/batch
Epoch: 4/20... Training Step: 13044... Training loss: 1.7516... 0.0420 sec/batch
Epoch: 4/20... Training Step: 13045... Training loss: 1.5149... 0.0410 sec/batch
Epoch: 4/20... Training Step: 13046... Training loss: 1.7361... 0.0421 sec/batch
Epoch: 4/20... Training Step: 13047... Training loss: 1.7066... 0.0405 sec/batch
Epoch: 4/20... Training Step: 13048... Training loss: 1.7123... 0.0400 sec/batch
Epoch: 4/20... Training Step: 13049... Training loss: 1.6683... 0.0399 sec/batch
Epoch: 4/20... Training Step: 13050... Training loss: 1.8444... 0.0585 sec/batch
Epoch: 4/20... Training Step: 13051... Training loss: 1.6307... 0.0583 sec/batch
Epoch: 4/20... Training Step: 13052... Training loss: 1.6891... 0.0556 sec/batch
Epoch: 4/20... Training Step: 13053... Training loss: 1.4898... 0.0733 sec/batch
Epoch: 4/20... Training Step: 13054... Training loss: 1.9178... 0.0546 sec/batch
Epoch: 4/20... Training Step: 13055... Training loss: 1.8568... 0.0626 sec/batch
Epoch: 4/20... Training Step: 13056... Training loss: 1.6625... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13057... Training loss: 1.8533... 0.0503 sec/batch
Epoch: 4/20... Training Step: 13058... Training loss: 1.6929... 0.0501 sec/batch
Epoch: 4/20... Training Step: 13059... Training loss: 1.7903... 0.0583 sec/batch
Epoch: 4/20... Training Step: 13060... Training loss: 1.6358... 0.0564 sec/batch
Epoch: 4/20... Training Step: 13061... Training loss: 1.6400... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13062... Training loss: 1.7087... 0.0572 sec/batch
Epoch: 4/20... Training Step: 13063... Training loss: 1.6937... 0.0494 sec/batch
Epoch: 4/20... Training Step: 13064... Training loss: 1.7748... 0.0574 sec/batch
Epoch: 4/20... Training Step: 13065... Training loss: 1.7128... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13066... Training loss: 1.5598... 0.0573 sec/batch
Epoch: 4/20... Training Step: 13067... Training loss: 1.6308... 0.0543 sec/batch
Epoch: 4/20... Training Step: 13068... Training loss: 1.5791... 0.0500 sec/batch
Epoch: 4/20... Training Step: 13069... Training loss: 1.6882... 0.0450 sec/batch
Epoch: 4/20... Training Step: 13070... Training loss: 1.5138... 0.0444 sec/batch
Epoch: 4/20... Training Step: 13071... Training loss: 1.6412... 0.0419 sec/batch
Epoch: 4/20... Training Step: 13072... Training loss: 1.7004... 0.0405 sec/batch
Epoch: 4/20... Training Step: 13073... Training loss: 1.6843... 0.0410 sec/batch
Epoch: 4/20... Training Step: 13074... Training loss: 1.6978... 0.0402 sec/batch
Epoch: 4/20... Training Step: 13075... Training loss: 1.6711... 0.0422 sec/batch
Epoch: 4/20... Training Step: 13076... Training loss: 1.7767... 0.0434 sec/batch
Epoch: 4/20... Training Step: 13077... Training loss: 1.6419... 0.0422 sec/batch
Epoch: 4/20... Training Step: 13078... Training loss: 1.7529... 0.0423 sec/batch
Epoch: 4/20... Training Step: 13079... Training loss: 1.6685... 0.0483 sec/batch
Epoch: 4/20... Training Step: 13080... Training loss: 1.5865... 0.0458 sec/batch
Epoch: 4/20... Training Step: 13081... Training loss: 1.4874... 0.0411 sec/batch
Epoch: 4/20... Training Step: 13082... Training loss: 1.7287... 0.0452 sec/batch
Epoch: 4/20... Training Step: 13083... Training loss: 1.5885... 0.0437 sec/batch
Epoch: 4/20... Training Step: 13084... Training loss: 1.5790... 0.0434 sec/batch
Epoch: 4/20... Training Step: 13085... Training loss: 1.6745... 0.0408 sec/batch
Epoch: 4/20... Training Step: 13086... Training loss: 1.5870... 0.0442 sec/batch
Epoch: 4/20... Training Step: 13087... Training loss: 1.6235... 0.0419 sec/batch
Epoch: 4/20... Training Step: 13088... Training loss: 1.4926... 0.0403 sec/batch
Epoch: 4/20... Training Step: 13089... Training loss: 1.4550... 0.0437 sec/batch
Epoch: 4/20... Training Step: 13090... Training loss: 1.6932... 0.0460 sec/batch
Epoch: 4/20... Training Step: 13091... Training loss: 1.7558... 0.0432 sec/batch
Epoch: 4/20... Training Step: 13092... Training loss: 1.6418... 0.0412 sec/batch
Epoch: 4/20... Training Step: 13093... Training loss: 1.5939... 0.0503 sec/batch
Epoch: 4/20... Training Step: 13094... Training loss: 1.7329... 0.0509 sec/batch
Epoch: 4/20... Training Step: 13095... Training loss: 1.6367... 0.0401 sec/batch
Epoch: 4/20... Training Step: 13096... Training loss: 1.6261... 0.0390 sec/batch
Epoch: 4/20... Training Step: 13097... Training loss: 1.7433... 0.0432 sec/batch
Epoch: 4/20... Training Step: 13098... Training loss: 1.8398... 0.0455 sec/batch
Epoch: 4/20... Training Step: 13099... Training loss: 1.5499... 0.0452 sec/batch
Epoch: 4/20... Training Step: 13100... Training loss: 1.6482... 0.0422 sec/batch
Epoch: 4/20... Training Step: 13101... Training loss: 1.6145... 0.0405 sec/batch
Epoch: 4/20... Training Step: 13102... Training loss: 1.7123... 0.0428 sec/batch
Epoch: 4/20... Training Step: 13103... Training loss: 1.6257... 0.0403 sec/batch
Epoch: 4/20... Training Step: 13104... Training loss: 1.6433... 0.0419 sec/batch
Epoch: 4/20... Training Step: 13105... Training loss: 1.8128... 0.0425 sec/batch
Epoch: 4/20... Training Step: 13106... Training loss: 1.7152... 0.0465 sec/batch
Epoch: 4/20... Training Step: 13107... Training loss: 1.5191... 0.0556 sec/batch
Epoch: 4/20... Training Step: 13108... Training loss: 1.6261... 0.0526 sec/batch
Epoch: 4/20... Training Step: 13109... Training loss: 1.6171... 0.0493 sec/batch
Epoch: 4/20... Training Step: 13110... Training loss: 1.6159... 0.0425 sec/batch
Epoch: 4/20... Training Step: 13111... Training loss: 1.5840... 0.0420 sec/batch
Epoch: 4/20... Training Step: 13112... Training loss: 1.7373... 0.0417 sec/batch
Epoch: 4/20... Training Step: 13113... Training loss: 1.6607... 0.0438 sec/batch
Epoch: 4/20... Training Step: 13114... Training loss: 1.6271... 0.0441 sec/batch
Epoch: 4/20... Training Step: 13115... Training loss: 1.6601... 0.0404 sec/batch
Epoch: 4/20... Training Step: 13116... Training loss: 1.6208... 0.0417 sec/batch
Epoch: 4/20... Training Step: 13117... Training loss: 1.5610... 0.0398 sec/batch
Epoch: 4/20... Training Step: 13118... Training loss: 1.6794... 0.0411 sec/batch
Epoch: 4/20... Training Step: 13119... Training loss: 1.7195... 0.0568 sec/batch
Epoch: 4/20... Training Step: 13120... Training loss: 1.7097... 0.0455 sec/batch
Epoch: 4/20... Training Step: 13121... Training loss: 1.6466... 0.0514 sec/batch
Epoch: 4/20... Training Step: 13122... Training loss: 1.7158... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13123... Training loss: 1.9055... 0.0412 sec/batch
Epoch: 4/20... Training Step: 13124... Training loss: 1.6029... 0.0473 sec/batch
Epoch: 4/20... Training Step: 13125... Training loss: 1.5659... 0.0523 sec/batch
Epoch: 4/20... Training Step: 13126... Training loss: 1.8270... 0.0444 sec/batch
Epoch: 4/20... Training Step: 13127... Training loss: 1.8596... 0.0451 sec/batch
Epoch: 4/20... Training Step: 13128... Training loss: 1.7655... 0.0505 sec/batch
Epoch: 4/20... Training Step: 13129... Training loss: 1.6429... 0.0697 sec/batch
Epoch: 4/20... Training Step: 13130... Training loss: 1.5722... 0.0558 sec/batch
Epoch: 4/20... Training Step: 13131... Training loss: 1.6857... 0.0796 sec/batch
Epoch: 4/20... Training Step: 13132... Training loss: 1.7033... 0.0934 sec/batch
Epoch: 4/20... Training Step: 13133... Training loss: 1.5148... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13134... Training loss: 1.5976... 0.0688 sec/batch
Epoch: 4/20... Training Step: 13135... Training loss: 1.7220... 0.0700 sec/batch
Epoch: 4/20... Training Step: 13136... Training loss: 1.6051... 0.0697 sec/batch
Epoch: 4/20... Training Step: 13137... Training loss: 1.8105... 0.0616 sec/batch
Epoch: 4/20... Training Step: 13138... Training loss: 1.7124... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13139... Training loss: 1.5785... 0.0622 sec/batch
Epoch: 4/20... Training Step: 13140... Training loss: 1.7625... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13141... Training loss: 1.4965... 0.0663 sec/batch
Epoch: 4/20... Training Step: 13142... Training loss: 1.7560... 0.0553 sec/batch
Epoch: 4/20... Training Step: 13143... Training loss: 1.6690... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13144... Training loss: 1.7585... 0.0628 sec/batch
Epoch: 4/20... Training Step: 13145... Training loss: 1.5747... 0.0526 sec/batch
Epoch: 4/20... Training Step: 13146... Training loss: 1.6166... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13147... Training loss: 1.7155... 0.0478 sec/batch
Epoch: 4/20... Training Step: 13148... Training loss: 1.4579... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13149... Training loss: 1.5487... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13150... Training loss: 1.5658... 0.0554 sec/batch
Epoch: 4/20... Training Step: 13151... Training loss: 1.5053... 0.0664 sec/batch
Epoch: 4/20... Training Step: 13152... Training loss: 1.6011... 0.0574 sec/batch
Epoch: 4/20... Training Step: 13153... Training loss: 1.5699... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13154... Training loss: 1.7540... 0.0545 sec/batch
Epoch: 4/20... Training Step: 13155... Training loss: 1.5491... 0.0689 sec/batch
Epoch: 4/20... Training Step: 13156... Training loss: 1.6558... 0.0576 sec/batch
Epoch: 4/20... Training Step: 13157... Training loss: 1.6810... 0.0612 sec/batch
Epoch: 4/20... Training Step: 13158... Training loss: 1.6423... 0.0627 sec/batch
Epoch: 4/20... Training Step: 13159... Training loss: 1.6794... 0.0648 sec/batch
Epoch: 4/20... Training Step: 13160... Training loss: 1.5548... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13161... Training loss: 1.7268... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13162... Training loss: 1.6086... 0.0582 sec/batch
Epoch: 4/20... Training Step: 13163... Training loss: 1.5404... 0.0622 sec/batch
Epoch: 4/20... Training Step: 13164... Training loss: 1.7740... 0.0682 sec/batch
Epoch: 4/20... Training Step: 13165... Training loss: 1.5855... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13166... Training loss: 1.6818... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13167... Training loss: 1.7737... 0.0691 sec/batch
Epoch: 4/20... Training Step: 13168... Training loss: 1.5979... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13169... Training loss: 1.6590... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13170... Training loss: 1.6841... 0.0517 sec/batch
Epoch: 4/20... Training Step: 13171... Training loss: 1.7999... 0.0662 sec/batch
Epoch: 4/20... Training Step: 13172... Training loss: 1.7061... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13173... Training loss: 1.6156... 0.0584 sec/batch
Epoch: 4/20... Training Step: 13174... Training loss: 1.5062... 0.0533 sec/batch
Epoch: 4/20... Training Step: 13175... Training loss: 1.5878... 0.0747 sec/batch
Epoch: 4/20... Training Step: 13176... Training loss: 1.6474... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13177... Training loss: 1.7398... 0.0819 sec/batch
Epoch: 4/20... Training Step: 13178... Training loss: 1.6334... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13179... Training loss: 1.6281... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13180... Training loss: 1.6764... 0.0531 sec/batch
Epoch: 4/20... Training Step: 13181... Training loss: 1.5770... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13182... Training loss: 1.6954... 0.0604 sec/batch
Epoch: 4/20... Training Step: 13183... Training loss: 1.7410... 0.0674 sec/batch
Epoch: 4/20... Training Step: 13184... Training loss: 1.5693... 0.0524 sec/batch
Epoch: 4/20... Training Step: 13185... Training loss: 1.7012... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13186... Training loss: 1.6131... 0.0625 sec/batch
Epoch: 4/20... Training Step: 13187... Training loss: 1.6987... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13188... Training loss: 1.7672... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13189... Training loss: 1.7384... 0.0604 sec/batch
Epoch: 4/20... Training Step: 13190... Training loss: 1.5539... 0.0627 sec/batch
Epoch: 4/20... Training Step: 13191... Training loss: 1.6955... 0.0593 sec/batch
Epoch: 4/20... Training Step: 13192... Training loss: 1.8394... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13193... Training loss: 1.6509... 0.0571 sec/batch
Epoch: 4/20... Training Step: 13194... Training loss: 1.7869... 0.0694 sec/batch
Epoch: 4/20... Training Step: 13195... Training loss: 1.8411... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13196... Training loss: 1.6101... 0.0584 sec/batch
Epoch: 4/20... Training Step: 13197... Training loss: 1.7412... 0.0549 sec/batch
Epoch: 4/20... Training Step: 13198... Training loss: 1.6426... 0.0592 sec/batch
Epoch: 4/20... Training Step: 13199... Training loss: 1.6703... 0.0638 sec/batch
Epoch: 4/20... Training Step: 13200... Training loss: 1.8512... 0.0740 sec/batch
Epoch: 4/20... Training Step: 13201... Training loss: 1.7960... 0.0688 sec/batch
Epoch: 4/20... Training Step: 13202... Training loss: 1.8203... 0.0772 sec/batch
Epoch: 4/20... Training Step: 13203... Training loss: 1.6902... 0.0527 sec/batch
Epoch: 4/20... Training Step: 13204... Training loss: 1.7091... 0.0568 sec/batch
Epoch: 4/20... Training Step: 13205... Training loss: 1.6899... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13206... Training loss: 1.7606... 0.0557 sec/batch
Epoch: 4/20... Training Step: 13207... Training loss: 1.5829... 0.0635 sec/batch
Epoch: 4/20... Training Step: 13208... Training loss: 1.5504... 0.0638 sec/batch
Epoch: 4/20... Training Step: 13209... Training loss: 1.5674... 0.0554 sec/batch
Epoch: 4/20... Training Step: 13210... Training loss: 1.6465... 0.0693 sec/batch
Epoch: 4/20... Training Step: 13211... Training loss: 1.7108... 0.0623 sec/batch
Epoch: 4/20... Training Step: 13212... Training loss: 1.6650... 0.0570 sec/batch
Epoch: 4/20... Training Step: 13213... Training loss: 1.6865... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13214... Training loss: 1.7683... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13215... Training loss: 1.6134... 0.0626 sec/batch
Epoch: 4/20... Training Step: 13216... Training loss: 1.8268... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13217... Training loss: 1.8505... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13218... Training loss: 1.7266... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13219... Training loss: 1.7352... 0.0584 sec/batch
Epoch: 4/20... Training Step: 13220... Training loss: 1.7002... 0.0544 sec/batch
Epoch: 4/20... Training Step: 13221... Training loss: 1.6993... 0.0627 sec/batch
Epoch: 4/20... Training Step: 13222... Training loss: 1.7387... 0.0527 sec/batch
Epoch: 4/20... Training Step: 13223... Training loss: 1.6388... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13224... Training loss: 1.7651... 0.0507 sec/batch
Epoch: 4/20... Training Step: 13225... Training loss: 1.7793... 0.0668 sec/batch
Epoch: 4/20... Training Step: 13226... Training loss: 1.7086... 0.0496 sec/batch
Epoch: 4/20... Training Step: 13227... Training loss: 1.6802... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13228... Training loss: 1.6588... 0.0530 sec/batch
Epoch: 4/20... Training Step: 13229... Training loss: 1.6599... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13230... Training loss: 1.7678... 0.0498 sec/batch
Epoch: 4/20... Training Step: 13231... Training loss: 1.7104... 0.0625 sec/batch
Epoch: 4/20... Training Step: 13232... Training loss: 1.7056... 0.0515 sec/batch
Epoch: 4/20... Training Step: 13233... Training loss: 1.7926... 0.0546 sec/batch
Epoch: 4/20... Training Step: 13234... Training loss: 1.6349... 0.0549 sec/batch
Epoch: 4/20... Training Step: 13235... Training loss: 1.7150... 0.0527 sec/batch
Epoch: 4/20... Training Step: 13236... Training loss: 1.6693... 0.0589 sec/batch
Epoch: 4/20... Training Step: 13237... Training loss: 1.6586... 0.0575 sec/batch
Epoch: 4/20... Training Step: 13238... Training loss: 1.7258... 0.0551 sec/batch
Epoch: 4/20... Training Step: 13239... Training loss: 1.7069... 0.0474 sec/batch
Epoch: 4/20... Training Step: 13240... Training loss: 1.6683... 0.0516 sec/batch
Epoch: 4/20... Training Step: 13241... Training loss: 1.7571... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13242... Training loss: 1.6510... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13243... Training loss: 1.6712... 0.0536 sec/batch
Epoch: 4/20... Training Step: 13244... Training loss: 1.7053... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13245... Training loss: 1.5743... 0.0548 sec/batch
Epoch: 4/20... Training Step: 13246... Training loss: 1.6674... 0.0556 sec/batch
Epoch: 4/20... Training Step: 13247... Training loss: 1.5610... 0.0541 sec/batch
Epoch: 4/20... Training Step: 13248... Training loss: 1.6328... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13249... Training loss: 1.6797... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13250... Training loss: 1.6502... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13251... Training loss: 1.5205... 0.0573 sec/batch
Epoch: 4/20... Training Step: 13252... Training loss: 1.6069... 0.0537 sec/batch
Epoch: 4/20... Training Step: 13253... Training loss: 1.7038... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13254... Training loss: 1.6913... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13255... Training loss: 1.7551... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13256... Training loss: 1.7008... 0.0611 sec/batch
Epoch: 4/20... Training Step: 13257... Training loss: 1.5924... 0.0714 sec/batch
Epoch: 4/20... Training Step: 13258... Training loss: 1.6545... 0.0572 sec/batch
Epoch: 4/20... Training Step: 13259... Training loss: 1.7564... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13260... Training loss: 1.5291... 0.0564 sec/batch
Epoch: 4/20... Training Step: 13261... Training loss: 1.5162... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13262... Training loss: 1.5857... 0.0514 sec/batch
Epoch: 4/20... Training Step: 13263... Training loss: 1.6631... 0.0548 sec/batch
Epoch: 4/20... Training Step: 13264... Training loss: 1.6555... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13265... Training loss: 1.6224... 0.0533 sec/batch
Epoch: 4/20... Training Step: 13266... Training loss: 1.6390... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13267... Training loss: 1.6944... 0.0574 sec/batch
Epoch: 4/20... Training Step: 13268... Training loss: 1.5786... 0.0532 sec/batch
Epoch: 4/20... Training Step: 13269... Training loss: 1.8076... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13270... Training loss: 1.5938... 0.0600 sec/batch
Epoch: 4/20... Training Step: 13271... Training loss: 1.7323... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13272... Training loss: 1.5987... 0.0541 sec/batch
Epoch: 4/20... Training Step: 13273... Training loss: 1.4922... 0.0708 sec/batch
Epoch: 4/20... Training Step: 13274... Training loss: 1.6369... 0.0537 sec/batch
Epoch: 4/20... Training Step: 13275... Training loss: 1.6155... 0.0501 sec/batch
Epoch: 4/20... Training Step: 13276... Training loss: 1.6600... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13277... Training loss: 1.5419... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13278... Training loss: 1.6341... 0.0524 sec/batch
Epoch: 4/20... Training Step: 13279... Training loss: 1.7700... 0.0585 sec/batch
Epoch: 4/20... Training Step: 13280... Training loss: 1.7301... 0.0524 sec/batch
Epoch: 4/20... Training Step: 13281... Training loss: 1.7054... 0.0531 sec/batch
Epoch: 4/20... Training Step: 13282... Training loss: 1.6029... 0.0499 sec/batch
Epoch: 4/20... Training Step: 13283... Training loss: 1.6543... 0.0640 sec/batch
Epoch: 4/20... Training Step: 13284... Training loss: 1.7387... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13285... Training loss: 1.6103... 0.0711 sec/batch
Epoch: 4/20... Training Step: 13286... Training loss: 1.6577... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13287... Training loss: 1.6737... 0.0603 sec/batch
Epoch: 4/20... Training Step: 13288... Training loss: 1.6492... 0.0703 sec/batch
Epoch: 4/20... Training Step: 13289... Training loss: 1.5466... 0.0638 sec/batch
Epoch: 4/20... Training Step: 13290... Training loss: 1.6500... 0.0564 sec/batch
Epoch: 4/20... Training Step: 13291... Training loss: 1.8317... 0.0542 sec/batch
Epoch: 4/20... Training Step: 13292... Training loss: 1.7400... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13293... Training loss: 1.8564... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13294... Training loss: 1.7285... 0.0549 sec/batch
Epoch: 4/20... Training Step: 13295... Training loss: 1.7215... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13296... Training loss: 1.9426... 0.0628 sec/batch
Epoch: 4/20... Training Step: 13297... Training loss: 1.6842... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13298... Training loss: 1.7269... 0.0631 sec/batch
Epoch: 4/20... Training Step: 13299... Training loss: 1.6834... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13300... Training loss: 1.7060... 0.0628 sec/batch
Epoch: 4/20... Training Step: 13301... Training loss: 1.6890... 0.0631 sec/batch
Epoch: 4/20... Training Step: 13302... Training loss: 1.6565... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13303... Training loss: 1.7669... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13304... Training loss: 1.6722... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13305... Training loss: 1.6981... 0.0710 sec/batch
Epoch: 4/20... Training Step: 13306... Training loss: 1.6052... 0.0630 sec/batch
Epoch: 4/20... Training Step: 13307... Training loss: 1.6306... 0.0647 sec/batch
Epoch: 4/20... Training Step: 13308... Training loss: 1.5768... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13309... Training loss: 1.7534... 0.0690 sec/batch
Epoch: 4/20... Training Step: 13310... Training loss: 1.7759... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13311... Training loss: 1.5106... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13312... Training loss: 1.7791... 0.0588 sec/batch
Epoch: 4/20... Training Step: 13313... Training loss: 1.6105... 0.0683 sec/batch
Epoch: 4/20... Training Step: 13314... Training loss: 1.7364... 0.0608 sec/batch
Epoch: 4/20... Training Step: 13315... Training loss: 1.7570... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13316... Training loss: 1.5441... 0.0517 sec/batch
Epoch: 4/20... Training Step: 13317... Training loss: 1.7099... 0.0603 sec/batch
Epoch: 4/20... Training Step: 13318... Training loss: 1.7670... 0.0622 sec/batch
Epoch: 4/20... Training Step: 13319... Training loss: 1.7201... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13320... Training loss: 1.6057... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13321... Training loss: 1.6109... 0.0735 sec/batch
Epoch: 4/20... Training Step: 13322... Training loss: 1.7237... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13323... Training loss: 1.6201... 0.0588 sec/batch
Epoch: 4/20... Training Step: 13324... Training loss: 1.9321... 0.0651 sec/batch
Epoch: 4/20... Training Step: 13325... Training loss: 1.6582... 0.0724 sec/batch
Epoch: 4/20... Training Step: 13326... Training loss: 1.6612... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13327... Training loss: 1.8815... 0.0596 sec/batch
Epoch: 4/20... Training Step: 13328... Training loss: 1.6682... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13329... Training loss: 1.6208... 0.0674 sec/batch
Epoch: 4/20... Training Step: 13330... Training loss: 1.7266... 0.0676 sec/batch
Epoch: 4/20... Training Step: 13331... Training loss: 1.8888... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13332... Training loss: 1.6882... 0.0524 sec/batch
Epoch: 4/20... Training Step: 13333... Training loss: 1.7661... 0.0674 sec/batch
Epoch: 4/20... Training Step: 13334... Training loss: 1.6297... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13335... Training loss: 1.6740... 0.0577 sec/batch
Epoch: 4/20... Training Step: 13336... Training loss: 1.6663... 0.0566 sec/batch
Epoch: 4/20... Training Step: 13337... Training loss: 1.5654... 0.0713 sec/batch
Epoch: 4/20... Training Step: 13338... Training loss: 1.7173... 0.0554 sec/batch
Epoch: 4/20... Training Step: 13339... Training loss: 1.6861... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13340... Training loss: 1.7153... 0.0664 sec/batch
Epoch: 4/20... Training Step: 13341... Training loss: 1.6865... 0.0595 sec/batch
Epoch: 4/20... Training Step: 13342... Training loss: 1.7371... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13343... Training loss: 1.8844... 0.0571 sec/batch
Epoch: 4/20... Training Step: 13344... Training loss: 1.6509... 0.0633 sec/batch
Epoch: 4/20... Training Step: 13345... Training loss: 1.6803... 0.0653 sec/batch
Epoch: 4/20... Training Step: 13346... Training loss: 1.7150... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13347... Training loss: 1.7188... 0.0867 sec/batch
Epoch: 4/20... Training Step: 13348... Training loss: 1.6419... 0.0985 sec/batch
Epoch: 4/20... Training Step: 13349... Training loss: 1.8364... 0.0922 sec/batch
Epoch: 4/20... Training Step: 13350... Training loss: 1.6184... 0.1048 sec/batch
Epoch: 4/20... Training Step: 13351... Training loss: 1.5264... 0.1267 sec/batch
Epoch: 4/20... Training Step: 13352... Training loss: 1.7674... 0.1015 sec/batch
Epoch: 4/20... Training Step: 13353... Training loss: 1.6704... 0.0821 sec/batch
Epoch: 4/20... Training Step: 13354... Training loss: 1.7871... 0.1023 sec/batch
Epoch: 4/20... Training Step: 13355... Training loss: 1.7010... 0.0832 sec/batch
Epoch: 4/20... Training Step: 13356... Training loss: 1.8006... 0.0970 sec/batch
Epoch: 4/20... Training Step: 13357... Training loss: 1.8200... 0.1125 sec/batch
Epoch: 4/20... Training Step: 13358... Training loss: 1.6227... 0.0936 sec/batch
Epoch: 4/20... Training Step: 13359... Training loss: 1.7614... 0.0843 sec/batch
Epoch: 4/20... Training Step: 13360... Training loss: 1.6855... 0.0932 sec/batch
Epoch: 4/20... Training Step: 13361... Training loss: 1.6637... 0.1029 sec/batch
Epoch: 4/20... Training Step: 13362... Training loss: 1.6072... 0.0959 sec/batch
Epoch: 4/20... Training Step: 13363... Training loss: 1.8471... 0.1012 sec/batch
Epoch: 4/20... Training Step: 13364... Training loss: 1.8094... 0.0978 sec/batch
Epoch: 4/20... Training Step: 13365... Training loss: 1.7819... 0.0893 sec/batch
Epoch: 4/20... Training Step: 13366... Training loss: 1.5639... 0.0895 sec/batch
Epoch: 4/20... Training Step: 13367... Training loss: 1.6670... 0.0915 sec/batch
Epoch: 4/20... Training Step: 13368... Training loss: 1.6704... 0.0921 sec/batch
Epoch: 4/20... Training Step: 13369... Training loss: 1.5995... 0.0881 sec/batch
Epoch: 4/20... Training Step: 13370... Training loss: 1.6843... 0.0829 sec/batch
Epoch: 4/20... Training Step: 13371... Training loss: 1.6705... 0.1158 sec/batch
Epoch: 4/20... Training Step: 13372... Training loss: 1.5402... 0.1137 sec/batch
Epoch: 4/20... Training Step: 13373... Training loss: 1.6586... 0.0943 sec/batch
Epoch: 4/20... Training Step: 13374... Training loss: 1.6900... 0.1004 sec/batch
Epoch: 4/20... Training Step: 13375... Training loss: 1.7148... 0.1543 sec/batch
Epoch: 4/20... Training Step: 13376... Training loss: 1.6471... 0.1025 sec/batch
Epoch: 4/20... Training Step: 13377... Training loss: 1.5968... 0.1343 sec/batch
Epoch: 4/20... Training Step: 13378... Training loss: 1.6418... 0.0931 sec/batch
Epoch: 4/20... Training Step: 13379... Training loss: 1.6405... 0.1023 sec/batch
Epoch: 4/20... Training Step: 13380... Training loss: 1.6585... 0.1339 sec/batch
Epoch: 4/20... Training Step: 13381... Training loss: 1.7052... 0.0926 sec/batch
Epoch: 4/20... Training Step: 13382... Training loss: 1.7597... 0.0906 sec/batch
Epoch: 4/20... Training Step: 13383... Training loss: 1.7190... 0.1053 sec/batch
Epoch: 4/20... Training Step: 13384... Training loss: 1.7480... 0.0977 sec/batch
Epoch: 4/20... Training Step: 13385... Training loss: 1.6543... 0.0884 sec/batch
Epoch: 4/20... Training Step: 13386... Training loss: 1.5789... 0.0999 sec/batch
Epoch: 4/20... Training Step: 13387... Training loss: 1.8173... 0.0776 sec/batch
Epoch: 4/20... Training Step: 13388... Training loss: 1.6982... 0.0686 sec/batch
Epoch: 4/20... Training Step: 13389... Training loss: 1.7018... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13390... Training loss: 1.6777... 0.0730 sec/batch
Epoch: 4/20... Training Step: 13391... Training loss: 1.6652... 0.0796 sec/batch
Epoch: 4/20... Training Step: 13392... Training loss: 1.6156... 0.0679 sec/batch
Epoch: 4/20... Training Step: 13393... Training loss: 1.6958... 0.0751 sec/batch
Epoch: 4/20... Training Step: 13394... Training loss: 1.8084... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13395... Training loss: 1.6600... 0.0768 sec/batch
Epoch: 4/20... Training Step: 13396... Training loss: 1.8090... 0.0632 sec/batch
Epoch: 4/20... Training Step: 13397... Training loss: 1.7322... 0.0682 sec/batch
Epoch: 4/20... Training Step: 13398... Training loss: 1.7361... 0.0663 sec/batch
Epoch: 4/20... Training Step: 13399... Training loss: 1.8153... 0.0849 sec/batch
Epoch: 4/20... Training Step: 13400... Training loss: 1.7181... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13401... Training loss: 1.8063... 0.0628 sec/batch
Epoch: 4/20... Training Step: 13402... Training loss: 1.6130... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13403... Training loss: 1.8006... 0.0637 sec/batch
Epoch: 4/20... Training Step: 13404... Training loss: 1.6087... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13405... Training loss: 1.6310... 0.0770 sec/batch
Epoch: 4/20... Training Step: 13406... Training loss: 1.5691... 0.0685 sec/batch
Epoch: 4/20... Training Step: 13407... Training loss: 1.7272... 0.0721 sec/batch
Epoch: 4/20... Training Step: 13408... Training loss: 1.6797... 0.0727 sec/batch
Epoch: 4/20... Training Step: 13409... Training loss: 1.6874... 0.0712 sec/batch
Epoch: 4/20... Training Step: 13410... Training loss: 1.6789... 0.0668 sec/batch
Epoch: 4/20... Training Step: 13411... Training loss: 1.6611... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13412... Training loss: 1.7142... 0.0667 sec/batch
Epoch: 4/20... Training Step: 13413... Training loss: 1.8610... 0.0587 sec/batch
Epoch: 4/20... Training Step: 13414... Training loss: 1.8447... 0.0642 sec/batch
Epoch: 4/20... Training Step: 13415... Training loss: 1.5710... 0.0731 sec/batch
Epoch: 4/20... Training Step: 13416... Training loss: 1.5866... 0.0707 sec/batch
Epoch: 4/20... Training Step: 13417... Training loss: 1.7438... 0.0617 sec/batch
Epoch: 4/20... Training Step: 13418... Training loss: 1.6228... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13419... Training loss: 1.5904... 0.0745 sec/batch
Epoch: 4/20... Training Step: 13420... Training loss: 1.6461... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13421... Training loss: 1.6323... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13422... Training loss: 1.5145... 0.0659 sec/batch
Epoch: 4/20... Training Step: 13423... Training loss: 1.6391... 0.0849 sec/batch
Epoch: 4/20... Training Step: 13424... Training loss: 1.6125... 0.0630 sec/batch
Epoch: 4/20... Training Step: 13425... Training loss: 1.6056... 0.0608 sec/batch
Epoch: 4/20... Training Step: 13426... Training loss: 1.7538... 0.0685 sec/batch
Epoch: 4/20... Training Step: 13427... Training loss: 1.7186... 0.0700 sec/batch
Epoch: 4/20... Training Step: 13428... Training loss: 1.6530... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13429... Training loss: 1.7528... 0.0608 sec/batch
Epoch: 4/20... Training Step: 13430... Training loss: 1.7132... 0.0655 sec/batch
Epoch: 4/20... Training Step: 13431... Training loss: 1.6446... 0.0722 sec/batch
Epoch: 4/20... Training Step: 13432... Training loss: 1.7841... 0.0662 sec/batch
Epoch: 4/20... Training Step: 13433... Training loss: 1.6971... 0.0573 sec/batch
Epoch: 4/20... Training Step: 13434... Training loss: 1.6741... 0.0656 sec/batch
Epoch: 4/20... Training Step: 13435... Training loss: 1.7854... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13436... Training loss: 1.7651... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13437... Training loss: 1.4494... 0.0694 sec/batch
Epoch: 4/20... Training Step: 13438... Training loss: 1.7075... 0.0786 sec/batch
Epoch: 4/20... Training Step: 13439... Training loss: 1.7281... 0.0736 sec/batch
Epoch: 4/20... Training Step: 13440... Training loss: 1.6710... 0.0665 sec/batch
Epoch: 4/20... Training Step: 13441... Training loss: 1.7653... 0.0659 sec/batch
Epoch: 4/20... Training Step: 13442... Training loss: 1.7889... 0.0714 sec/batch
Epoch: 4/20... Training Step: 13443... Training loss: 1.5825... 0.0545 sec/batch
Epoch: 4/20... Training Step: 13444... Training loss: 1.7214... 0.0647 sec/batch
Epoch: 4/20... Training Step: 13445... Training loss: 1.7350... 0.0696 sec/batch
Epoch: 4/20... Training Step: 13446... Training loss: 1.5771... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13447... Training loss: 1.5880... 0.0547 sec/batch
Epoch: 4/20... Training Step: 13448... Training loss: 1.7018... 0.0693 sec/batch
Epoch: 4/20... Training Step: 13449... Training loss: 1.9196... 0.0732 sec/batch
Epoch: 4/20... Training Step: 13450... Training loss: 1.7658... 0.0653 sec/batch
Epoch: 4/20... Training Step: 13451... Training loss: 1.7271... 0.0595 sec/batch
Epoch: 4/20... Training Step: 13452... Training loss: 1.7391... 0.0665 sec/batch
Epoch: 4/20... Training Step: 13453... Training loss: 1.6748... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13454... Training loss: 1.8164... 0.0576 sec/batch
Epoch: 4/20... Training Step: 13455... Training loss: 1.5549... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13456... Training loss: 1.7512... 0.0520 sec/batch
Epoch: 4/20... Training Step: 13457... Training loss: 1.5685... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13458... Training loss: 1.5399... 0.0593 sec/batch
Epoch: 4/20... Training Step: 13459... Training loss: 1.4923... 0.0738 sec/batch
Epoch: 4/20... Training Step: 13460... Training loss: 1.6016... 0.0593 sec/batch
Epoch: 4/20... Training Step: 13461... Training loss: 1.5637... 0.0672 sec/batch
Epoch: 4/20... Training Step: 13462... Training loss: 1.6847... 0.0632 sec/batch
Epoch: 4/20... Training Step: 13463... Training loss: 1.7926... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13464... Training loss: 1.7727... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13465... Training loss: 1.7660... 0.0743 sec/batch
Epoch: 4/20... Training Step: 13466... Training loss: 1.7166... 0.0557 sec/batch
Epoch: 4/20... Training Step: 13467... Training loss: 1.8162... 0.0553 sec/batch
Epoch: 4/20... Training Step: 13468... Training loss: 1.7216... 0.0531 sec/batch
Epoch: 4/20... Training Step: 13469... Training loss: 1.6972... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13470... Training loss: 1.6375... 0.0505 sec/batch
Epoch: 4/20... Training Step: 13471... Training loss: 1.6100... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13472... Training loss: 1.7086... 0.0617 sec/batch
Epoch: 4/20... Training Step: 13473... Training loss: 1.6404... 0.0682 sec/batch
Epoch: 4/20... Training Step: 13474... Training loss: 1.7207... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13475... Training loss: 1.7802... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13476... Training loss: 1.8551... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13477... Training loss: 1.6436... 0.0616 sec/batch
Epoch: 4/20... Training Step: 13478... Training loss: 1.6817... 0.0638 sec/batch
Epoch: 4/20... Training Step: 13479... Training loss: 1.6441... 0.0582 sec/batch
Epoch: 4/20... Training Step: 13480... Training loss: 1.6918... 0.0526 sec/batch
Epoch: 4/20... Training Step: 13481... Training loss: 1.6477... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13482... Training loss: 1.6343... 0.0559 sec/batch
Epoch: 4/20... Training Step: 13483... Training loss: 1.7017... 0.0640 sec/batch
Epoch: 4/20... Training Step: 13484... Training loss: 1.6005... 0.0613 sec/batch
Epoch: 4/20... Training Step: 13485... Training loss: 1.6819... 0.0700 sec/batch
Epoch: 4/20... Training Step: 13486... Training loss: 1.7235... 0.0665 sec/batch
Epoch: 4/20... Training Step: 13487... Training loss: 1.6831... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13488... Training loss: 1.8245... 0.0585 sec/batch
Epoch: 4/20... Training Step: 13489... Training loss: 1.7271... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13490... Training loss: 1.7241... 0.0533 sec/batch
Epoch: 4/20... Training Step: 13491... Training loss: 1.8292... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13492... Training loss: 1.7395... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13493... Training loss: 1.6997... 0.0569 sec/batch
Epoch: 4/20... Training Step: 13494... Training loss: 1.6363... 0.0622 sec/batch
Epoch: 4/20... Training Step: 13495... Training loss: 1.6883... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13496... Training loss: 1.5661... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13497... Training loss: 1.6542... 0.0777 sec/batch
Epoch: 4/20... Training Step: 13498... Training loss: 1.6474... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13499... Training loss: 1.7661... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13500... Training loss: 1.7124... 0.0580 sec/batch
Epoch: 4/20... Training Step: 13501... Training loss: 1.6139... 0.0816 sec/batch
Epoch: 4/20... Training Step: 13502... Training loss: 1.7002... 0.0573 sec/batch
Epoch: 4/20... Training Step: 13503... Training loss: 1.6006... 0.0602 sec/batch
Epoch: 4/20... Training Step: 13504... Training loss: 1.6973... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13505... Training loss: 1.6775... 0.0688 sec/batch
Epoch: 4/20... Training Step: 13506... Training loss: 1.6353... 0.0596 sec/batch
Epoch: 4/20... Training Step: 13507... Training loss: 1.6009... 0.0545 sec/batch
Epoch: 4/20... Training Step: 13508... Training loss: 1.7089... 0.0659 sec/batch
Epoch: 4/20... Training Step: 13509... Training loss: 1.7565... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13510... Training loss: 1.6892... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13511... Training loss: 1.7522... 0.0578 sec/batch
Epoch: 4/20... Training Step: 13512... Training loss: 1.6891... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13513... Training loss: 1.7398... 0.0734 sec/batch
Epoch: 4/20... Training Step: 13514... Training loss: 1.8067... 0.0611 sec/batch
Epoch: 4/20... Training Step: 13515... Training loss: 1.8154... 0.0573 sec/batch
Epoch: 4/20... Training Step: 13516... Training loss: 1.7483... 0.0539 sec/batch
Epoch: 4/20... Training Step: 13517... Training loss: 1.9360... 0.0669 sec/batch
Epoch: 4/20... Training Step: 13518... Training loss: 1.9201... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13519... Training loss: 1.8261... 0.0535 sec/batch
Epoch: 4/20... Training Step: 13520... Training loss: 1.7451... 0.0626 sec/batch
Epoch: 4/20... Training Step: 13521... Training loss: 1.8653... 0.0728 sec/batch
Epoch: 4/20... Training Step: 13522... Training loss: 1.6601... 0.0521 sec/batch
Epoch: 4/20... Training Step: 13523... Training loss: 1.7706... 0.0580 sec/batch
Epoch: 4/20... Training Step: 13524... Training loss: 1.7745... 0.0518 sec/batch
Epoch: 4/20... Training Step: 13525... Training loss: 1.8356... 0.0553 sec/batch
Epoch: 4/20... Training Step: 13526... Training loss: 1.6753... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13527... Training loss: 1.8055... 0.0518 sec/batch
Epoch: 4/20... Training Step: 13528... Training loss: 1.8036... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13529... Training loss: 1.7429... 0.0641 sec/batch
Epoch: 4/20... Training Step: 13530... Training loss: 1.7955... 0.0503 sec/batch
Epoch: 4/20... Training Step: 13531... Training loss: 1.7459... 0.0525 sec/batch
Epoch: 4/20... Training Step: 13532... Training loss: 1.6439... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13533... Training loss: 1.6620... 0.0559 sec/batch
Epoch: 4/20... Training Step: 13534... Training loss: 1.5880... 0.0567 sec/batch
Epoch: 4/20... Training Step: 13535... Training loss: 1.7880... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13536... Training loss: 1.7125... 0.0564 sec/batch
Epoch: 4/20... Training Step: 13537... Training loss: 1.5849... 0.0624 sec/batch
Epoch: 4/20... Training Step: 13538... Training loss: 1.7804... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13539... Training loss: 1.7424... 0.0604 sec/batch
Epoch: 4/20... Training Step: 13540... Training loss: 1.7423... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13541... Training loss: 1.7090... 0.0663 sec/batch
Epoch: 4/20... Training Step: 13542... Training loss: 1.5254... 0.0503 sec/batch
Epoch: 4/20... Training Step: 13543... Training loss: 1.6698... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13544... Training loss: 1.6288... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13545... Training loss: 1.6889... 0.0665 sec/batch
Epoch: 4/20... Training Step: 13546... Training loss: 1.7630... 0.0653 sec/batch
Epoch: 4/20... Training Step: 13547... Training loss: 1.6721... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13548... Training loss: 1.6123... 0.0554 sec/batch
Epoch: 4/20... Training Step: 13549... Training loss: 1.6636... 0.0641 sec/batch
Epoch: 4/20... Training Step: 13550... Training loss: 1.7428... 0.0632 sec/batch
Epoch: 4/20... Training Step: 13551... Training loss: 1.5318... 0.0603 sec/batch
Epoch: 4/20... Training Step: 13552... Training loss: 1.6511... 0.0542 sec/batch
Epoch: 4/20... Training Step: 13553... Training loss: 1.5792... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13554... Training loss: 1.6416... 0.0574 sec/batch
Epoch: 4/20... Training Step: 13555... Training loss: 1.6288... 0.0542 sec/batch
Epoch: 4/20... Training Step: 13556... Training loss: 1.6377... 0.0525 sec/batch
Epoch: 4/20... Training Step: 13557... Training loss: 1.5665... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13558... Training loss: 1.7479... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13559... Training loss: 1.5268... 0.0610 sec/batch
Epoch: 4/20... Training Step: 13560... Training loss: 1.6332... 0.0610 sec/batch
Epoch: 4/20... Training Step: 13561... Training loss: 1.5882... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13562... Training loss: 1.7814... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13563... Training loss: 1.6957... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13564... Training loss: 1.5566... 0.0582 sec/batch
Epoch: 4/20... Training Step: 13565... Training loss: 1.6233... 0.0653 sec/batch
Epoch: 4/20... Training Step: 13566... Training loss: 1.5171... 0.0549 sec/batch
Epoch: 4/20... Training Step: 13567... Training loss: 1.6552... 0.0589 sec/batch
Epoch: 4/20... Training Step: 13568... Training loss: 1.7411... 0.0525 sec/batch
Epoch: 4/20... Training Step: 13569... Training loss: 1.8071... 0.0676 sec/batch
Epoch: 4/20... Training Step: 13570... Training loss: 1.5579... 0.0520 sec/batch
Epoch: 4/20... Training Step: 13571... Training loss: 1.6749... 0.0623 sec/batch
Epoch: 4/20... Training Step: 13572... Training loss: 1.5383... 0.0529 sec/batch
Epoch: 4/20... Training Step: 13573... Training loss: 1.7224... 0.0715 sec/batch
Epoch: 4/20... Training Step: 13574... Training loss: 1.7776... 0.0628 sec/batch
Epoch: 4/20... Training Step: 13575... Training loss: 2.0102... 0.0647 sec/batch
Epoch: 4/20... Training Step: 13576... Training loss: 1.7549... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13577... Training loss: 1.6704... 0.0739 sec/batch
Epoch: 4/20... Training Step: 13578... Training loss: 1.7934... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13579... Training loss: 1.6578... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13580... Training loss: 1.8094... 0.0613 sec/batch
Epoch: 4/20... Training Step: 13581... Training loss: 1.6264... 0.0679 sec/batch
Epoch: 4/20... Training Step: 13582... Training loss: 1.7379... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13583... Training loss: 1.6653... 0.0665 sec/batch
Epoch: 4/20... Training Step: 13584... Training loss: 1.7178... 0.0576 sec/batch
Epoch: 4/20... Training Step: 13585... Training loss: 1.8442... 0.0701 sec/batch
Epoch: 4/20... Training Step: 13586... Training loss: 1.6291... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13587... Training loss: 1.8300... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13588... Training loss: 1.7864... 0.0716 sec/batch
Epoch: 4/20... Training Step: 13589... Training loss: 1.7720... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13590... Training loss: 1.5758... 0.0668 sec/batch
Epoch: 4/20... Training Step: 13591... Training loss: 1.7199... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13592... Training loss: 1.6912... 0.0660 sec/batch
Epoch: 4/20... Training Step: 13593... Training loss: 1.7660... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13594... Training loss: 1.6603... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13595... Training loss: 1.6657... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13596... Training loss: 1.7507... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13597... Training loss: 1.6932... 0.0651 sec/batch
Epoch: 4/20... Training Step: 13598... Training loss: 1.8008... 0.0577 sec/batch
Epoch: 4/20... Training Step: 13599... Training loss: 1.7182... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13600... Training loss: 1.6241... 0.0681 sec/batch
Epoch: 4/20... Training Step: 13601... Training loss: 1.6742... 0.0577 sec/batch
Epoch: 4/20... Training Step: 13602... Training loss: 1.6211... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13603... Training loss: 1.5189... 0.0712 sec/batch
Epoch: 4/20... Training Step: 13604... Training loss: 1.6379... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13605... Training loss: 1.7508... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13606... Training loss: 1.6304... 0.0602 sec/batch
Epoch: 4/20... Training Step: 13607... Training loss: 1.5740... 0.0693 sec/batch
Epoch: 4/20... Training Step: 13608... Training loss: 1.7570... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13609... Training loss: 1.7032... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13610... Training loss: 1.7085... 0.0553 sec/batch
Epoch: 4/20... Training Step: 13611... Training loss: 1.5498... 0.0716 sec/batch
Epoch: 4/20... Training Step: 13612... Training loss: 1.6607... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13613... Training loss: 1.6241... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13614... Training loss: 1.6999... 0.0595 sec/batch
Epoch: 4/20... Training Step: 13615... Training loss: 1.7469... 0.0587 sec/batch
Epoch: 4/20... Training Step: 13616... Training loss: 1.8030... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13617... Training loss: 1.6845... 0.0630 sec/batch
Epoch: 4/20... Training Step: 13618... Training loss: 1.6171... 0.0656 sec/batch
Epoch: 4/20... Training Step: 13619... Training loss: 1.7732... 0.0696 sec/batch
Epoch: 4/20... Training Step: 13620... Training loss: 1.6290... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13621... Training loss: 1.7291... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13622... Training loss: 1.6585... 0.0577 sec/batch
Epoch: 4/20... Training Step: 13623... Training loss: 1.5023... 0.0687 sec/batch
Epoch: 4/20... Training Step: 13624... Training loss: 1.7545... 0.0674 sec/batch
Epoch: 4/20... Training Step: 13625... Training loss: 1.7424... 0.0602 sec/batch
Epoch: 4/20... Training Step: 13626... Training loss: 1.7713... 0.0604 sec/batch
Epoch: 4/20... Training Step: 13627... Training loss: 1.5068... 0.0704 sec/batch
Epoch: 4/20... Training Step: 13628... Training loss: 1.7161... 0.0600 sec/batch
Epoch: 4/20... Training Step: 13629... Training loss: 1.6430... 0.0668 sec/batch
Epoch: 4/20... Training Step: 13630... Training loss: 1.6328... 0.0588 sec/batch
Epoch: 4/20... Training Step: 13631... Training loss: 1.7078... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13632... Training loss: 1.7064... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13633... Training loss: 1.6625... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13634... Training loss: 1.5120... 0.0726 sec/batch
Epoch: 4/20... Training Step: 13635... Training loss: 1.6634... 0.0943 sec/batch
Epoch: 4/20... Training Step: 13636... Training loss: 1.7161... 0.0861 sec/batch
Epoch: 4/20... Training Step: 13637... Training loss: 1.5769... 0.0858 sec/batch
Epoch: 4/20... Training Step: 13638... Training loss: 1.6992... 0.1001 sec/batch
Epoch: 4/20... Training Step: 13639... Training loss: 1.6474... 0.0815 sec/batch
Epoch: 4/20... Training Step: 13640... Training loss: 1.5503... 0.0914 sec/batch
Epoch: 4/20... Training Step: 13641... Training loss: 1.6586... 0.0902 sec/batch
Epoch: 4/20... Training Step: 13642... Training loss: 1.8394... 0.0865 sec/batch
Epoch: 4/20... Training Step: 13643... Training loss: 1.7249... 0.0892 sec/batch
Epoch: 4/20... Training Step: 13644... Training loss: 1.9101... 0.1003 sec/batch
Epoch: 4/20... Training Step: 13645... Training loss: 1.7861... 0.0897 sec/batch
Epoch: 4/20... Training Step: 13646... Training loss: 1.6930... 0.0901 sec/batch
Epoch: 4/20... Training Step: 13647... Training loss: 1.6028... 0.0986 sec/batch
Epoch: 4/20... Training Step: 13648... Training loss: 1.6315... 0.0929 sec/batch
Epoch: 4/20... Training Step: 13649... Training loss: 1.6922... 0.0885 sec/batch
Epoch: 4/20... Training Step: 13650... Training loss: 1.7069... 0.0949 sec/batch
Epoch: 4/20... Training Step: 13651... Training loss: 1.6721... 0.0907 sec/batch
Epoch: 4/20... Training Step: 13652... Training loss: 1.6387... 0.0891 sec/batch
Epoch: 4/20... Training Step: 13653... Training loss: 1.7625... 0.1020 sec/batch
Epoch: 4/20... Training Step: 13654... Training loss: 1.6358... 0.0795 sec/batch
Epoch: 4/20... Training Step: 13655... Training loss: 1.7033... 0.0888 sec/batch
Epoch: 4/20... Training Step: 13656... Training loss: 1.4909... 0.1147 sec/batch
Epoch: 4/20... Training Step: 13657... Training loss: 1.7084... 0.0871 sec/batch
Epoch: 4/20... Training Step: 13658... Training loss: 1.7564... 0.0941 sec/batch
Epoch: 4/20... Training Step: 13659... Training loss: 1.4889... 0.0880 sec/batch
Epoch: 4/20... Training Step: 13660... Training loss: 1.7495... 0.0963 sec/batch
Epoch: 4/20... Training Step: 13661... Training loss: 1.7241... 0.0977 sec/batch
Epoch: 4/20... Training Step: 13662... Training loss: 1.6108... 0.0891 sec/batch
Epoch: 4/20... Training Step: 13663... Training loss: 1.5768... 0.0825 sec/batch
Epoch: 4/20... Training Step: 13664... Training loss: 1.7045... 0.0795 sec/batch
Epoch: 4/20... Training Step: 13665... Training loss: 1.6728... 0.0978 sec/batch
Epoch: 4/20... Training Step: 13666... Training loss: 1.5350... 0.0830 sec/batch
Epoch: 4/20... Training Step: 13667... Training loss: 1.5889... 0.0856 sec/batch
Epoch: 4/20... Training Step: 13668... Training loss: 1.6903... 0.0886 sec/batch
Epoch: 4/20... Training Step: 13669... Training loss: 1.7414... 0.0904 sec/batch
Epoch: 4/20... Training Step: 13670... Training loss: 1.6026... 0.0820 sec/batch
Epoch: 4/20... Training Step: 13671... Training loss: 1.6907... 0.1023 sec/batch
Epoch: 4/20... Training Step: 13672... Training loss: 1.7077... 0.0852 sec/batch
Epoch: 4/20... Training Step: 13673... Training loss: 1.6105... 0.0849 sec/batch
Epoch: 4/20... Training Step: 13674... Training loss: 1.7365... 0.1007 sec/batch
Epoch: 4/20... Training Step: 13675... Training loss: 1.7104... 0.0824 sec/batch
Epoch: 4/20... Training Step: 13676... Training loss: 1.7087... 0.0851 sec/batch
Epoch: 4/20... Training Step: 13677... Training loss: 1.5247... 0.0846 sec/batch
Epoch: 4/20... Training Step: 13678... Training loss: 1.6625... 0.0767 sec/batch
Epoch: 4/20... Training Step: 13679... Training loss: 1.7778... 0.0647 sec/batch
Epoch: 4/20... Training Step: 13680... Training loss: 1.7949... 0.0674 sec/batch
Epoch: 4/20... Training Step: 13681... Training loss: 1.5429... 0.0683 sec/batch
Epoch: 4/20... Training Step: 13682... Training loss: 1.6068... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13683... Training loss: 1.6292... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13684... Training loss: 1.8194... 0.0820 sec/batch
Epoch: 4/20... Training Step: 13685... Training loss: 1.7571... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13686... Training loss: 1.5050... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13687... Training loss: 1.6880... 0.0654 sec/batch
Epoch: 4/20... Training Step: 13688... Training loss: 1.5102... 0.0682 sec/batch
Epoch: 4/20... Training Step: 13689... Training loss: 1.6230... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13690... Training loss: 1.7275... 0.0679 sec/batch
Epoch: 4/20... Training Step: 13691... Training loss: 1.7179... 0.0678 sec/batch
Epoch: 4/20... Training Step: 13692... Training loss: 1.6215... 0.0738 sec/batch
Epoch: 4/20... Training Step: 13693... Training loss: 1.6710... 0.0672 sec/batch
Epoch: 4/20... Training Step: 13694... Training loss: 1.5943... 0.0694 sec/batch
Epoch: 4/20... Training Step: 13695... Training loss: 1.6404... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13696... Training loss: 1.6762... 0.0652 sec/batch
Epoch: 4/20... Training Step: 13697... Training loss: 1.6999... 0.0677 sec/batch
Epoch: 4/20... Training Step: 13698... Training loss: 1.6268... 0.0694 sec/batch
Epoch: 4/20... Training Step: 13699... Training loss: 1.6893... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13700... Training loss: 1.6174... 0.0654 sec/batch
Epoch: 4/20... Training Step: 13701... Training loss: 1.6851... 0.0561 sec/batch
Epoch: 4/20... Training Step: 13702... Training loss: 1.8296... 0.0732 sec/batch
Epoch: 4/20... Training Step: 13703... Training loss: 1.5503... 0.0690 sec/batch
Epoch: 4/20... Training Step: 13704... Training loss: 1.4767... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13705... Training loss: 1.4618... 0.0680 sec/batch
Epoch: 4/20... Training Step: 13706... Training loss: 1.6667... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13707... Training loss: 1.4588... 0.0688 sec/batch
Epoch: 4/20... Training Step: 13708... Training loss: 1.5653... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13709... Training loss: 1.6221... 0.0638 sec/batch
Epoch: 4/20... Training Step: 13710... Training loss: 1.6560... 0.0761 sec/batch
Epoch: 4/20... Training Step: 13711... Training loss: 1.6551... 0.0664 sec/batch
Epoch: 4/20... Training Step: 13712... Training loss: 1.6237... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13713... Training loss: 1.6287... 0.0671 sec/batch
Epoch: 4/20... Training Step: 13714... Training loss: 1.6207... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13715... Training loss: 1.5400... 0.0656 sec/batch
Epoch: 4/20... Training Step: 13716... Training loss: 1.5668... 0.0695 sec/batch
Epoch: 4/20... Training Step: 13717... Training loss: 1.5321... 0.0667 sec/batch
Epoch: 4/20... Training Step: 13718... Training loss: 1.7505... 0.0693 sec/batch
Epoch: 4/20... Training Step: 13719... Training loss: 1.5586... 0.0688 sec/batch
Epoch: 4/20... Training Step: 13720... Training loss: 1.4368... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13721... Training loss: 1.6091... 0.0791 sec/batch
Epoch: 4/20... Training Step: 13722... Training loss: 1.6681... 0.0676 sec/batch
Epoch: 4/20... Training Step: 13723... Training loss: 1.5730... 0.0602 sec/batch
Epoch: 4/20... Training Step: 13724... Training loss: 1.6885... 0.0698 sec/batch
Epoch: 4/20... Training Step: 13725... Training loss: 1.6316... 0.0671 sec/batch
Epoch: 4/20... Training Step: 13726... Training loss: 1.7994... 0.0662 sec/batch
Epoch: 4/20... Training Step: 13727... Training loss: 1.5934... 0.0784 sec/batch
Epoch: 4/20... Training Step: 13728... Training loss: 1.5694... 0.0717 sec/batch
Epoch: 4/20... Training Step: 13729... Training loss: 1.6430... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13730... Training loss: 1.6894... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13731... Training loss: 1.6517... 0.0697 sec/batch
Epoch: 4/20... Training Step: 13732... Training loss: 1.7811... 0.0723 sec/batch
Epoch: 4/20... Training Step: 13733... Training loss: 1.6303... 0.0667 sec/batch
Epoch: 4/20... Training Step: 13734... Training loss: 1.6876... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13735... Training loss: 1.7143... 0.0654 sec/batch
Epoch: 4/20... Training Step: 13736... Training loss: 1.6029... 0.0698 sec/batch
Epoch: 4/20... Training Step: 13737... Training loss: 1.7711... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13738... Training loss: 1.6553... 0.0647 sec/batch
Epoch: 4/20... Training Step: 13739... Training loss: 1.7918... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13740... Training loss: 1.6567... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13741... Training loss: 1.7586... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13742... Training loss: 1.6803... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13743... Training loss: 1.6608... 0.0586 sec/batch
Epoch: 4/20... Training Step: 13744... Training loss: 1.7376... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13745... Training loss: 1.6456... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13746... Training loss: 1.6471... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13747... Training loss: 1.6739... 0.0554 sec/batch
Epoch: 4/20... Training Step: 13748... Training loss: 1.6277... 0.0695 sec/batch
Epoch: 4/20... Training Step: 13749... Training loss: 1.6043... 0.0626 sec/batch
Epoch: 4/20... Training Step: 13750... Training loss: 1.5934... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13751... Training loss: 1.5965... 0.0623 sec/batch
Epoch: 4/20... Training Step: 13752... Training loss: 1.7510... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13753... Training loss: 1.4674... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13754... Training loss: 1.5054... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13755... Training loss: 1.7214... 0.0624 sec/batch
Epoch: 4/20... Training Step: 13756... Training loss: 1.7357... 0.0756 sec/batch
Epoch: 4/20... Training Step: 13757... Training loss: 1.6468... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13758... Training loss: 1.7027... 0.0613 sec/batch
Epoch: 4/20... Training Step: 13759... Training loss: 1.5992... 0.0611 sec/batch
Epoch: 4/20... Training Step: 13760... Training loss: 1.7612... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13761... Training loss: 1.6027... 0.0664 sec/batch
Epoch: 4/20... Training Step: 13762... Training loss: 1.7442... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13763... Training loss: 1.8239... 0.0587 sec/batch
Epoch: 4/20... Training Step: 13764... Training loss: 1.8352... 0.0656 sec/batch
Epoch: 4/20... Training Step: 13765... Training loss: 1.6909... 0.0584 sec/batch
Epoch: 4/20... Training Step: 13766... Training loss: 1.5535... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13767... Training loss: 1.9044... 0.0635 sec/batch
Epoch: 4/20... Training Step: 13768... Training loss: 1.8371... 0.0686 sec/batch
Epoch: 4/20... Training Step: 13769... Training loss: 1.6105... 0.0652 sec/batch
Epoch: 4/20... Training Step: 13770... Training loss: 1.7730... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13771... Training loss: 1.7925... 0.0617 sec/batch
Epoch: 4/20... Training Step: 13772... Training loss: 1.6466... 0.0689 sec/batch
Epoch: 4/20... Training Step: 13773... Training loss: 1.7104... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13774... Training loss: 1.7990... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13775... Training loss: 1.6519... 0.0623 sec/batch
Epoch: 4/20... Training Step: 13776... Training loss: 1.5193... 0.0719 sec/batch
Epoch: 4/20... Training Step: 13777... Training loss: 1.7619... 0.0626 sec/batch
Epoch: 4/20... Training Step: 13778... Training loss: 1.6321... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13779... Training loss: 1.6543... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13780... Training loss: 1.6121... 0.0729 sec/batch
Epoch: 4/20... Training Step: 13781... Training loss: 1.6455... 0.0610 sec/batch
Epoch: 4/20... Training Step: 13782... Training loss: 1.4695... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13783... Training loss: 1.6373... 0.0525 sec/batch
Epoch: 4/20... Training Step: 13784... Training loss: 1.5977... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13785... Training loss: 1.6883... 0.0521 sec/batch
Epoch: 4/20... Training Step: 13786... Training loss: 1.7613... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13787... Training loss: 1.5786... 0.0592 sec/batch
Epoch: 4/20... Training Step: 13788... Training loss: 1.6999... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13789... Training loss: 1.5337... 0.0659 sec/batch
Epoch: 4/20... Training Step: 13790... Training loss: 1.7454... 0.0572 sec/batch
Epoch: 4/20... Training Step: 13791... Training loss: 1.8464... 0.0618 sec/batch
Epoch: 4/20... Training Step: 13792... Training loss: 1.5415... 0.0635 sec/batch
Epoch: 4/20... Training Step: 13793... Training loss: 1.7804... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13794... Training loss: 1.8136... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13795... Training loss: 1.5059... 0.0618 sec/batch
Epoch: 4/20... Training Step: 13796... Training loss: 1.6673... 0.0693 sec/batch
Epoch: 4/20... Training Step: 13797... Training loss: 1.7471... 0.0611 sec/batch
Epoch: 4/20... Training Step: 13798... Training loss: 1.7942... 0.0604 sec/batch
Epoch: 4/20... Training Step: 13799... Training loss: 1.5796... 0.0616 sec/batch
Epoch: 4/20... Training Step: 13800... Training loss: 1.6436... 0.0633 sec/batch
Epoch: 4/20... Training Step: 13801... Training loss: 1.7938... 0.0566 sec/batch
Epoch: 4/20... Training Step: 13802... Training loss: 1.7424... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13803... Training loss: 1.6633... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13804... Training loss: 1.6449... 0.0556 sec/batch
Epoch: 4/20... Training Step: 13805... Training loss: 1.5484... 0.0585 sec/batch
Epoch: 4/20... Training Step: 13806... Training loss: 1.6687... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13807... Training loss: 1.7439... 0.0622 sec/batch
Epoch: 4/20... Training Step: 13808... Training loss: 1.7983... 0.0612 sec/batch
Epoch: 4/20... Training Step: 13809... Training loss: 1.6511... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13810... Training loss: 1.5480... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13811... Training loss: 1.6980... 0.0535 sec/batch
Epoch: 4/20... Training Step: 13812... Training loss: 1.7156... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13813... Training loss: 1.7033... 0.0520 sec/batch
Epoch: 4/20... Training Step: 13814... Training loss: 1.4615... 0.0525 sec/batch
Epoch: 4/20... Training Step: 13815... Training loss: 1.5596... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13816... Training loss: 1.7326... 0.0523 sec/batch
Epoch: 4/20... Training Step: 13817... Training loss: 1.7257... 0.0539 sec/batch
Epoch: 4/20... Training Step: 13818... Training loss: 1.7473... 0.0537 sec/batch
Epoch: 4/20... Training Step: 13819... Training loss: 1.6820... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13820... Training loss: 1.6321... 0.0633 sec/batch
Epoch: 4/20... Training Step: 13821... Training loss: 1.5761... 0.0583 sec/batch
Epoch: 4/20... Training Step: 13822... Training loss: 1.7671... 0.0616 sec/batch
Epoch: 4/20... Training Step: 13823... Training loss: 1.6101... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13824... Training loss: 1.7009... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13825... Training loss: 1.6123... 0.0677 sec/batch
Epoch: 4/20... Training Step: 13826... Training loss: 1.6675... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13827... Training loss: 1.6271... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13828... Training loss: 1.7929... 0.0600 sec/batch
Epoch: 4/20... Training Step: 13829... Training loss: 1.6239... 0.0582 sec/batch
Epoch: 4/20... Training Step: 13830... Training loss: 1.7503... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13831... Training loss: 1.7783... 0.0698 sec/batch
Epoch: 4/20... Training Step: 13832... Training loss: 1.5580... 0.0555 sec/batch
Epoch: 4/20... Training Step: 13833... Training loss: 1.6024... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13834... Training loss: 1.6482... 0.0592 sec/batch
Epoch: 4/20... Training Step: 13835... Training loss: 1.6173... 0.0596 sec/batch
Epoch: 4/20... Training Step: 13836... Training loss: 1.5223... 0.0558 sec/batch
Epoch: 4/20... Training Step: 13837... Training loss: 1.5902... 0.0548 sec/batch
Epoch: 4/20... Training Step: 13838... Training loss: 1.6918... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13839... Training loss: 1.6659... 0.0776 sec/batch
Epoch: 4/20... Training Step: 13840... Training loss: 1.6844... 0.0690 sec/batch
Epoch: 4/20... Training Step: 13841... Training loss: 1.6430... 0.0760 sec/batch
Epoch: 4/20... Training Step: 13842... Training loss: 1.5883... 0.0705 sec/batch
Epoch: 4/20... Training Step: 13843... Training loss: 1.6400... 0.0795 sec/batch
Epoch: 4/20... Training Step: 13844... Training loss: 1.9659... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13845... Training loss: 1.6299... 0.0678 sec/batch
Epoch: 4/20... Training Step: 13846... Training loss: 1.6129... 0.0663 sec/batch
Epoch: 4/20... Training Step: 13847... Training loss: 1.8057... 0.0639 sec/batch
Epoch: 4/20... Training Step: 13848... Training loss: 2.0155... 0.0532 sec/batch
Epoch: 4/20... Training Step: 13849... Training loss: 1.7692... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13850... Training loss: 1.9012... 0.0592 sec/batch
Epoch: 4/20... Training Step: 13851... Training loss: 1.6573... 0.0536 sec/batch
Epoch: 4/20... Training Step: 13852... Training loss: 1.6566... 0.0585 sec/batch
Epoch: 4/20... Training Step: 13853... Training loss: 1.6685... 0.0669 sec/batch
Epoch: 4/20... Training Step: 13854... Training loss: 1.6646... 0.0634 sec/batch
Epoch: 4/20... Training Step: 13855... Training loss: 1.7598... 0.0671 sec/batch
Epoch: 4/20... Training Step: 13856... Training loss: 1.5189... 0.0521 sec/batch
Epoch: 4/20... Training Step: 13857... Training loss: 1.6595... 0.0704 sec/batch
Epoch: 4/20... Training Step: 13858... Training loss: 1.6723... 0.0555 sec/batch
Epoch: 4/20... Training Step: 13859... Training loss: 1.7380... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13860... Training loss: 1.5968... 0.0616 sec/batch
Epoch: 4/20... Training Step: 13861... Training loss: 1.6983... 0.0667 sec/batch
Epoch: 4/20... Training Step: 13862... Training loss: 1.7017... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13863... Training loss: 1.6673... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13864... Training loss: 1.6512... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13865... Training loss: 1.8801... 0.0651 sec/batch
Epoch: 4/20... Training Step: 13866... Training loss: 1.8056... 0.0918 sec/batch
Epoch: 4/20... Training Step: 13867... Training loss: 1.6744... 0.0707 sec/batch
Epoch: 4/20... Training Step: 13868... Training loss: 1.6381... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13869... Training loss: 1.6369... 0.0722 sec/batch
Epoch: 4/20... Training Step: 13870... Training loss: 1.7191... 0.0624 sec/batch
Epoch: 4/20... Training Step: 13871... Training loss: 1.6828... 0.0618 sec/batch
Epoch: 4/20... Training Step: 13872... Training loss: 1.6948... 0.0703 sec/batch
Epoch: 4/20... Training Step: 13873... Training loss: 1.5689... 0.0619 sec/batch
Epoch: 4/20... Training Step: 13874... Training loss: 1.4493... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13875... Training loss: 1.6209... 0.0593 sec/batch
Epoch: 4/20... Training Step: 13876... Training loss: 1.5553... 0.0729 sec/batch
Epoch: 4/20... Training Step: 13877... Training loss: 1.7091... 0.0670 sec/batch
Epoch: 4/20... Training Step: 13878... Training loss: 1.6980... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13879... Training loss: 1.9054... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13880... Training loss: 1.7453... 0.0649 sec/batch
Epoch: 4/20... Training Step: 13881... Training loss: 1.6759... 0.0630 sec/batch
Epoch: 4/20... Training Step: 13882... Training loss: 1.6999... 0.0597 sec/batch
Epoch: 4/20... Training Step: 13883... Training loss: 1.7210... 0.0599 sec/batch
Epoch: 4/20... Training Step: 13884... Training loss: 1.7209... 0.0668 sec/batch
Epoch: 4/20... Training Step: 13885... Training loss: 1.8690... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13886... Training loss: 1.5150... 0.0667 sec/batch
Epoch: 4/20... Training Step: 13887... Training loss: 1.6604... 0.0590 sec/batch
Epoch: 4/20... Training Step: 13888... Training loss: 1.6568... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13889... Training loss: 1.7055... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13890... Training loss: 1.6250... 0.0562 sec/batch
Epoch: 4/20... Training Step: 13891... Training loss: 1.7967... 0.0617 sec/batch
Epoch: 4/20... Training Step: 13892... Training loss: 1.6517... 0.0760 sec/batch
Epoch: 4/20... Training Step: 13893... Training loss: 1.7224... 0.0651 sec/batch
Epoch: 4/20... Training Step: 13894... Training loss: 1.8555... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13895... Training loss: 1.8772... 0.0623 sec/batch
Epoch: 4/20... Training Step: 13896... Training loss: 1.6864... 0.0624 sec/batch
Epoch: 4/20... Training Step: 13897... Training loss: 1.6358... 0.0601 sec/batch
Epoch: 4/20... Training Step: 13898... Training loss: 1.7279... 0.0652 sec/batch
Epoch: 4/20... Training Step: 13899... Training loss: 1.6272... 0.0614 sec/batch
Epoch: 4/20... Training Step: 13900... Training loss: 1.6519... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13901... Training loss: 1.5782... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13902... Training loss: 1.7233... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13903... Training loss: 1.9236... 0.0603 sec/batch
Epoch: 4/20... Training Step: 13904... Training loss: 1.8341... 0.0631 sec/batch
Epoch: 4/20... Training Step: 13905... Training loss: 1.6690... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13906... Training loss: 1.7602... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13907... Training loss: 1.6490... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13908... Training loss: 1.7109... 0.0696 sec/batch
Epoch: 4/20... Training Step: 13909... Training loss: 1.6780... 0.0598 sec/batch
Epoch: 4/20... Training Step: 13910... Training loss: 1.6547... 0.0607 sec/batch
Epoch: 4/20... Training Step: 13911... Training loss: 1.5598... 0.0544 sec/batch
Epoch: 4/20... Training Step: 13912... Training loss: 1.5702... 0.0722 sec/batch
Epoch: 4/20... Training Step: 13913... Training loss: 1.5096... 0.0549 sec/batch
Epoch: 4/20... Training Step: 13914... Training loss: 1.5382... 0.0657 sec/batch
Epoch: 4/20... Training Step: 13915... Training loss: 1.6836... 0.0675 sec/batch
Epoch: 4/20... Training Step: 13916... Training loss: 1.5263... 0.0643 sec/batch
Epoch: 4/20... Training Step: 13917... Training loss: 1.6712... 0.0596 sec/batch
Epoch: 4/20... Training Step: 13918... Training loss: 1.6077... 0.0695 sec/batch
Epoch: 4/20... Training Step: 13919... Training loss: 1.5663... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13920... Training loss: 1.6982... 0.0673 sec/batch
Epoch: 4/20... Training Step: 13921... Training loss: 1.5941... 0.0558 sec/batch
Epoch: 4/20... Training Step: 13922... Training loss: 1.5895... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13923... Training loss: 1.6121... 0.0617 sec/batch
Epoch: 4/20... Training Step: 13924... Training loss: 1.6201... 0.0684 sec/batch
Epoch: 4/20... Training Step: 13925... Training loss: 1.6656... 0.0588 sec/batch
Epoch: 4/20... Training Step: 13926... Training loss: 1.6226... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13927... Training loss: 1.5472... 0.0613 sec/batch
Epoch: 4/20... Training Step: 13928... Training loss: 1.4902... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13929... Training loss: 1.4973... 0.0629 sec/batch
Epoch: 4/20... Training Step: 13930... Training loss: 1.5685... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13931... Training loss: 1.7492... 0.0533 sec/batch
Epoch: 4/20... Training Step: 13932... Training loss: 1.6018... 0.0661 sec/batch
Epoch: 4/20... Training Step: 13933... Training loss: 1.8199... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13934... Training loss: 1.6033... 0.0642 sec/batch
Epoch: 4/20... Training Step: 13935... Training loss: 1.5828... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13936... Training loss: 1.5723... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13937... Training loss: 1.6363... 0.0572 sec/batch
Epoch: 4/20... Training Step: 13938... Training loss: 1.4953... 0.0518 sec/batch
Epoch: 4/20... Training Step: 13939... Training loss: 1.6921... 0.0654 sec/batch
Epoch: 4/20... Training Step: 13940... Training loss: 1.5492... 0.0683 sec/batch
Epoch: 4/20... Training Step: 13941... Training loss: 1.6988... 0.0644 sec/batch
Epoch: 4/20... Training Step: 13942... Training loss: 1.7079... 0.0658 sec/batch
Epoch: 4/20... Training Step: 13943... Training loss: 1.6911... 0.0581 sec/batch
Epoch: 4/20... Training Step: 13944... Training loss: 1.7539... 0.0620 sec/batch
Epoch: 4/20... Training Step: 13945... Training loss: 1.6497... 0.0552 sec/batch
Epoch: 4/20... Training Step: 13946... Training loss: 1.6699... 0.0536 sec/batch
Epoch: 4/20... Training Step: 13947... Training loss: 1.4863... 0.0660 sec/batch
Epoch: 4/20... Training Step: 13948... Training loss: 1.6188... 0.0679 sec/batch
Epoch: 4/20... Training Step: 13949... Training loss: 1.6312... 0.0571 sec/batch
Epoch: 4/20... Training Step: 13950... Training loss: 1.7664... 0.0553 sec/batch
Epoch: 4/20... Training Step: 13951... Training loss: 1.6780... 0.0584 sec/batch
Epoch: 4/20... Training Step: 13952... Training loss: 1.6637... 0.0613 sec/batch
Epoch: 4/20... Training Step: 13953... Training loss: 1.6845... 0.0560 sec/batch
Epoch: 4/20... Training Step: 13954... Training loss: 1.7925... 0.0576 sec/batch
Epoch: 4/20... Training Step: 13955... Training loss: 1.8005... 0.0594 sec/batch
Epoch: 4/20... Training Step: 13956... Training loss: 1.6660... 0.0651 sec/batch
Epoch: 4/20... Training Step: 13957... Training loss: 1.7349... 0.0662 sec/batch
Epoch: 4/20... Training Step: 13958... Training loss: 1.5663... 0.0593 sec/batch
Epoch: 4/20... Training Step: 13959... Training loss: 1.7654... 0.0609 sec/batch
Epoch: 4/20... Training Step: 13960... Training loss: 1.8127... 0.0627 sec/batch
Epoch: 4/20... Training Step: 13961... Training loss: 1.6984... 0.0612 sec/batch
Epoch: 4/20... Training Step: 13962... Training loss: 1.6324... 0.0519 sec/batch
Epoch: 4/20... Training Step: 13963... Training loss: 1.5645... 0.0559 sec/batch
Epoch: 4/20... Training Step: 13964... Training loss: 1.5096... 0.0563 sec/batch
Epoch: 4/20... Training Step: 13965... Training loss: 1.5689... 0.0606 sec/batch
Epoch: 4/20... Training Step: 13966... Training loss: 1.6183... 0.0575 sec/batch
Epoch: 4/20... Training Step: 13967... Training loss: 1.5938... 0.0511 sec/batch
Epoch: 4/20... Training Step: 13968... Training loss: 1.6327... 0.0621 sec/batch
Epoch: 4/20... Training Step: 13969... Training loss: 1.6140... 0.0518 sec/batch
Epoch: 4/20... Training Step: 13970... Training loss: 1.6326... 0.0580 sec/batch
Epoch: 4/20... Training Step: 13971... Training loss: 1.7059... 0.0630 sec/batch
Epoch: 4/20... Training Step: 13972... Training loss: 1.6681... 0.0737 sec/batch
Epoch: 4/20... Training Step: 13973... Training loss: 1.5728... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13974... Training loss: 1.7965... 0.0645 sec/batch
Epoch: 4/20... Training Step: 13975... Training loss: 1.6202... 0.0505 sec/batch
Epoch: 4/20... Training Step: 13976... Training loss: 1.7942... 0.0571 sec/batch
Epoch: 4/20... Training Step: 13977... Training loss: 1.6430... 0.0509 sec/batch
Epoch: 4/20... Training Step: 13978... Training loss: 1.7223... 0.0579 sec/batch
Epoch: 4/20... Training Step: 13979... Training loss: 1.7217... 0.0574 sec/batch
Epoch: 4/20... Training Step: 13980... Training loss: 1.6062... 0.0709 sec/batch
Epoch: 4/20... Training Step: 13981... Training loss: 1.7060... 0.0631 sec/batch
Epoch: 4/20... Training Step: 13982... Training loss: 1.8279... 0.0605 sec/batch
Epoch: 4/20... Training Step: 13983... Training loss: 1.6121... 0.0636 sec/batch
Epoch: 4/20... Training Step: 13984... Training loss: 1.8970... 0.0615 sec/batch
Epoch: 4/20... Training Step: 13985... Training loss: 1.6419... 0.0507 sec/batch
Epoch: 4/20... Training Step: 13986... Training loss: 1.6822... 0.0633 sec/batch
Epoch: 4/20... Training Step: 13987... Training loss: 1.5991... 0.0580 sec/batch
Epoch: 4/20... Training Step: 13988... Training loss: 1.6715... 0.0666 sec/batch
Epoch: 4/20... Training Step: 13989... Training loss: 1.7096... 0.0591 sec/batch
Epoch: 4/20... Training Step: 13990... Training loss: 1.6052... 0.0555 sec/batch
Epoch: 4/20... Training Step: 13991... Training loss: 1.6800... 0.0583 sec/batch
Epoch: 4/20... Training Step: 13992... Training loss: 1.6401... 0.0608 sec/batch
Epoch: 4/20... Training Step: 13993... Training loss: 1.6828... 0.0650 sec/batch
Epoch: 4/20... Training Step: 13994... Training loss: 1.6376... 0.0603 sec/batch
Epoch: 4/20... Training Step: 13995... Training loss: 1.6131... 0.0627 sec/batch
Epoch: 4/20... Training Step: 13996... Training loss: 1.6670... 0.0703 sec/batch
Epoch: 4/20... Training Step: 13997... Training loss: 1.7152... 0.0646 sec/batch
Epoch: 4/20... Training Step: 13998... Training loss: 1.8035... 0.0595 sec/batch
Epoch: 4/20... Training Step: 13999... Training loss: 1.6197... 0.0649 sec/batch
Epoch: 4/20... Training Step: 14000... Training loss: 1.6477... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14001... Training loss: 1.7632... 0.0574 sec/batch
Epoch: 4/20... Training Step: 14002... Training loss: 1.5718... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14003... Training loss: 1.5784... 0.0753 sec/batch
Epoch: 4/20... Training Step: 14004... Training loss: 1.6244... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14005... Training loss: 1.7040... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14006... Training loss: 1.6572... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14007... Training loss: 1.6377... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14008... Training loss: 1.6188... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14009... Training loss: 1.5098... 0.0561 sec/batch
Epoch: 4/20... Training Step: 14010... Training loss: 1.6484... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14011... Training loss: 1.5863... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14012... Training loss: 1.6244... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14013... Training loss: 1.7635... 0.0611 sec/batch
Epoch: 4/20... Training Step: 14014... Training loss: 1.7378... 0.0570 sec/batch
Epoch: 4/20... Training Step: 14015... Training loss: 1.7768... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14016... Training loss: 1.7720... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14017... Training loss: 1.6419... 0.0597 sec/batch
Epoch: 4/20... Training Step: 14018... Training loss: 1.7145... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14019... Training loss: 1.6082... 0.0639 sec/batch
Epoch: 4/20... Training Step: 14020... Training loss: 1.6477... 0.0585 sec/batch
Epoch: 4/20... Training Step: 14021... Training loss: 1.5655... 0.0570 sec/batch
Epoch: 4/20... Training Step: 14022... Training loss: 1.6898... 0.0542 sec/batch
Epoch: 4/20... Training Step: 14023... Training loss: 1.6521... 0.0684 sec/batch
Epoch: 4/20... Training Step: 14024... Training loss: 1.5876... 0.0577 sec/batch
Epoch: 4/20... Training Step: 14025... Training loss: 1.5429... 0.0617 sec/batch
Epoch: 4/20... Training Step: 14026... Training loss: 1.6194... 0.0564 sec/batch
Epoch: 4/20... Training Step: 14027... Training loss: 1.6261... 0.0590 sec/batch
Epoch: 4/20... Training Step: 14028... Training loss: 1.6307... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14029... Training loss: 1.5714... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14030... Training loss: 1.5238... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14031... Training loss: 1.7530... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14032... Training loss: 1.7962... 0.0582 sec/batch
Epoch: 4/20... Training Step: 14033... Training loss: 1.6335... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14034... Training loss: 1.6482... 0.0646 sec/batch
Epoch: 4/20... Training Step: 14035... Training loss: 1.7090... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14036... Training loss: 1.6803... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14037... Training loss: 1.7853... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14038... Training loss: 1.5456... 0.0565 sec/batch
Epoch: 4/20... Training Step: 14039... Training loss: 1.5391... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14040... Training loss: 1.7661... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14041... Training loss: 1.6291... 0.0663 sec/batch
Epoch: 4/20... Training Step: 14042... Training loss: 1.6005... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14043... Training loss: 1.7606... 0.0781 sec/batch
Epoch: 4/20... Training Step: 14044... Training loss: 1.6573... 0.0566 sec/batch
Epoch: 4/20... Training Step: 14045... Training loss: 1.6302... 0.0671 sec/batch
Epoch: 4/20... Training Step: 14046... Training loss: 1.6338... 0.0591 sec/batch
Epoch: 4/20... Training Step: 14047... Training loss: 1.6184... 0.0573 sec/batch
Epoch: 4/20... Training Step: 14048... Training loss: 1.6615... 0.0556 sec/batch
Epoch: 4/20... Training Step: 14049... Training loss: 1.6915... 0.0537 sec/batch
Epoch: 4/20... Training Step: 14050... Training loss: 1.6879... 0.0576 sec/batch
Epoch: 4/20... Training Step: 14051... Training loss: 1.6279... 0.0696 sec/batch
Epoch: 4/20... Training Step: 14052... Training loss: 1.7448... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14053... Training loss: 1.6357... 0.0561 sec/batch
Epoch: 4/20... Training Step: 14054... Training loss: 1.6509... 0.0650 sec/batch
Epoch: 4/20... Training Step: 14055... Training loss: 1.7087... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14056... Training loss: 1.6877... 0.0920 sec/batch
Epoch: 4/20... Training Step: 14057... Training loss: 1.5308... 0.0804 sec/batch
Epoch: 4/20... Training Step: 14058... Training loss: 1.9717... 0.0973 sec/batch
Epoch: 4/20... Training Step: 14059... Training loss: 1.6961... 0.0813 sec/batch
Epoch: 4/20... Training Step: 14060... Training loss: 1.6511... 0.0843 sec/batch
Epoch: 4/20... Training Step: 14061... Training loss: 1.7956... 0.0884 sec/batch
Epoch: 4/20... Training Step: 14062... Training loss: 1.6713... 0.0881 sec/batch
Epoch: 4/20... Training Step: 14063... Training loss: 1.5619... 0.0936 sec/batch
Epoch: 4/20... Training Step: 14064... Training loss: 1.7135... 0.0861 sec/batch
Epoch: 4/20... Training Step: 14065... Training loss: 1.6883... 0.0819 sec/batch
Epoch: 4/20... Training Step: 14066... Training loss: 1.4670... 0.0883 sec/batch
Epoch: 4/20... Training Step: 14067... Training loss: 1.6756... 0.1028 sec/batch
Epoch: 4/20... Training Step: 14068... Training loss: 1.6447... 0.0890 sec/batch
Epoch: 4/20... Training Step: 14069... Training loss: 1.8123... 0.0825 sec/batch
Epoch: 4/20... Training Step: 14070... Training loss: 1.6802... 0.0937 sec/batch
Epoch: 4/20... Training Step: 14071... Training loss: 1.6064... 0.0832 sec/batch
Epoch: 4/20... Training Step: 14072... Training loss: 1.6042... 0.0852 sec/batch
Epoch: 4/20... Training Step: 14073... Training loss: 1.5309... 0.0904 sec/batch
Epoch: 4/20... Training Step: 14074... Training loss: 1.6550... 0.0814 sec/batch
Epoch: 4/20... Training Step: 14075... Training loss: 1.7348... 0.0893 sec/batch
Epoch: 4/20... Training Step: 14076... Training loss: 1.6203... 0.0922 sec/batch
Epoch: 4/20... Training Step: 14077... Training loss: 1.6595... 0.0847 sec/batch
Epoch: 4/20... Training Step: 14078... Training loss: 1.5726... 0.0823 sec/batch
Epoch: 4/20... Training Step: 14079... Training loss: 1.6928... 0.0818 sec/batch
Epoch: 4/20... Training Step: 14080... Training loss: 1.5683... 0.0834 sec/batch
Epoch: 4/20... Training Step: 14081... Training loss: 1.5551... 0.0869 sec/batch
Epoch: 4/20... Training Step: 14082... Training loss: 1.7126... 0.0970 sec/batch
Epoch: 4/20... Training Step: 14083... Training loss: 1.8268... 0.0866 sec/batch
Epoch: 4/20... Training Step: 14084... Training loss: 1.5961... 0.0880 sec/batch
Epoch: 4/20... Training Step: 14085... Training loss: 1.6620... 0.0865 sec/batch
Epoch: 4/20... Training Step: 14086... Training loss: 1.7406... 0.0959 sec/batch
Epoch: 4/20... Training Step: 14087... Training loss: 1.6644... 0.0877 sec/batch
Epoch: 4/20... Training Step: 14088... Training loss: 1.7440... 0.1035 sec/batch
Epoch: 4/20... Training Step: 14089... Training loss: 1.6965... 0.0894 sec/batch
Epoch: 4/20... Training Step: 14090... Training loss: 1.7188... 0.0773 sec/batch
Epoch: 4/20... Training Step: 14091... Training loss: 1.5917... 0.1095 sec/batch
Epoch: 4/20... Training Step: 14092... Training loss: 1.8986... 0.0837 sec/batch
Epoch: 4/20... Training Step: 14093... Training loss: 1.6372... 0.1023 sec/batch
Epoch: 4/20... Training Step: 14094... Training loss: 1.6732... 0.0812 sec/batch
Epoch: 4/20... Training Step: 14095... Training loss: 1.7394... 0.0974 sec/batch
Epoch: 4/20... Training Step: 14096... Training loss: 1.6988... 0.0778 sec/batch
Epoch: 4/20... Training Step: 14097... Training loss: 1.7400... 0.0894 sec/batch
Epoch: 4/20... Training Step: 14098... Training loss: 1.5959... 0.0822 sec/batch
Epoch: 4/20... Training Step: 14099... Training loss: 1.7097... 0.0961 sec/batch
Epoch: 4/20... Training Step: 14100... Training loss: 1.6751... 0.0925 sec/batch
Epoch: 4/20... Training Step: 14101... Training loss: 1.7423... 0.0690 sec/batch
Epoch: 4/20... Training Step: 14102... Training loss: 1.5940... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14103... Training loss: 1.5700... 0.0798 sec/batch
Epoch: 4/20... Training Step: 14104... Training loss: 1.6183... 0.0694 sec/batch
Epoch: 4/20... Training Step: 14105... Training loss: 1.6607... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14106... Training loss: 1.7572... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14107... Training loss: 1.6418... 0.0867 sec/batch
Epoch: 4/20... Training Step: 14108... Training loss: 1.7516... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14109... Training loss: 1.6682... 0.0694 sec/batch
Epoch: 4/20... Training Step: 14110... Training loss: 1.6734... 0.0718 sec/batch
Epoch: 4/20... Training Step: 14111... Training loss: 1.8364... 0.0665 sec/batch
Epoch: 4/20... Training Step: 14112... Training loss: 1.7939... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14113... Training loss: 1.8560... 0.0674 sec/batch
Epoch: 4/20... Training Step: 14114... Training loss: 1.5663... 0.0735 sec/batch
Epoch: 4/20... Training Step: 14115... Training loss: 1.6455... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14116... Training loss: 1.6750... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14117... Training loss: 1.7532... 0.0679 sec/batch
Epoch: 4/20... Training Step: 14118... Training loss: 1.6519... 0.0696 sec/batch
Epoch: 4/20... Training Step: 14119... Training loss: 1.6734... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14120... Training loss: 1.7280... 0.0821 sec/batch
Epoch: 4/20... Training Step: 14121... Training loss: 1.7352... 0.0680 sec/batch
Epoch: 4/20... Training Step: 14122... Training loss: 1.5254... 0.0633 sec/batch
Epoch: 4/20... Training Step: 14123... Training loss: 1.6470... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14124... Training loss: 1.7492... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14125... Training loss: 1.7452... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14126... Training loss: 1.9510... 0.0676 sec/batch
Epoch: 4/20... Training Step: 14127... Training loss: 1.7934... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14128... Training loss: 1.6051... 0.0646 sec/batch
Epoch: 4/20... Training Step: 14129... Training loss: 1.6756... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14130... Training loss: 1.7188... 0.0675 sec/batch
Epoch: 4/20... Training Step: 14131... Training loss: 1.6305... 0.0675 sec/batch
Epoch: 4/20... Training Step: 14132... Training loss: 1.6936... 0.0734 sec/batch
Epoch: 4/20... Training Step: 14133... Training loss: 1.5571... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14134... Training loss: 1.4767... 0.0733 sec/batch
Epoch: 4/20... Training Step: 14135... Training loss: 1.8438... 0.0821 sec/batch
Epoch: 4/20... Training Step: 14136... Training loss: 1.4579... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14137... Training loss: 1.5940... 0.0798 sec/batch
Epoch: 4/20... Training Step: 14138... Training loss: 1.7158... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14139... Training loss: 1.6505... 0.0640 sec/batch
Epoch: 4/20... Training Step: 14140... Training loss: 1.6688... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14141... Training loss: 1.7365... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14142... Training loss: 1.6750... 0.0677 sec/batch
Epoch: 4/20... Training Step: 14143... Training loss: 1.6308... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14144... Training loss: 1.6493... 0.0676 sec/batch
Epoch: 4/20... Training Step: 14145... Training loss: 1.5800... 0.0689 sec/batch
Epoch: 4/20... Training Step: 14146... Training loss: 1.4208... 0.0706 sec/batch
Epoch: 4/20... Training Step: 14147... Training loss: 1.6631... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14148... Training loss: 1.5649... 0.0669 sec/batch
Epoch: 4/20... Training Step: 14149... Training loss: 1.7340... 0.0860 sec/batch
Epoch: 4/20... Training Step: 14150... Training loss: 1.9039... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14151... Training loss: 1.7335... 0.0697 sec/batch
Epoch: 4/20... Training Step: 14152... Training loss: 1.6915... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14153... Training loss: 1.7712... 0.0646 sec/batch
Epoch: 4/20... Training Step: 14154... Training loss: 1.7590... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14155... Training loss: 2.0782... 0.0752 sec/batch
Epoch: 4/20... Training Step: 14156... Training loss: 1.7995... 0.0688 sec/batch
Epoch: 4/20... Training Step: 14157... Training loss: 1.6007... 0.0692 sec/batch
Epoch: 4/20... Training Step: 14158... Training loss: 1.7318... 0.0713 sec/batch
Epoch: 4/20... Training Step: 14159... Training loss: 1.5443... 0.0669 sec/batch
Epoch: 4/20... Training Step: 14160... Training loss: 1.7095... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14161... Training loss: 1.5829... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14162... Training loss: 1.7445... 0.0678 sec/batch
Epoch: 4/20... Training Step: 14163... Training loss: 1.5171... 0.0591 sec/batch
Epoch: 4/20... Training Step: 14164... Training loss: 1.6886... 0.0670 sec/batch
Epoch: 4/20... Training Step: 14165... Training loss: 1.6323... 0.0569 sec/batch
Epoch: 4/20... Training Step: 14166... Training loss: 1.6413... 0.0729 sec/batch
Epoch: 4/20... Training Step: 14167... Training loss: 1.6827... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14168... Training loss: 1.6454... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14169... Training loss: 1.5416... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14170... Training loss: 1.8500... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14171... Training loss: 1.8573... 0.0751 sec/batch
Epoch: 4/20... Training Step: 14172... Training loss: 1.5077... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14173... Training loss: 1.5701... 0.0566 sec/batch
Epoch: 4/20... Training Step: 14174... Training loss: 1.7067... 0.0718 sec/batch
Epoch: 4/20... Training Step: 14175... Training loss: 1.6411... 0.0701 sec/batch
Epoch: 4/20... Training Step: 14176... Training loss: 1.6173... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14177... Training loss: 1.6238... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14178... Training loss: 1.7821... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14179... Training loss: 1.6879... 0.0705 sec/batch
Epoch: 4/20... Training Step: 14180... Training loss: 1.5703... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14181... Training loss: 1.6235... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14182... Training loss: 1.7674... 0.0655 sec/batch
Epoch: 4/20... Training Step: 14183... Training loss: 1.5162... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14184... Training loss: 1.5528... 0.0703 sec/batch
Epoch: 4/20... Training Step: 14185... Training loss: 1.4848... 0.0574 sec/batch
Epoch: 4/20... Training Step: 14186... Training loss: 1.5852... 0.0653 sec/batch
Epoch: 4/20... Training Step: 14187... Training loss: 1.6171... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14188... Training loss: 1.6906... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14189... Training loss: 1.7725... 0.0536 sec/batch
Epoch: 4/20... Training Step: 14190... Training loss: 1.6171... 0.0744 sec/batch
Epoch: 4/20... Training Step: 14191... Training loss: 1.6786... 0.0535 sec/batch
Epoch: 4/20... Training Step: 14192... Training loss: 1.6340... 0.0652 sec/batch
Epoch: 4/20... Training Step: 14193... Training loss: 1.7001... 0.0544 sec/batch
Epoch: 4/20... Training Step: 14194... Training loss: 1.5341... 0.0765 sec/batch
Epoch: 4/20... Training Step: 14195... Training loss: 1.5892... 0.0622 sec/batch
Epoch: 4/20... Training Step: 14196... Training loss: 1.5195... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14197... Training loss: 1.5957... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14198... Training loss: 1.5088... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14199... Training loss: 1.6336... 0.0672 sec/batch
Epoch: 4/20... Training Step: 14200... Training loss: 1.6572... 0.0548 sec/batch
Epoch: 4/20... Training Step: 14201... Training loss: 1.6054... 0.0691 sec/batch
Epoch: 4/20... Training Step: 14202... Training loss: 1.6084... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14203... Training loss: 1.5563... 0.0589 sec/batch
Epoch: 4/20... Training Step: 14204... Training loss: 1.6555... 0.0502 sec/batch
Epoch: 4/20... Training Step: 14205... Training loss: 1.6681... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14206... Training loss: 1.5688... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14207... Training loss: 1.6395... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14208... Training loss: 1.5641... 0.0534 sec/batch
Epoch: 4/20... Training Step: 14209... Training loss: 1.7723... 0.0731 sec/batch
Epoch: 4/20... Training Step: 14210... Training loss: 1.5156... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14211... Training loss: 1.7572... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14212... Training loss: 1.7380... 0.0538 sec/batch
Epoch: 4/20... Training Step: 14213... Training loss: 1.6751... 0.0617 sec/batch
Epoch: 4/20... Training Step: 14214... Training loss: 1.5821... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14215... Training loss: 1.6758... 0.0602 sec/batch
Epoch: 4/20... Training Step: 14216... Training loss: 1.7991... 0.0598 sec/batch
Epoch: 4/20... Training Step: 14217... Training loss: 1.6444... 0.0745 sec/batch
Epoch: 4/20... Training Step: 14218... Training loss: 1.7884... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14219... Training loss: 1.6151... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14220... Training loss: 1.7177... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14221... Training loss: 1.5667... 0.0594 sec/batch
Epoch: 4/20... Training Step: 14222... Training loss: 1.5698... 0.0548 sec/batch
Epoch: 4/20... Training Step: 14223... Training loss: 1.7352... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14224... Training loss: 1.6918... 0.0597 sec/batch
Epoch: 4/20... Training Step: 14225... Training loss: 1.5816... 0.0649 sec/batch
Epoch: 4/20... Training Step: 14226... Training loss: 1.6773... 0.0573 sec/batch
Epoch: 4/20... Training Step: 14227... Training loss: 1.5120... 0.0523 sec/batch
Epoch: 4/20... Training Step: 14228... Training loss: 1.6618... 0.0524 sec/batch
Epoch: 4/20... Training Step: 14229... Training loss: 1.7127... 0.0518 sec/batch
Epoch: 4/20... Training Step: 14230... Training loss: 1.6341... 0.0700 sec/batch
Epoch: 4/20... Training Step: 14231... Training loss: 1.6173... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14232... Training loss: 1.5792... 0.0595 sec/batch
Epoch: 4/20... Training Step: 14233... Training loss: 1.5803... 0.0689 sec/batch
Epoch: 4/20... Training Step: 14234... Training loss: 1.7357... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14235... Training loss: 1.8101... 0.0570 sec/batch
Epoch: 4/20... Training Step: 14236... Training loss: 1.7377... 0.0514 sec/batch
Epoch: 4/20... Training Step: 14237... Training loss: 1.7019... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14238... Training loss: 1.6840... 0.0573 sec/batch
Epoch: 4/20... Training Step: 14239... Training loss: 1.7825... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14240... Training loss: 1.5714... 0.0535 sec/batch
Epoch: 4/20... Training Step: 14241... Training loss: 1.5112... 0.0649 sec/batch
Epoch: 4/20... Training Step: 14242... Training loss: 1.7062... 0.0539 sec/batch
Epoch: 4/20... Training Step: 14243... Training loss: 1.5214... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14244... Training loss: 1.5536... 0.0492 sec/batch
Epoch: 4/20... Training Step: 14245... Training loss: 1.6348... 0.0525 sec/batch
Epoch: 4/20... Training Step: 14246... Training loss: 1.7021... 0.0574 sec/batch
Epoch: 4/20... Training Step: 14247... Training loss: 1.8843... 0.0516 sec/batch
Epoch: 4/20... Training Step: 14248... Training loss: 1.8707... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14249... Training loss: 1.6807... 0.0564 sec/batch
Epoch: 4/20... Training Step: 14250... Training loss: 1.7296... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14251... Training loss: 1.6445... 0.0602 sec/batch
Epoch: 4/20... Training Step: 14252... Training loss: 1.5704... 0.0571 sec/batch
Epoch: 4/20... Training Step: 14253... Training loss: 1.6150... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14254... Training loss: 1.5822... 0.0493 sec/batch
Epoch: 4/20... Training Step: 14255... Training loss: 1.6322... 0.0556 sec/batch
Epoch: 4/20... Training Step: 14256... Training loss: 1.4507... 0.0556 sec/batch
Epoch: 4/20... Training Step: 14257... Training loss: 1.6391... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14258... Training loss: 1.6398... 0.0589 sec/batch
Epoch: 4/20... Training Step: 14259... Training loss: 1.6637... 0.0622 sec/batch
Epoch: 4/20... Training Step: 14260... Training loss: 1.7529... 0.0514 sec/batch
Epoch: 4/20... Training Step: 14261... Training loss: 1.9021... 0.0533 sec/batch
Epoch: 4/20... Training Step: 14262... Training loss: 1.7200... 0.0553 sec/batch
Epoch: 4/20... Training Step: 14263... Training loss: 1.7889... 0.0533 sec/batch
Epoch: 4/20... Training Step: 14264... Training loss: 1.6745... 0.0577 sec/batch
Epoch: 4/20... Training Step: 14265... Training loss: 1.6812... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14266... Training loss: 1.6892... 0.0573 sec/batch
Epoch: 4/20... Training Step: 14267... Training loss: 1.6762... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14268... Training loss: 1.6842... 0.0598 sec/batch
Epoch: 4/20... Training Step: 14269... Training loss: 1.6468... 0.0652 sec/batch
Epoch: 4/20... Training Step: 14270... Training loss: 1.8146... 0.0577 sec/batch
Epoch: 4/20... Training Step: 14271... Training loss: 1.6667... 0.0524 sec/batch
Epoch: 4/20... Training Step: 14272... Training loss: 1.8273... 0.0611 sec/batch
Epoch: 4/20... Training Step: 14273... Training loss: 1.6558... 0.0567 sec/batch
Epoch: 4/20... Training Step: 14274... Training loss: 1.7959... 0.0536 sec/batch
Epoch: 4/20... Training Step: 14275... Training loss: 1.6313... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14276... Training loss: 1.5764... 0.0578 sec/batch
Epoch: 4/20... Training Step: 14277... Training loss: 1.6448... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14278... Training loss: 1.6256... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14279... Training loss: 1.6974... 0.0553 sec/batch
Epoch: 4/20... Training Step: 14280... Training loss: 1.6630... 0.0544 sec/batch
Epoch: 4/20... Training Step: 14281... Training loss: 1.6625... 0.0598 sec/batch
Epoch: 4/20... Training Step: 14282... Training loss: 1.7023... 0.0515 sec/batch
Epoch: 4/20... Training Step: 14283... Training loss: 1.6232... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14284... Training loss: 1.6699... 0.0498 sec/batch
Epoch: 4/20... Training Step: 14285... Training loss: 1.6972... 0.0552 sec/batch
Epoch: 4/20... Training Step: 14286... Training loss: 1.7368... 0.0534 sec/batch
Epoch: 4/20... Training Step: 14287... Training loss: 1.7552... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14288... Training loss: 1.6463... 0.0569 sec/batch
Epoch: 4/20... Training Step: 14289... Training loss: 1.7234... 0.0562 sec/batch
Epoch: 4/20... Training Step: 14290... Training loss: 1.7098... 0.0550 sec/batch
Epoch: 4/20... Training Step: 14291... Training loss: 1.6128... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14292... Training loss: 1.7155... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14293... Training loss: 1.7533... 0.0697 sec/batch
Epoch: 4/20... Training Step: 14294... Training loss: 1.7085... 0.0611 sec/batch
Epoch: 4/20... Training Step: 14295... Training loss: 1.7101... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14296... Training loss: 1.5900... 0.0610 sec/batch
Epoch: 4/20... Training Step: 14297... Training loss: 1.5276... 0.0724 sec/batch
Epoch: 4/20... Training Step: 14298... Training loss: 1.7248... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14299... Training loss: 1.7032... 0.0590 sec/batch
Epoch: 4/20... Training Step: 14300... Training loss: 1.7049... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14301... Training loss: 1.7556... 0.0700 sec/batch
Epoch: 4/20... Training Step: 14302... Training loss: 1.7433... 0.0528 sec/batch
Epoch: 4/20... Training Step: 14303... Training loss: 1.8500... 0.0589 sec/batch
Epoch: 4/20... Training Step: 14304... Training loss: 1.7519... 0.0699 sec/batch
Epoch: 4/20... Training Step: 14305... Training loss: 1.6626... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14306... Training loss: 1.7233... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14307... Training loss: 1.6554... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14308... Training loss: 1.6473... 0.0604 sec/batch
Epoch: 4/20... Training Step: 14309... Training loss: 1.7021... 0.0757 sec/batch
Epoch: 4/20... Training Step: 14310... Training loss: 1.7342... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14311... Training loss: 1.6834... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14312... Training loss: 1.5957... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14313... Training loss: 1.6186... 0.0686 sec/batch
Epoch: 4/20... Training Step: 14314... Training loss: 1.6519... 0.0536 sec/batch
Epoch: 4/20... Training Step: 14315... Training loss: 1.5499... 0.0693 sec/batch
Epoch: 4/20... Training Step: 14316... Training loss: 1.6773... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14317... Training loss: 1.6022... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14318... Training loss: 1.6125... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14319... Training loss: 1.5377... 0.0671 sec/batch
Epoch: 4/20... Training Step: 14320... Training loss: 1.6280... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14321... Training loss: 1.6850... 0.0669 sec/batch
Epoch: 4/20... Training Step: 14322... Training loss: 1.6636... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14323... Training loss: 1.7650... 0.0598 sec/batch
Epoch: 4/20... Training Step: 14324... Training loss: 1.5794... 0.0588 sec/batch
Epoch: 4/20... Training Step: 14325... Training loss: 1.6752... 0.0652 sec/batch
Epoch: 4/20... Training Step: 14326... Training loss: 1.5697... 0.0570 sec/batch
Epoch: 4/20... Training Step: 14327... Training loss: 1.5935... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14328... Training loss: 1.6170... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14329... Training loss: 1.6343... 0.0631 sec/batch
Epoch: 4/20... Training Step: 14330... Training loss: 1.8080... 0.0682 sec/batch
Epoch: 4/20... Training Step: 14331... Training loss: 1.7249... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14332... Training loss: 1.7209... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14333... Training loss: 1.7050... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14334... Training loss: 1.6606... 0.0578 sec/batch
Epoch: 4/20... Training Step: 14335... Training loss: 1.5708... 0.0588 sec/batch
Epoch: 4/20... Training Step: 14336... Training loss: 1.6577... 0.0587 sec/batch
Epoch: 4/20... Training Step: 14337... Training loss: 1.7385... 0.0696 sec/batch
Epoch: 4/20... Training Step: 14338... Training loss: 1.7805... 0.0640 sec/batch
Epoch: 4/20... Training Step: 14339... Training loss: 1.7423... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14340... Training loss: 1.6850... 0.0558 sec/batch
Epoch: 4/20... Training Step: 14341... Training loss: 1.6969... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14342... Training loss: 1.8331... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14343... Training loss: 1.7109... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14344... Training loss: 1.6788... 0.0522 sec/batch
Epoch: 4/20... Training Step: 14345... Training loss: 1.5547... 0.0671 sec/batch
Epoch: 4/20... Training Step: 14346... Training loss: 1.7491... 0.0639 sec/batch
Epoch: 4/20... Training Step: 14347... Training loss: 1.6550... 0.0597 sec/batch
Epoch: 4/20... Training Step: 14348... Training loss: 1.6864... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14349... Training loss: 1.6997... 0.0788 sec/batch
Epoch: 4/20... Training Step: 14350... Training loss: 1.7002... 0.0573 sec/batch
Epoch: 4/20... Training Step: 14351... Training loss: 1.6588... 0.0613 sec/batch
Epoch: 4/20... Training Step: 14352... Training loss: 1.6850... 0.0714 sec/batch
Epoch: 4/20... Training Step: 14353... Training loss: 1.6037... 0.0968 sec/batch
Epoch: 4/20... Training Step: 14354... Training loss: 1.8079... 0.0909 sec/batch
Epoch: 4/20... Training Step: 14355... Training loss: 1.6795... 0.0932 sec/batch
Epoch: 4/20... Training Step: 14356... Training loss: 1.6251... 0.0897 sec/batch
Epoch: 4/20... Training Step: 14357... Training loss: 1.7854... 0.0888 sec/batch
Epoch: 4/20... Training Step: 14358... Training loss: 1.5595... 0.0857 sec/batch
Epoch: 4/20... Training Step: 14359... Training loss: 1.6386... 0.0953 sec/batch
Epoch: 4/20... Training Step: 14360... Training loss: 1.7625... 0.0845 sec/batch
Epoch: 4/20... Training Step: 14361... Training loss: 1.9468... 0.0911 sec/batch
Epoch: 4/20... Training Step: 14362... Training loss: 1.6652... 0.0928 sec/batch
Epoch: 4/20... Training Step: 14363... Training loss: 1.7530... 0.0959 sec/batch
Epoch: 4/20... Training Step: 14364... Training loss: 1.6137... 0.0814 sec/batch
Epoch: 4/20... Training Step: 14365... Training loss: 1.5770... 0.1021 sec/batch
Epoch: 4/20... Training Step: 14366... Training loss: 1.7143... 0.0837 sec/batch
Epoch: 4/20... Training Step: 14367... Training loss: 1.8091... 0.0851 sec/batch
Epoch: 4/20... Training Step: 14368... Training loss: 1.7405... 0.0764 sec/batch
Epoch: 4/20... Training Step: 14369... Training loss: 1.8450... 0.0804 sec/batch
Epoch: 4/20... Training Step: 14370... Training loss: 1.7244... 0.0899 sec/batch
Epoch: 4/20... Training Step: 14371... Training loss: 1.5649... 0.0927 sec/batch
Epoch: 4/20... Training Step: 14372... Training loss: 1.7088... 0.0816 sec/batch
Epoch: 4/20... Training Step: 14373... Training loss: 1.5374... 0.0878 sec/batch
Epoch: 4/20... Training Step: 14374... Training loss: 1.4948... 0.0913 sec/batch
Epoch: 4/20... Training Step: 14375... Training loss: 1.7205... 0.0851 sec/batch
Epoch: 4/20... Training Step: 14376... Training loss: 1.6162... 0.0864 sec/batch
Epoch: 4/20... Training Step: 14377... Training loss: 1.7899... 0.0822 sec/batch
Epoch: 4/20... Training Step: 14378... Training loss: 1.6855... 0.0924 sec/batch
Epoch: 4/20... Training Step: 14379... Training loss: 1.6313... 0.0805 sec/batch
Epoch: 4/20... Training Step: 14380... Training loss: 1.7030... 0.0929 sec/batch
Epoch: 4/20... Training Step: 14381... Training loss: 1.6423... 0.0842 sec/batch
Epoch: 4/20... Training Step: 14382... Training loss: 1.6545... 0.0830 sec/batch
Epoch: 4/20... Training Step: 14383... Training loss: 1.5446... 0.0938 sec/batch
Epoch: 4/20... Training Step: 14384... Training loss: 1.6198... 0.0812 sec/batch
Epoch: 4/20... Training Step: 14385... Training loss: 1.4772... 0.0822 sec/batch
Epoch: 4/20... Training Step: 14386... Training loss: 1.5977... 0.0970 sec/batch
Epoch: 4/20... Training Step: 14387... Training loss: 1.5937... 0.0827 sec/batch
Epoch: 4/20... Training Step: 14388... Training loss: 1.5974... 0.0968 sec/batch
Epoch: 4/20... Training Step: 14389... Training loss: 1.6436... 0.0899 sec/batch
Epoch: 4/20... Training Step: 14390... Training loss: 1.7148... 0.0872 sec/batch
Epoch: 4/20... Training Step: 14391... Training loss: 1.5069... 0.0881 sec/batch
Epoch: 4/20... Training Step: 14392... Training loss: 1.6322... 0.0923 sec/batch
Epoch: 4/20... Training Step: 14393... Training loss: 1.8571... 0.0848 sec/batch
Epoch: 4/20... Training Step: 14394... Training loss: 1.6243... 0.0893 sec/batch
Epoch: 4/20... Training Step: 14395... Training loss: 1.6390... 0.0916 sec/batch
Epoch: 4/20... Training Step: 14396... Training loss: 1.6229... 0.0785 sec/batch
Epoch: 4/20... Training Step: 14397... Training loss: 1.6049... 0.0815 sec/batch
Epoch: 4/20... Training Step: 14398... Training loss: 1.5376... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14399... Training loss: 1.7277... 0.0741 sec/batch
Epoch: 4/20... Training Step: 14400... Training loss: 1.7323... 0.0781 sec/batch
Epoch: 4/20... Training Step: 14401... Training loss: 1.7671... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14402... Training loss: 1.7116... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14403... Training loss: 1.6254... 0.0717 sec/batch
Epoch: 4/20... Training Step: 14404... Training loss: 1.6488... 0.0788 sec/batch
Epoch: 4/20... Training Step: 14405... Training loss: 1.6546... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14406... Training loss: 1.5936... 0.0741 sec/batch
Epoch: 4/20... Training Step: 14407... Training loss: 1.5689... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14408... Training loss: 1.6494... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14409... Training loss: 1.7211... 0.0617 sec/batch
Epoch: 4/20... Training Step: 14410... Training loss: 1.6224... 0.0799 sec/batch
Epoch: 4/20... Training Step: 14411... Training loss: 1.5827... 0.0692 sec/batch
Epoch: 4/20... Training Step: 14412... Training loss: 1.5264... 0.0650 sec/batch
Epoch: 4/20... Training Step: 14413... Training loss: 1.5832... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14414... Training loss: 1.5477... 0.0754 sec/batch
Epoch: 4/20... Training Step: 14415... Training loss: 1.6768... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14416... Training loss: 1.6389... 0.0676 sec/batch
Epoch: 4/20... Training Step: 14417... Training loss: 1.5390... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14418... Training loss: 1.6568... 0.0705 sec/batch
Epoch: 4/20... Training Step: 14419... Training loss: 1.7931... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14420... Training loss: 1.7302... 0.0549 sec/batch
Epoch: 4/20... Training Step: 14421... Training loss: 1.6659... 0.0771 sec/batch
Epoch: 4/20... Training Step: 14422... Training loss: 1.5227... 0.0577 sec/batch
Epoch: 4/20... Training Step: 14423... Training loss: 1.5774... 0.0650 sec/batch
Epoch: 4/20... Training Step: 14424... Training loss: 1.7046... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14425... Training loss: 1.5468... 0.0715 sec/batch
Epoch: 4/20... Training Step: 14426... Training loss: 1.6899... 0.0696 sec/batch
Epoch: 4/20... Training Step: 14427... Training loss: 1.6589... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14428... Training loss: 1.7011... 0.0695 sec/batch
Epoch: 4/20... Training Step: 14429... Training loss: 1.7537... 0.0745 sec/batch
Epoch: 4/20... Training Step: 14430... Training loss: 1.7873... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14431... Training loss: 1.6212... 0.0678 sec/batch
Epoch: 4/20... Training Step: 14432... Training loss: 1.6785... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14433... Training loss: 1.6082... 0.0671 sec/batch
Epoch: 4/20... Training Step: 14434... Training loss: 1.7094... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14435... Training loss: 1.7348... 0.0757 sec/batch
Epoch: 4/20... Training Step: 14436... Training loss: 1.5871... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14437... Training loss: 1.7645... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14438... Training loss: 1.6458... 0.0653 sec/batch
Epoch: 4/20... Training Step: 14439... Training loss: 1.7402... 0.0816 sec/batch
Epoch: 4/20... Training Step: 14440... Training loss: 1.6214... 0.0670 sec/batch
Epoch: 4/20... Training Step: 14441... Training loss: 1.5732... 0.0733 sec/batch
Epoch: 4/20... Training Step: 14442... Training loss: 1.5556... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14443... Training loss: 1.6737... 0.0639 sec/batch
Epoch: 4/20... Training Step: 14444... Training loss: 1.5733... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14445... Training loss: 1.6561... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14446... Training loss: 1.6684... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14447... Training loss: 1.6840... 0.0688 sec/batch
Epoch: 4/20... Training Step: 14448... Training loss: 1.7020... 0.0678 sec/batch
Epoch: 4/20... Training Step: 14449... Training loss: 1.7499... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14450... Training loss: 1.7151... 0.0733 sec/batch
Epoch: 4/20... Training Step: 14451... Training loss: 1.7211... 0.0668 sec/batch
Epoch: 4/20... Training Step: 14452... Training loss: 1.7224... 0.0631 sec/batch
Epoch: 4/20... Training Step: 14453... Training loss: 1.5770... 0.0764 sec/batch
Epoch: 4/20... Training Step: 14454... Training loss: 1.7047... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14455... Training loss: 1.6988... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14456... Training loss: 1.6060... 0.0697 sec/batch
Epoch: 4/20... Training Step: 14457... Training loss: 1.6910... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14458... Training loss: 1.6445... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14459... Training loss: 1.6289... 0.0637 sec/batch
Epoch: 4/20... Training Step: 14460... Training loss: 1.7060... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14461... Training loss: 1.5490... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14462... Training loss: 1.7570... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14463... Training loss: 1.5431... 0.0678 sec/batch
Epoch: 4/20... Training Step: 14464... Training loss: 1.5862... 0.0631 sec/batch
Epoch: 4/20... Training Step: 14465... Training loss: 1.6717... 0.0639 sec/batch
Epoch: 4/20... Training Step: 14466... Training loss: 1.7379... 0.0585 sec/batch
Epoch: 4/20... Training Step: 14467... Training loss: 1.6198... 0.0729 sec/batch
Epoch: 4/20... Training Step: 14468... Training loss: 1.6363... 0.0632 sec/batch
Epoch: 4/20... Training Step: 14469... Training loss: 1.7318... 0.0602 sec/batch
Epoch: 4/20... Training Step: 14470... Training loss: 1.5445... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14471... Training loss: 1.4800... 0.0677 sec/batch
Epoch: 4/20... Training Step: 14472... Training loss: 1.6359... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14473... Training loss: 1.6239... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14474... Training loss: 1.4604... 0.0675 sec/batch
Epoch: 4/20... Training Step: 14475... Training loss: 1.5179... 0.0739 sec/batch
Epoch: 4/20... Training Step: 14476... Training loss: 1.6490... 0.0738 sec/batch
Epoch: 4/20... Training Step: 14477... Training loss: 1.5644... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14478... Training loss: 1.6600... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14479... Training loss: 1.5632... 0.0701 sec/batch
Epoch: 4/20... Training Step: 14480... Training loss: 1.5652... 0.0638 sec/batch
Epoch: 4/20... Training Step: 14481... Training loss: 1.6095... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14482... Training loss: 1.6844... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14483... Training loss: 1.6396... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14484... Training loss: 1.7026... 0.0617 sec/batch
Epoch: 4/20... Training Step: 14485... Training loss: 1.5522... 0.0593 sec/batch
Epoch: 4/20... Training Step: 14486... Training loss: 1.6361... 0.0714 sec/batch
Epoch: 4/20... Training Step: 14487... Training loss: 1.7509... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14488... Training loss: 1.6765... 0.0587 sec/batch
Epoch: 4/20... Training Step: 14489... Training loss: 1.6666... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14490... Training loss: 1.6923... 0.0740 sec/batch
Epoch: 4/20... Training Step: 14491... Training loss: 1.5858... 0.0576 sec/batch
Epoch: 4/20... Training Step: 14492... Training loss: 1.8134... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14493... Training loss: 1.5300... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14494... Training loss: 1.6299... 0.0638 sec/batch
Epoch: 4/20... Training Step: 14495... Training loss: 1.6567... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14496... Training loss: 1.8256... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14497... Training loss: 1.6925... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14498... Training loss: 1.8939... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14499... Training loss: 1.7149... 0.0638 sec/batch
Epoch: 4/20... Training Step: 14500... Training loss: 1.7242... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14501... Training loss: 1.6864... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14502... Training loss: 1.7101... 0.0724 sec/batch
Epoch: 4/20... Training Step: 14503... Training loss: 1.7396... 0.0591 sec/batch
Epoch: 4/20... Training Step: 14504... Training loss: 1.6165... 0.0631 sec/batch
Epoch: 4/20... Training Step: 14505... Training loss: 1.6003... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14506... Training loss: 1.6742... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14507... Training loss: 1.5043... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14508... Training loss: 1.5590... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14509... Training loss: 1.5632... 0.0569 sec/batch
Epoch: 4/20... Training Step: 14510... Training loss: 1.5815... 0.0670 sec/batch
Epoch: 4/20... Training Step: 14511... Training loss: 1.6328... 0.0574 sec/batch
Epoch: 4/20... Training Step: 14512... Training loss: 1.5706... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14513... Training loss: 1.7463... 0.0611 sec/batch
Epoch: 4/20... Training Step: 14514... Training loss: 1.6665... 0.0689 sec/batch
Epoch: 4/20... Training Step: 14515... Training loss: 1.6098... 0.0637 sec/batch
Epoch: 4/20... Training Step: 14516... Training loss: 1.6069... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14517... Training loss: 1.6185... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14518... Training loss: 1.6804... 0.0767 sec/batch
Epoch: 4/20... Training Step: 14519... Training loss: 1.5579... 0.0637 sec/batch
Epoch: 4/20... Training Step: 14520... Training loss: 1.7146... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14521... Training loss: 1.7537... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14522... Training loss: 1.5786... 0.0756 sec/batch
Epoch: 4/20... Training Step: 14523... Training loss: 1.6935... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14524... Training loss: 1.6328... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14525... Training loss: 1.4027... 0.0725 sec/batch
Epoch: 4/20... Training Step: 14526... Training loss: 1.6096... 0.0711 sec/batch
Epoch: 4/20... Training Step: 14527... Training loss: 1.6434... 0.0700 sec/batch
Epoch: 4/20... Training Step: 14528... Training loss: 1.7097... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14529... Training loss: 1.5542... 0.0733 sec/batch
Epoch: 4/20... Training Step: 14530... Training loss: 1.5284... 0.0788 sec/batch
Epoch: 4/20... Training Step: 14531... Training loss: 1.8461... 0.0738 sec/batch
Epoch: 4/20... Training Step: 14532... Training loss: 1.4770... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14533... Training loss: 1.6646... 0.0690 sec/batch
Epoch: 4/20... Training Step: 14534... Training loss: 1.7977... 0.0652 sec/batch
Epoch: 4/20... Training Step: 14535... Training loss: 1.7725... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14536... Training loss: 1.6854... 0.0765 sec/batch
Epoch: 4/20... Training Step: 14537... Training loss: 1.6504... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14538... Training loss: 1.7239... 0.0702 sec/batch
Epoch: 4/20... Training Step: 14539... Training loss: 1.4716... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14540... Training loss: 1.8186... 0.0742 sec/batch
Epoch: 4/20... Training Step: 14541... Training loss: 1.7432... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14542... Training loss: 1.5526... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14543... Training loss: 1.6718... 0.0698 sec/batch
Epoch: 4/20... Training Step: 14544... Training loss: 1.6605... 0.0697 sec/batch
Epoch: 4/20... Training Step: 14545... Training loss: 1.5359... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14546... Training loss: 1.7566... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14547... Training loss: 1.7062... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14548... Training loss: 1.5904... 0.0754 sec/batch
Epoch: 4/20... Training Step: 14549... Training loss: 1.6912... 0.0709 sec/batch
Epoch: 4/20... Training Step: 14550... Training loss: 1.5256... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14551... Training loss: 1.8054... 0.0590 sec/batch
Epoch: 4/20... Training Step: 14552... Training loss: 1.6670... 0.0701 sec/batch
Epoch: 4/20... Training Step: 14553... Training loss: 1.6214... 0.0632 sec/batch
Epoch: 4/20... Training Step: 14554... Training loss: 1.6371... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14555... Training loss: 1.6793... 0.0686 sec/batch
Epoch: 4/20... Training Step: 14556... Training loss: 1.6803... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14557... Training loss: 1.6424... 0.0680 sec/batch
Epoch: 4/20... Training Step: 14558... Training loss: 1.6902... 0.0664 sec/batch
Epoch: 4/20... Training Step: 14559... Training loss: 1.6266... 0.0676 sec/batch
Epoch: 4/20... Training Step: 14560... Training loss: 1.7898... 0.0770 sec/batch
Epoch: 4/20... Training Step: 14561... Training loss: 1.7642... 0.0637 sec/batch
Epoch: 4/20... Training Step: 14562... Training loss: 1.7764... 0.0670 sec/batch
Epoch: 4/20... Training Step: 14563... Training loss: 1.7253... 0.0672 sec/batch
Epoch: 4/20... Training Step: 14564... Training loss: 1.7387... 0.0622 sec/batch
Epoch: 4/20... Training Step: 14565... Training loss: 1.6070... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14566... Training loss: 1.6395... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14567... Training loss: 1.7470... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14568... Training loss: 1.6249... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14569... Training loss: 1.5652... 0.0650 sec/batch
Epoch: 4/20... Training Step: 14570... Training loss: 1.7683... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14571... Training loss: 1.5098... 0.0654 sec/batch
Epoch: 4/20... Training Step: 14572... Training loss: 1.6557... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14573... Training loss: 1.5839... 0.0618 sec/batch
Epoch: 4/20... Training Step: 14574... Training loss: 1.6369... 0.0739 sec/batch
Epoch: 4/20... Training Step: 14575... Training loss: 1.8006... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14576... Training loss: 1.6818... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14577... Training loss: 1.6505... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14578... Training loss: 1.5239... 0.0715 sec/batch
Epoch: 4/20... Training Step: 14579... Training loss: 1.7536... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14580... Training loss: 1.6336... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14581... Training loss: 2.0294... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14582... Training loss: 1.6263... 0.0653 sec/batch
Epoch: 4/20... Training Step: 14583... Training loss: 1.7461... 0.0568 sec/batch
Epoch: 4/20... Training Step: 14584... Training loss: 1.7606... 0.0613 sec/batch
Epoch: 4/20... Training Step: 14585... Training loss: 1.6279... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14586... Training loss: 1.6336... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14587... Training loss: 1.6167... 0.0530 sec/batch
Epoch: 4/20... Training Step: 14588... Training loss: 1.6866... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14589... Training loss: 1.6827... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14590... Training loss: 1.5484... 0.0580 sec/batch
Epoch: 4/20... Training Step: 14591... Training loss: 1.7134... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14592... Training loss: 1.5983... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14593... Training loss: 1.7530... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14594... Training loss: 1.5868... 0.0665 sec/batch
Epoch: 4/20... Training Step: 14595... Training loss: 1.5947... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14596... Training loss: 1.5643... 0.0613 sec/batch
Epoch: 4/20... Training Step: 14597... Training loss: 1.6156... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14598... Training loss: 1.6056... 0.0706 sec/batch
Epoch: 4/20... Training Step: 14599... Training loss: 1.5260... 0.0633 sec/batch
Epoch: 4/20... Training Step: 14600... Training loss: 1.6659... 0.0571 sec/batch
Epoch: 4/20... Training Step: 14601... Training loss: 1.7605... 0.0732 sec/batch
Epoch: 4/20... Training Step: 14602... Training loss: 1.6535... 0.0595 sec/batch
Epoch: 4/20... Training Step: 14603... Training loss: 1.5560... 0.0683 sec/batch
Epoch: 4/20... Training Step: 14604... Training loss: 1.7907... 0.0586 sec/batch
Epoch: 4/20... Training Step: 14605... Training loss: 1.8916... 0.0691 sec/batch
Epoch: 4/20... Training Step: 14606... Training loss: 1.6548... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14607... Training loss: 1.6292... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14608... Training loss: 1.6426... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14609... Training loss: 1.8450... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14610... Training loss: 1.8070... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14611... Training loss: 1.7661... 0.0537 sec/batch
Epoch: 4/20... Training Step: 14612... Training loss: 1.7531... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14613... Training loss: 1.7099... 0.0766 sec/batch
Epoch: 4/20... Training Step: 14614... Training loss: 1.8375... 0.0535 sec/batch
Epoch: 4/20... Training Step: 14615... Training loss: 1.8028... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14616... Training loss: 1.7068... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14617... Training loss: 1.5959... 0.0671 sec/batch
Epoch: 4/20... Training Step: 14618... Training loss: 1.6629... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14619... Training loss: 1.6308... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14620... Training loss: 1.6210... 0.0562 sec/batch
Epoch: 4/20... Training Step: 14621... Training loss: 1.5549... 0.0567 sec/batch
Epoch: 4/20... Training Step: 14622... Training loss: 1.7327... 0.0633 sec/batch
Epoch: 4/20... Training Step: 14623... Training loss: 1.6768... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14624... Training loss: 1.7012... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14625... Training loss: 1.6066... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14626... Training loss: 1.6864... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14627... Training loss: 1.6260... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14628... Training loss: 1.6371... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14629... Training loss: 1.6464... 0.0717 sec/batch
Epoch: 4/20... Training Step: 14630... Training loss: 1.5233... 0.0610 sec/batch
Epoch: 4/20... Training Step: 14631... Training loss: 1.7060... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14632... Training loss: 1.5953... 0.0672 sec/batch
Epoch: 4/20... Training Step: 14633... Training loss: 1.6787... 0.0702 sec/batch
Epoch: 4/20... Training Step: 14634... Training loss: 1.6429... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14635... Training loss: 1.6704... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14636... Training loss: 1.8701... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14637... Training loss: 1.7540... 0.0776 sec/batch
Epoch: 4/20... Training Step: 14638... Training loss: 1.8196... 0.0547 sec/batch
Epoch: 4/20... Training Step: 14639... Training loss: 1.7114... 0.0618 sec/batch
Epoch: 4/20... Training Step: 14640... Training loss: 1.6572... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14641... Training loss: 1.7240... 0.0696 sec/batch
Epoch: 4/20... Training Step: 14642... Training loss: 1.6401... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14643... Training loss: 1.7413... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14644... Training loss: 1.6478... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14645... Training loss: 1.6112... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14646... Training loss: 1.7719... 0.0588 sec/batch
Epoch: 4/20... Training Step: 14647... Training loss: 1.5976... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14648... Training loss: 1.6752... 0.0516 sec/batch
Epoch: 4/20... Training Step: 14649... Training loss: 1.8411... 0.0550 sec/batch
Epoch: 4/20... Training Step: 14650... Training loss: 1.5546... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14651... Training loss: 1.6535... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14652... Training loss: 1.7301... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14653... Training loss: 1.8623... 0.0594 sec/batch
Epoch: 4/20... Training Step: 14654... Training loss: 1.6772... 0.0520 sec/batch
Epoch: 4/20... Training Step: 14655... Training loss: 1.7464... 0.0974 sec/batch
Epoch: 4/20... Training Step: 14656... Training loss: 1.8169... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14657... Training loss: 1.6199... 0.0687 sec/batch
Epoch: 4/20... Training Step: 14658... Training loss: 1.7158... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14659... Training loss: 1.7485... 0.0576 sec/batch
Epoch: 4/20... Training Step: 14660... Training loss: 1.6267... 0.0669 sec/batch
Epoch: 4/20... Training Step: 14661... Training loss: 1.5731... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14662... Training loss: 1.7821... 0.0528 sec/batch
Epoch: 4/20... Training Step: 14663... Training loss: 1.6556... 0.0562 sec/batch
Epoch: 4/20... Training Step: 14664... Training loss: 1.6311... 0.0563 sec/batch
Epoch: 4/20... Training Step: 14665... Training loss: 1.8115... 0.0543 sec/batch
Epoch: 4/20... Training Step: 14666... Training loss: 1.6788... 0.0536 sec/batch
Epoch: 4/20... Training Step: 14667... Training loss: 1.7100... 0.0549 sec/batch
Epoch: 4/20... Training Step: 14668... Training loss: 1.6545... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14669... Training loss: 1.6273... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14670... Training loss: 1.6770... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14671... Training loss: 1.5744... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14672... Training loss: 1.6525... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14673... Training loss: 1.7329... 0.0571 sec/batch
Epoch: 4/20... Training Step: 14674... Training loss: 1.5231... 0.0637 sec/batch
Epoch: 4/20... Training Step: 14675... Training loss: 1.7663... 0.0554 sec/batch
Epoch: 4/20... Training Step: 14676... Training loss: 1.6165... 0.0580 sec/batch
Epoch: 4/20... Training Step: 14677... Training loss: 1.5480... 0.0506 sec/batch
Epoch: 4/20... Training Step: 14678... Training loss: 1.7567... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14679... Training loss: 1.8338... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14680... Training loss: 1.7452... 0.0582 sec/batch
Epoch: 4/20... Training Step: 14681... Training loss: 1.6848... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14682... Training loss: 1.7628... 0.0541 sec/batch
Epoch: 4/20... Training Step: 14683... Training loss: 1.6107... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14684... Training loss: 1.7067... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14685... Training loss: 1.7080... 0.0553 sec/batch
Epoch: 4/20... Training Step: 14686... Training loss: 1.6413... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14687... Training loss: 1.6902... 0.0582 sec/batch
Epoch: 4/20... Training Step: 14688... Training loss: 1.5833... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14689... Training loss: 1.5940... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14690... Training loss: 1.5883... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14691... Training loss: 1.5803... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14692... Training loss: 1.7945... 0.0658 sec/batch
Epoch: 4/20... Training Step: 14693... Training loss: 1.5468... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14694... Training loss: 1.5359... 0.0591 sec/batch
Epoch: 4/20... Training Step: 14695... Training loss: 1.5764... 0.0580 sec/batch
Epoch: 4/20... Training Step: 14696... Training loss: 1.7866... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14697... Training loss: 1.5607... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14698... Training loss: 1.6383... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14699... Training loss: 1.5462... 0.0580 sec/batch
Epoch: 4/20... Training Step: 14700... Training loss: 1.5502... 0.0622 sec/batch
Epoch: 4/20... Training Step: 14701... Training loss: 1.6128... 0.0657 sec/batch
Epoch: 4/20... Training Step: 14702... Training loss: 1.5879... 0.0558 sec/batch
Epoch: 4/20... Training Step: 14703... Training loss: 1.5490... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14704... Training loss: 1.5250... 0.0597 sec/batch
Epoch: 4/20... Training Step: 14705... Training loss: 1.5291... 0.0594 sec/batch
Epoch: 4/20... Training Step: 14706... Training loss: 1.7436... 0.0655 sec/batch
Epoch: 4/20... Training Step: 14707... Training loss: 1.5195... 0.0739 sec/batch
Epoch: 4/20... Training Step: 14708... Training loss: 1.7155... 0.0755 sec/batch
Epoch: 4/20... Training Step: 14709... Training loss: 1.6633... 0.0640 sec/batch
Epoch: 4/20... Training Step: 14710... Training loss: 1.5426... 0.0537 sec/batch
Epoch: 4/20... Training Step: 14711... Training loss: 1.7378... 0.0653 sec/batch
Epoch: 4/20... Training Step: 14712... Training loss: 1.7533... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14713... Training loss: 1.6639... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14714... Training loss: 1.5988... 0.0635 sec/batch
Epoch: 4/20... Training Step: 14715... Training loss: 1.6914... 0.0628 sec/batch
Epoch: 4/20... Training Step: 14716... Training loss: 1.7652... 0.0633 sec/batch
Epoch: 4/20... Training Step: 14717... Training loss: 1.7507... 0.0640 sec/batch
Epoch: 4/20... Training Step: 14718... Training loss: 1.6729... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14719... Training loss: 1.7164... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14720... Training loss: 1.7587... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14721... Training loss: 1.6682... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14722... Training loss: 1.5857... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14723... Training loss: 1.6217... 0.0638 sec/batch
Epoch: 4/20... Training Step: 14724... Training loss: 1.5726... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14725... Training loss: 1.8030... 0.0596 sec/batch
Epoch: 4/20... Training Step: 14726... Training loss: 1.8366... 0.0630 sec/batch
Epoch: 4/20... Training Step: 14727... Training loss: 1.6657... 0.0679 sec/batch
Epoch: 4/20... Training Step: 14728... Training loss: 1.8002... 0.0567 sec/batch
Epoch: 4/20... Training Step: 14729... Training loss: 1.6050... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14730... Training loss: 1.7512... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14731... Training loss: 1.7085... 0.0729 sec/batch
Epoch: 4/20... Training Step: 14732... Training loss: 1.6659... 0.0587 sec/batch
Epoch: 4/20... Training Step: 14733... Training loss: 1.7570... 0.0556 sec/batch
Epoch: 4/20... Training Step: 14734... Training loss: 1.5358... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14735... Training loss: 1.8267... 0.0777 sec/batch
Epoch: 4/20... Training Step: 14736... Training loss: 1.7937... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14737... Training loss: 1.6505... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14738... Training loss: 1.7100... 0.0587 sec/batch
Epoch: 4/20... Training Step: 14739... Training loss: 1.6734... 0.0726 sec/batch
Epoch: 4/20... Training Step: 14740... Training loss: 1.7166... 0.0595 sec/batch
Epoch: 4/20... Training Step: 14741... Training loss: 1.6919... 0.0564 sec/batch
Epoch: 4/20... Training Step: 14742... Training loss: 1.7421... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14743... Training loss: 1.6840... 0.0765 sec/batch
Epoch: 4/20... Training Step: 14744... Training loss: 1.8145... 0.0546 sec/batch
Epoch: 4/20... Training Step: 14745... Training loss: 1.7356... 0.0631 sec/batch
Epoch: 4/20... Training Step: 14746... Training loss: 1.8023... 0.0593 sec/batch
Epoch: 4/20... Training Step: 14747... Training loss: 1.8307... 0.0659 sec/batch
Epoch: 4/20... Training Step: 14748... Training loss: 1.7102... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14749... Training loss: 1.6647... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14750... Training loss: 1.6659... 0.0714 sec/batch
Epoch: 4/20... Training Step: 14751... Training loss: 1.7675... 0.0640 sec/batch
Epoch: 4/20... Training Step: 14752... Training loss: 1.7859... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14753... Training loss: 1.4958... 0.0549 sec/batch
Epoch: 4/20... Training Step: 14754... Training loss: 1.5673... 0.0560 sec/batch
Epoch: 4/20... Training Step: 14755... Training loss: 1.7150... 0.0755 sec/batch
Epoch: 4/20... Training Step: 14756... Training loss: 1.6446... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14757... Training loss: 1.6918... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14758... Training loss: 1.6458... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14759... Training loss: 1.6209... 0.0694 sec/batch
Epoch: 4/20... Training Step: 14760... Training loss: 1.7792... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14761... Training loss: 1.7162... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14762... Training loss: 1.8180... 0.0604 sec/batch
Epoch: 4/20... Training Step: 14763... Training loss: 1.8317... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14764... Training loss: 1.6141... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14765... Training loss: 1.6701... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14766... Training loss: 1.8024... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14767... Training loss: 1.6919... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14768... Training loss: 1.8168... 0.0593 sec/batch
Epoch: 4/20... Training Step: 14769... Training loss: 1.4449... 0.0548 sec/batch
Epoch: 4/20... Training Step: 14770... Training loss: 1.5903... 0.0588 sec/batch
Epoch: 4/20... Training Step: 14771... Training loss: 1.7303... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14772... Training loss: 1.5237... 0.0538 sec/batch
Epoch: 4/20... Training Step: 14773... Training loss: 1.6534... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14774... Training loss: 1.6632... 0.0535 sec/batch
Epoch: 4/20... Training Step: 14775... Training loss: 1.5553... 0.0613 sec/batch
Epoch: 4/20... Training Step: 14776... Training loss: 1.5598... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14777... Training loss: 1.5863... 0.0574 sec/batch
Epoch: 4/20... Training Step: 14778... Training loss: 1.6292... 0.0515 sec/batch
Epoch: 4/20... Training Step: 14779... Training loss: 1.7580... 0.0546 sec/batch
Epoch: 4/20... Training Step: 14780... Training loss: 1.6387... 0.0819 sec/batch
Epoch: 4/20... Training Step: 14781... Training loss: 1.5979... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14782... Training loss: 1.5317... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14783... Training loss: 1.6744... 0.0679 sec/batch
Epoch: 4/20... Training Step: 14784... Training loss: 1.6563... 0.0537 sec/batch
Epoch: 4/20... Training Step: 14785... Training loss: 1.7075... 0.0586 sec/batch
Epoch: 4/20... Training Step: 14786... Training loss: 1.8061... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14787... Training loss: 1.7280... 0.0555 sec/batch
Epoch: 4/20... Training Step: 14788... Training loss: 1.5959... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14789... Training loss: 1.8600... 0.0515 sec/batch
Epoch: 4/20... Training Step: 14790... Training loss: 1.6644... 0.0730 sec/batch
Epoch: 4/20... Training Step: 14791... Training loss: 1.8037... 0.0539 sec/batch
Epoch: 4/20... Training Step: 14792... Training loss: 1.8037... 0.0594 sec/batch
Epoch: 4/20... Training Step: 14793... Training loss: 1.7208... 0.0571 sec/batch
Epoch: 4/20... Training Step: 14794... Training loss: 1.6532... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14795... Training loss: 1.6847... 0.0604 sec/batch
Epoch: 4/20... Training Step: 14796... Training loss: 1.5784... 0.0536 sec/batch
Epoch: 4/20... Training Step: 14797... Training loss: 1.6726... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14798... Training loss: 1.7015... 0.0599 sec/batch
Epoch: 4/20... Training Step: 14799... Training loss: 1.7673... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14800... Training loss: 1.5895... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14801... Training loss: 1.7807... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14802... Training loss: 1.5623... 0.0546 sec/batch
Epoch: 4/20... Training Step: 14803... Training loss: 1.7312... 0.0678 sec/batch
Epoch: 4/20... Training Step: 14804... Training loss: 1.6644... 0.0946 sec/batch
Epoch: 4/20... Training Step: 14805... Training loss: 1.5770... 0.0689 sec/batch
Epoch: 4/20... Training Step: 14806... Training loss: 1.5835... 0.0710 sec/batch
Epoch: 4/20... Training Step: 14807... Training loss: 1.5902... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14808... Training loss: 1.7602... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14809... Training loss: 1.6369... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14810... Training loss: 1.5162... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14811... Training loss: 1.5265... 0.0825 sec/batch
Epoch: 4/20... Training Step: 14812... Training loss: 1.7035... 0.0654 sec/batch
Epoch: 4/20... Training Step: 14813... Training loss: 1.5464... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14814... Training loss: 1.6591... 0.0581 sec/batch
Epoch: 4/20... Training Step: 14815... Training loss: 1.7225... 0.0662 sec/batch
Epoch: 4/20... Training Step: 14816... Training loss: 1.6498... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14817... Training loss: 1.6156... 0.0585 sec/batch
Epoch: 4/20... Training Step: 14818... Training loss: 1.6858... 0.0540 sec/batch
Epoch: 4/20... Training Step: 14819... Training loss: 1.5176... 0.0710 sec/batch
Epoch: 4/20... Training Step: 14820... Training loss: 1.7117... 0.0533 sec/batch
Epoch: 4/20... Training Step: 14821... Training loss: 1.5334... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14822... Training loss: 1.7612... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14823... Training loss: 1.6740... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14824... Training loss: 1.5896... 0.0554 sec/batch
Epoch: 4/20... Training Step: 14825... Training loss: 1.8201... 0.0559 sec/batch
Epoch: 4/20... Training Step: 14826... Training loss: 1.6348... 0.0645 sec/batch
Epoch: 4/20... Training Step: 14827... Training loss: 1.7120... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14828... Training loss: 1.7607... 0.0539 sec/batch
Epoch: 4/20... Training Step: 14829... Training loss: 1.8135... 0.0585 sec/batch
Epoch: 4/20... Training Step: 14830... Training loss: 1.7565... 0.0680 sec/batch
Epoch: 4/20... Training Step: 14831... Training loss: 1.7660... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14832... Training loss: 1.5932... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14833... Training loss: 1.6378... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14834... Training loss: 1.8357... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14835... Training loss: 1.6522... 0.0740 sec/batch
Epoch: 4/20... Training Step: 14836... Training loss: 1.5591... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14837... Training loss: 1.6550... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14838... Training loss: 1.7461... 0.0575 sec/batch
Epoch: 4/20... Training Step: 14839... Training loss: 1.6146... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14840... Training loss: 1.6068... 0.0553 sec/batch
Epoch: 4/20... Training Step: 14841... Training loss: 1.5337... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14842... Training loss: 1.5364... 0.0625 sec/batch
Epoch: 4/20... Training Step: 14843... Training loss: 1.6087... 0.0672 sec/batch
Epoch: 4/20... Training Step: 14844... Training loss: 1.6342... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14845... Training loss: 1.6689... 0.0627 sec/batch
Epoch: 4/20... Training Step: 14846... Training loss: 1.6585... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14847... Training loss: 1.6898... 0.0685 sec/batch
Epoch: 4/20... Training Step: 14848... Training loss: 1.5812... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14849... Training loss: 1.6276... 0.0502 sec/batch
Epoch: 4/20... Training Step: 14850... Training loss: 1.7235... 0.0560 sec/batch
Epoch: 4/20... Training Step: 14851... Training loss: 1.7428... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14852... Training loss: 1.7056... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14853... Training loss: 1.6638... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14854... Training loss: 1.6107... 0.0653 sec/batch
Epoch: 4/20... Training Step: 14855... Training loss: 1.5953... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14856... Training loss: 1.7479... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14857... Training loss: 1.6145... 0.0596 sec/batch
Epoch: 4/20... Training Step: 14858... Training loss: 1.5674... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14859... Training loss: 1.5342... 0.0723 sec/batch
Epoch: 4/20... Training Step: 14860... Training loss: 1.4522... 0.0636 sec/batch
Epoch: 4/20... Training Step: 14861... Training loss: 1.5729... 0.0667 sec/batch
Epoch: 4/20... Training Step: 14862... Training loss: 1.6149... 0.0651 sec/batch
Epoch: 4/20... Training Step: 14863... Training loss: 1.5744... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14864... Training loss: 1.5479... 0.0583 sec/batch
Epoch: 4/20... Training Step: 14865... Training loss: 1.5172... 0.0565 sec/batch
Epoch: 4/20... Training Step: 14866... Training loss: 1.6388... 0.0620 sec/batch
Epoch: 4/20... Training Step: 14867... Training loss: 1.5949... 0.0689 sec/batch
Epoch: 4/20... Training Step: 14868... Training loss: 1.5717... 0.0607 sec/batch
Epoch: 4/20... Training Step: 14869... Training loss: 1.5930... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14870... Training loss: 1.6660... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14871... Training loss: 1.6623... 0.0682 sec/batch
Epoch: 4/20... Training Step: 14872... Training loss: 1.7942... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14873... Training loss: 1.6085... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14874... Training loss: 1.4964... 0.0589 sec/batch
Epoch: 4/20... Training Step: 14875... Training loss: 1.5687... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14876... Training loss: 1.6158... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14877... Training loss: 1.5851... 0.0624 sec/batch
Epoch: 4/20... Training Step: 14878... Training loss: 1.6891... 0.0605 sec/batch
Epoch: 4/20... Training Step: 14879... Training loss: 1.5567... 0.0744 sec/batch
Epoch: 4/20... Training Step: 14880... Training loss: 1.5629... 0.0568 sec/batch
Epoch: 4/20... Training Step: 14881... Training loss: 1.7193... 0.0654 sec/batch
Epoch: 4/20... Training Step: 14882... Training loss: 1.5693... 0.0647 sec/batch
Epoch: 4/20... Training Step: 14883... Training loss: 1.5494... 0.0672 sec/batch
Epoch: 4/20... Training Step: 14884... Training loss: 1.7274... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14885... Training loss: 1.7217... 0.0566 sec/batch
Epoch: 4/20... Training Step: 14886... Training loss: 1.6484... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14887... Training loss: 1.5937... 0.0691 sec/batch
Epoch: 4/20... Training Step: 14888... Training loss: 1.7670... 0.0648 sec/batch
Epoch: 4/20... Training Step: 14889... Training loss: 1.6888... 0.0563 sec/batch
Epoch: 4/20... Training Step: 14890... Training loss: 1.8948... 0.0554 sec/batch
Epoch: 4/20... Training Step: 14891... Training loss: 1.7762... 0.0699 sec/batch
Epoch: 4/20... Training Step: 14892... Training loss: 1.7413... 0.0617 sec/batch
Epoch: 4/20... Training Step: 14893... Training loss: 1.4841... 0.0652 sec/batch
Epoch: 4/20... Training Step: 14894... Training loss: 1.5185... 0.0604 sec/batch
Epoch: 4/20... Training Step: 14895... Training loss: 1.6978... 0.0632 sec/batch
Epoch: 4/20... Training Step: 14896... Training loss: 1.7621... 0.0576 sec/batch
Epoch: 4/20... Training Step: 14897... Training loss: 1.5909... 0.0601 sec/batch
Epoch: 4/20... Training Step: 14898... Training loss: 1.7777... 0.0594 sec/batch
Epoch: 4/20... Training Step: 14899... Training loss: 1.7585... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14900... Training loss: 1.6996... 0.0618 sec/batch
Epoch: 4/20... Training Step: 14901... Training loss: 1.7105... 0.0540 sec/batch
Epoch: 4/20... Training Step: 14902... Training loss: 1.6473... 0.0544 sec/batch
Epoch: 4/20... Training Step: 14903... Training loss: 1.4998... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14904... Training loss: 1.7101... 0.0579 sec/batch
Epoch: 4/20... Training Step: 14905... Training loss: 1.7417... 0.0533 sec/batch
Epoch: 4/20... Training Step: 14906... Training loss: 1.8637... 0.0578 sec/batch
Epoch: 4/20... Training Step: 14907... Training loss: 1.8498... 0.0568 sec/batch
Epoch: 4/20... Training Step: 14908... Training loss: 1.6846... 0.0538 sec/batch
Epoch: 4/20... Training Step: 14909... Training loss: 1.5926... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14910... Training loss: 1.6108... 0.0592 sec/batch
Epoch: 4/20... Training Step: 14911... Training loss: 1.6373... 0.0609 sec/batch
Epoch: 4/20... Training Step: 14912... Training loss: 1.5029... 0.0612 sec/batch
Epoch: 4/20... Training Step: 14913... Training loss: 1.6584... 0.0540 sec/batch
Epoch: 4/20... Training Step: 14914... Training loss: 1.6189... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14915... Training loss: 1.7874... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14916... Training loss: 1.6241... 0.0544 sec/batch
Epoch: 4/20... Training Step: 14917... Training loss: 1.5904... 0.0524 sec/batch
Epoch: 4/20... Training Step: 14918... Training loss: 1.7677... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14919... Training loss: 1.6858... 0.0524 sec/batch
Epoch: 4/20... Training Step: 14920... Training loss: 1.7710... 0.0545 sec/batch
Epoch: 4/20... Training Step: 14921... Training loss: 1.5586... 0.0593 sec/batch
Epoch: 4/20... Training Step: 14922... Training loss: 1.8418... 0.0596 sec/batch
Epoch: 4/20... Training Step: 14923... Training loss: 1.5821... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14924... Training loss: 1.7172... 0.0533 sec/batch
Epoch: 4/20... Training Step: 14925... Training loss: 1.7710... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14926... Training loss: 1.6613... 0.0559 sec/batch
Epoch: 4/20... Training Step: 14927... Training loss: 1.6686... 0.0551 sec/batch
Epoch: 4/20... Training Step: 14928... Training loss: 1.7191... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14929... Training loss: 1.5876... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14930... Training loss: 1.8580... 0.0568 sec/batch
Epoch: 4/20... Training Step: 14931... Training loss: 1.7001... 0.0687 sec/batch
Epoch: 4/20... Training Step: 14932... Training loss: 1.6396... 0.0558 sec/batch
Epoch: 4/20... Training Step: 14933... Training loss: 1.6415... 0.0614 sec/batch
Epoch: 4/20... Training Step: 14934... Training loss: 1.6281... 0.0537 sec/batch
Epoch: 4/20... Training Step: 14935... Training loss: 1.6352... 0.0761 sec/batch
Epoch: 4/20... Training Step: 14936... Training loss: 1.5733... 0.0521 sec/batch
Epoch: 4/20... Training Step: 14937... Training loss: 1.6617... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14938... Training loss: 1.7157... 0.0549 sec/batch
Epoch: 4/20... Training Step: 14939... Training loss: 1.7908... 0.0615 sec/batch
Epoch: 4/20... Training Step: 14940... Training loss: 1.5634... 0.0531 sec/batch
Epoch: 4/20... Training Step: 14941... Training loss: 1.7138... 0.0584 sec/batch
Epoch: 4/20... Training Step: 14942... Training loss: 1.7422... 0.0527 sec/batch
Epoch: 4/20... Training Step: 14943... Training loss: 1.6042... 0.0707 sec/batch
Epoch: 4/20... Training Step: 14944... Training loss: 1.6713... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14945... Training loss: 1.4225... 0.0551 sec/batch
Epoch: 4/20... Training Step: 14946... Training loss: 1.8474... 0.0531 sec/batch
Epoch: 4/20... Training Step: 14947... Training loss: 1.6845... 0.0745 sec/batch
Epoch: 4/20... Training Step: 14948... Training loss: 1.8476... 0.0591 sec/batch
Epoch: 4/20... Training Step: 14949... Training loss: 1.6066... 0.0598 sec/batch
Epoch: 4/20... Training Step: 14950... Training loss: 1.7827... 0.0623 sec/batch
Epoch: 4/20... Training Step: 14951... Training loss: 1.5867... 0.0596 sec/batch
Epoch: 4/20... Training Step: 14952... Training loss: 1.6972... 0.0593 sec/batch
Epoch: 4/20... Training Step: 14953... Training loss: 1.6836... 0.0621 sec/batch
Epoch: 4/20... Training Step: 14954... Training loss: 1.6828... 0.0600 sec/batch
Epoch: 4/20... Training Step: 14955... Training loss: 1.7858... 0.0643 sec/batch
Epoch: 4/20... Training Step: 14956... Training loss: 1.7354... 0.0513 sec/batch
Epoch: 4/20... Training Step: 14957... Training loss: 1.5884... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14958... Training loss: 1.8519... 0.0586 sec/batch
Epoch: 4/20... Training Step: 14959... Training loss: 1.7138... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14960... Training loss: 1.7505... 0.0597 sec/batch
Epoch: 4/20... Training Step: 14961... Training loss: 1.7427... 0.0582 sec/batch
Epoch: 4/20... Training Step: 14962... Training loss: 1.7249... 0.0634 sec/batch
Epoch: 4/20... Training Step: 14963... Training loss: 1.8683... 0.0716 sec/batch
Epoch: 4/20... Training Step: 14964... Training loss: 1.6434... 0.0639 sec/batch
Epoch: 4/20... Training Step: 14965... Training loss: 1.6340... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14966... Training loss: 1.7072... 0.0644 sec/batch
Epoch: 4/20... Training Step: 14967... Training loss: 1.7919... 0.0715 sec/batch
Epoch: 4/20... Training Step: 14968... Training loss: 1.7175... 0.0542 sec/batch
Epoch: 4/20... Training Step: 14969... Training loss: 1.7970... 0.0559 sec/batch
Epoch: 4/20... Training Step: 14970... Training loss: 1.7039... 0.0642 sec/batch
Epoch: 4/20... Training Step: 14971... Training loss: 1.8190... 0.0608 sec/batch
Epoch: 4/20... Training Step: 14972... Training loss: 1.6844... 0.0661 sec/batch
Epoch: 4/20... Training Step: 14973... Training loss: 1.7714... 0.0595 sec/batch
Epoch: 4/20... Training Step: 14974... Training loss: 1.6094... 0.0590 sec/batch
Epoch: 4/20... Training Step: 14975... Training loss: 1.5807... 0.0656 sec/batch
Epoch: 4/20... Training Step: 14976... Training loss: 1.6599... 0.0629 sec/batch
Epoch: 4/20... Training Step: 14977... Training loss: 1.7800... 0.0589 sec/batch
Epoch: 4/20... Training Step: 14978... Training loss: 1.6475... 0.0578 sec/batch
Epoch: 4/20... Training Step: 14979... Training loss: 1.6594... 0.0838 sec/batch
Epoch: 4/20... Training Step: 14980... Training loss: 1.6190... 0.0708 sec/batch
Epoch: 4/20... Training Step: 14981... Training loss: 1.5900... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14982... Training loss: 1.6354... 0.0641 sec/batch
Epoch: 4/20... Training Step: 14983... Training loss: 1.7059... 0.0655 sec/batch
Epoch: 4/20... Training Step: 14984... Training loss: 1.5050... 0.0834 sec/batch
Epoch: 4/20... Training Step: 14985... Training loss: 1.7119... 0.0603 sec/batch
Epoch: 4/20... Training Step: 14986... Training loss: 1.6783... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14987... Training loss: 1.7074... 0.0619 sec/batch
Epoch: 4/20... Training Step: 14988... Training loss: 1.6648... 0.0606 sec/batch
Epoch: 4/20... Training Step: 14989... Training loss: 1.5848... 0.0610 sec/batch
Epoch: 4/20... Training Step: 14990... Training loss: 1.6648... 0.0698 sec/batch
Epoch: 4/20... Training Step: 14991... Training loss: 1.6385... 0.0572 sec/batch
Epoch: 4/20... Training Step: 14992... Training loss: 1.5101... 0.0616 sec/batch
Epoch: 4/20... Training Step: 14993... Training loss: 1.7154... 0.0660 sec/batch
Epoch: 4/20... Training Step: 14994... Training loss: 1.5782... 0.0694 sec/batch
Epoch: 4/20... Training Step: 14995... Training loss: 1.7244... 0.0681 sec/batch
Epoch: 4/20... Training Step: 14996... Training loss: 1.6117... 0.0626 sec/batch
Epoch: 4/20... Training Step: 14997... Training loss: 1.5111... 0.0565 sec/batch
Epoch: 4/20... Training Step: 14998... Training loss: 1.6508... 0.0673 sec/batch
Epoch: 4/20... Training Step: 14999... Training loss: 1.6041... 0.0622 sec/batch
Epoch: 4/20... Training Step: 15000... Training loss: 1.6386... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15001... Training loss: 1.4709... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15002... Training loss: 1.5537... 0.0688 sec/batch
Epoch: 4/20... Training Step: 15003... Training loss: 1.7582... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15004... Training loss: 1.7734... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15005... Training loss: 1.7450... 0.0678 sec/batch
Epoch: 4/20... Training Step: 15006... Training loss: 1.6402... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15007... Training loss: 1.7176... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15008... Training loss: 1.6861... 0.0539 sec/batch
Epoch: 4/20... Training Step: 15009... Training loss: 1.6004... 0.0533 sec/batch
Epoch: 4/20... Training Step: 15010... Training loss: 1.5976... 0.0569 sec/batch
Epoch: 4/20... Training Step: 15011... Training loss: 1.7423... 0.0576 sec/batch
Epoch: 4/20... Training Step: 15012... Training loss: 1.6180... 0.0627 sec/batch
Epoch: 4/20... Training Step: 15013... Training loss: 1.4774... 0.0681 sec/batch
Epoch: 4/20... Training Step: 15014... Training loss: 1.6455... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15015... Training loss: 1.5665... 0.0709 sec/batch
Epoch: 4/20... Training Step: 15016... Training loss: 1.6392... 0.0647 sec/batch
Epoch: 4/20... Training Step: 15017... Training loss: 1.6850... 0.0692 sec/batch
Epoch: 4/20... Training Step: 15018... Training loss: 1.5509... 0.0532 sec/batch
Epoch: 4/20... Training Step: 15019... Training loss: 1.5100... 0.0625 sec/batch
Epoch: 4/20... Training Step: 15020... Training loss: 1.6065... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15021... Training loss: 1.6579... 0.0721 sec/batch
Epoch: 4/20... Training Step: 15022... Training loss: 1.5946... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15023... Training loss: 1.6745... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15024... Training loss: 1.5527... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15025... Training loss: 1.7608... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15026... Training loss: 1.5734... 0.0535 sec/batch
Epoch: 4/20... Training Step: 15027... Training loss: 1.5512... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15028... Training loss: 1.5919... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15029... Training loss: 1.6971... 0.0654 sec/batch
Epoch: 4/20... Training Step: 15030... Training loss: 1.6690... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15031... Training loss: 1.6292... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15032... Training loss: 1.5789... 0.0682 sec/batch
Epoch: 4/20... Training Step: 15033... Training loss: 1.7527... 0.0648 sec/batch
Epoch: 4/20... Training Step: 15034... Training loss: 1.6437... 0.0649 sec/batch
Epoch: 4/20... Training Step: 15035... Training loss: 1.5907... 0.0634 sec/batch
Epoch: 4/20... Training Step: 15036... Training loss: 1.6621... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15037... Training loss: 1.6670... 0.0691 sec/batch
Epoch: 4/20... Training Step: 15038... Training loss: 1.5694... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15039... Training loss: 1.4906... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15040... Training loss: 1.6404... 0.0657 sec/batch
Epoch: 4/20... Training Step: 15041... Training loss: 1.6712... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15042... Training loss: 1.6189... 0.0629 sec/batch
Epoch: 4/20... Training Step: 15043... Training loss: 1.7886... 0.0582 sec/batch
Epoch: 4/20... Training Step: 15044... Training loss: 1.7134... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15045... Training loss: 1.7276... 0.0571 sec/batch
Epoch: 4/20... Training Step: 15046... Training loss: 1.7362... 0.0544 sec/batch
Epoch: 4/20... Training Step: 15047... Training loss: 1.5749... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15048... Training loss: 1.6555... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15049... Training loss: 1.7642... 0.0521 sec/batch
Epoch: 4/20... Training Step: 15050... Training loss: 1.7535... 0.0676 sec/batch
Epoch: 4/20... Training Step: 15051... Training loss: 1.7222... 0.0574 sec/batch
Epoch: 4/20... Training Step: 15052... Training loss: 1.5113... 0.0595 sec/batch
Epoch: 4/20... Training Step: 15053... Training loss: 1.6924... 0.0631 sec/batch
Epoch: 4/20... Training Step: 15054... Training loss: 1.5961... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15055... Training loss: 1.6654... 0.0629 sec/batch
Epoch: 4/20... Training Step: 15056... Training loss: 1.6174... 0.0511 sec/batch
Epoch: 4/20... Training Step: 15057... Training loss: 1.6380... 0.0722 sec/batch
Epoch: 4/20... Training Step: 15058... Training loss: 1.7058... 0.0533 sec/batch
Epoch: 4/20... Training Step: 15059... Training loss: 1.5832... 0.0535 sec/batch
Epoch: 4/20... Training Step: 15060... Training loss: 1.5778... 0.0570 sec/batch
Epoch: 4/20... Training Step: 15061... Training loss: 1.6216... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15062... Training loss: 1.6745... 0.0519 sec/batch
Epoch: 4/20... Training Step: 15063... Training loss: 1.6262... 0.0627 sec/batch
Epoch: 4/20... Training Step: 15064... Training loss: 1.5197... 0.0493 sec/batch
Epoch: 4/20... Training Step: 15065... Training loss: 1.7231... 0.0529 sec/batch
Epoch: 4/20... Training Step: 15066... Training loss: 1.7794... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15067... Training loss: 1.6139... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15068... Training loss: 1.6416... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15069... Training loss: 1.5836... 0.0653 sec/batch
Epoch: 4/20... Training Step: 15070... Training loss: 1.5969... 0.0541 sec/batch
Epoch: 4/20... Training Step: 15071... Training loss: 1.7568... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15072... Training loss: 1.6600... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15073... Training loss: 1.7607... 0.0701 sec/batch
Epoch: 4/20... Training Step: 15074... Training loss: 1.5891... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15075... Training loss: 1.5756... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15076... Training loss: 1.6171... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15077... Training loss: 1.5979... 0.0669 sec/batch
Epoch: 4/20... Training Step: 15078... Training loss: 1.6120... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15079... Training loss: 1.6220... 0.0528 sec/batch
Epoch: 4/20... Training Step: 15080... Training loss: 1.6846... 0.0662 sec/batch
Epoch: 4/20... Training Step: 15081... Training loss: 1.7197... 0.0645 sec/batch
Epoch: 4/20... Training Step: 15082... Training loss: 1.6472... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15083... Training loss: 1.6485... 0.0620 sec/batch
Epoch: 4/20... Training Step: 15084... Training loss: 1.5241... 0.0580 sec/batch
Epoch: 4/20... Training Step: 15085... Training loss: 1.5957... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15086... Training loss: 1.6589... 0.0631 sec/batch
Epoch: 4/20... Training Step: 15087... Training loss: 1.7702... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15088... Training loss: 1.6058... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15089... Training loss: 1.7000... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15090... Training loss: 1.5387... 0.0571 sec/batch
Epoch: 4/20... Training Step: 15091... Training loss: 1.6677... 0.0584 sec/batch
Epoch: 4/20... Training Step: 15092... Training loss: 1.6057... 0.0722 sec/batch
Epoch: 4/20... Training Step: 15093... Training loss: 1.6352... 0.0707 sec/batch
Epoch: 4/20... Training Step: 15094... Training loss: 1.6440... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15095... Training loss: 1.7705... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15096... Training loss: 1.7649... 0.0570 sec/batch
Epoch: 4/20... Training Step: 15097... Training loss: 1.4544... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15098... Training loss: 1.8052... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15099... Training loss: 1.6181... 0.0620 sec/batch
Epoch: 4/20... Training Step: 15100... Training loss: 1.5897... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15101... Training loss: 1.5563... 0.0703 sec/batch
Epoch: 4/20... Training Step: 15102... Training loss: 1.5391... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15103... Training loss: 1.6570... 0.0650 sec/batch
Epoch: 4/20... Training Step: 15104... Training loss: 1.5953... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15105... Training loss: 1.6421... 0.0689 sec/batch
Epoch: 4/20... Training Step: 15106... Training loss: 1.6730... 0.0635 sec/batch
Epoch: 4/20... Training Step: 15107... Training loss: 1.6706... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15108... Training loss: 1.7334... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15109... Training loss: 1.6253... 0.0653 sec/batch
Epoch: 4/20... Training Step: 15110... Training loss: 1.7568... 0.0679 sec/batch
Epoch: 4/20... Training Step: 15111... Training loss: 1.8065... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15112... Training loss: 1.5665... 0.0528 sec/batch
Epoch: 4/20... Training Step: 15113... Training loss: 1.5886... 0.0791 sec/batch
Epoch: 4/20... Training Step: 15114... Training loss: 1.8067... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15115... Training loss: 1.6183... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15116... Training loss: 1.6922... 0.0630 sec/batch
Epoch: 4/20... Training Step: 15117... Training loss: 1.6840... 0.0752 sec/batch
Epoch: 4/20... Training Step: 15118... Training loss: 1.7720... 0.0665 sec/batch
Epoch: 4/20... Training Step: 15119... Training loss: 1.5867... 0.0599 sec/batch
Epoch: 4/20... Training Step: 15120... Training loss: 1.5616... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15121... Training loss: 1.7824... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15122... Training loss: 1.7800... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15123... Training loss: 1.7769... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15124... Training loss: 1.7106... 0.0637 sec/batch
Epoch: 4/20... Training Step: 15125... Training loss: 1.7408... 0.0544 sec/batch
Epoch: 4/20... Training Step: 15126... Training loss: 1.7435... 0.0584 sec/batch
Epoch: 4/20... Training Step: 15127... Training loss: 1.6423... 0.0631 sec/batch
Epoch: 4/20... Training Step: 15128... Training loss: 1.5433... 0.0595 sec/batch
Epoch: 4/20... Training Step: 15129... Training loss: 1.5732... 0.0710 sec/batch
Epoch: 4/20... Training Step: 15130... Training loss: 1.6841... 0.0505 sec/batch
Epoch: 4/20... Training Step: 15131... Training loss: 1.5619... 0.0580 sec/batch
Epoch: 4/20... Training Step: 15132... Training loss: 1.7102... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15133... Training loss: 1.6292... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15134... Training loss: 1.5995... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15135... Training loss: 1.6678... 0.0623 sec/batch
Epoch: 4/20... Training Step: 15136... Training loss: 1.5263... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15137... Training loss: 1.6744... 0.0648 sec/batch
Epoch: 4/20... Training Step: 15138... Training loss: 1.6337... 0.0674 sec/batch
Epoch: 4/20... Training Step: 15139... Training loss: 1.6177... 0.0579 sec/batch
Epoch: 4/20... Training Step: 15140... Training loss: 1.7296... 0.0589 sec/batch
Epoch: 4/20... Training Step: 15141... Training loss: 1.7573... 0.0707 sec/batch
Epoch: 4/20... Training Step: 15142... Training loss: 1.6205... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15143... Training loss: 1.6558... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15144... Training loss: 1.5863... 0.0604 sec/batch
Epoch: 4/20... Training Step: 15145... Training loss: 1.5899... 0.0683 sec/batch
Epoch: 4/20... Training Step: 15146... Training loss: 1.6807... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15147... Training loss: 1.5695... 0.0657 sec/batch
Epoch: 4/20... Training Step: 15148... Training loss: 1.6102... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15149... Training loss: 1.6403... 0.0718 sec/batch
Epoch: 4/20... Training Step: 15150... Training loss: 1.6430... 0.0682 sec/batch
Epoch: 4/20... Training Step: 15151... Training loss: 1.7122... 0.0616 sec/batch
Epoch: 4/20... Training Step: 15152... Training loss: 1.5462... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15153... Training loss: 1.6877... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15154... Training loss: 1.7303... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15155... Training loss: 1.7015... 0.0591 sec/batch
Epoch: 4/20... Training Step: 15156... Training loss: 1.7346... 0.0534 sec/batch
Epoch: 4/20... Training Step: 15157... Training loss: 1.6964... 0.0665 sec/batch
Epoch: 4/20... Training Step: 15158... Training loss: 1.7689... 0.0556 sec/batch
Epoch: 4/20... Training Step: 15159... Training loss: 1.7649... 0.0527 sec/batch
Epoch: 4/20... Training Step: 15160... Training loss: 1.6718... 0.0566 sec/batch
Epoch: 4/20... Training Step: 15161... Training loss: 1.8497... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15162... Training loss: 1.8074... 0.0566 sec/batch
Epoch: 4/20... Training Step: 15163... Training loss: 1.8930... 0.0575 sec/batch
Epoch: 4/20... Training Step: 15164... Training loss: 1.6458... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15165... Training loss: 1.7173... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15166... Training loss: 1.6584... 0.0521 sec/batch
Epoch: 4/20... Training Step: 15167... Training loss: 1.5660... 0.0582 sec/batch
Epoch: 4/20... Training Step: 15168... Training loss: 1.6352... 0.0627 sec/batch
Epoch: 4/20... Training Step: 15169... Training loss: 1.7928... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15170... Training loss: 1.6232... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15171... Training loss: 1.6867... 0.0599 sec/batch
Epoch: 4/20... Training Step: 15172... Training loss: 1.6743... 0.0562 sec/batch
Epoch: 4/20... Training Step: 15173... Training loss: 1.6077... 0.0668 sec/batch
Epoch: 4/20... Training Step: 15174... Training loss: 1.5456... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15175... Training loss: 1.5668... 0.0571 sec/batch
Epoch: 4/20... Training Step: 15176... Training loss: 1.6646... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15177... Training loss: 1.5757... 0.0551 sec/batch
Epoch: 4/20... Training Step: 15178... Training loss: 1.6941... 0.0623 sec/batch
Epoch: 4/20... Training Step: 15179... Training loss: 1.5369... 0.0611 sec/batch
Epoch: 4/20... Training Step: 15180... Training loss: 1.7373... 0.0533 sec/batch
Epoch: 4/20... Training Step: 15181... Training loss: 1.6708... 0.0544 sec/batch
Epoch: 4/20... Training Step: 15182... Training loss: 1.6594... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15183... Training loss: 1.6978... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15184... Training loss: 1.6142... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15185... Training loss: 1.6045... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15186... Training loss: 1.6199... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15187... Training loss: 1.7026... 0.0658 sec/batch
Epoch: 4/20... Training Step: 15188... Training loss: 1.6469... 0.0560 sec/batch
Epoch: 4/20... Training Step: 15189... Training loss: 1.6083... 0.0611 sec/batch
Epoch: 4/20... Training Step: 15190... Training loss: 1.5486... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15191... Training loss: 1.6755... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15192... Training loss: 1.5449... 0.0542 sec/batch
Epoch: 4/20... Training Step: 15193... Training loss: 1.7205... 0.0708 sec/batch
Epoch: 4/20... Training Step: 15194... Training loss: 1.6465... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15195... Training loss: 1.5900... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15196... Training loss: 1.6139... 0.0534 sec/batch
Epoch: 4/20... Training Step: 15197... Training loss: 1.6337... 0.0601 sec/batch
Epoch: 4/20... Training Step: 15198... Training loss: 1.6175... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15199... Training loss: 1.7349... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15200... Training loss: 1.8749... 0.0535 sec/batch
Epoch: 4/20... Training Step: 15201... Training loss: 1.7646... 0.0525 sec/batch
Epoch: 4/20... Training Step: 15202... Training loss: 1.6478... 0.0573 sec/batch
Epoch: 4/20... Training Step: 15203... Training loss: 1.6716... 0.0548 sec/batch
Epoch: 4/20... Training Step: 15204... Training loss: 1.6948... 0.0579 sec/batch
Epoch: 4/20... Training Step: 15205... Training loss: 1.6890... 0.0576 sec/batch
Epoch: 4/20... Training Step: 15206... Training loss: 1.7320... 0.0531 sec/batch
Epoch: 4/20... Training Step: 15207... Training loss: 1.7288... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15208... Training loss: 1.6856... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15209... Training loss: 1.7469... 0.0575 sec/batch
Epoch: 4/20... Training Step: 15210... Training loss: 1.7330... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15211... Training loss: 1.7457... 0.0611 sec/batch
Epoch: 4/20... Training Step: 15212... Training loss: 1.7733... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15213... Training loss: 1.6637... 0.0650 sec/batch
Epoch: 4/20... Training Step: 15214... Training loss: 1.7735... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15215... Training loss: 1.8607... 0.0601 sec/batch
Epoch: 4/20... Training Step: 15216... Training loss: 1.7088... 0.0533 sec/batch
Epoch: 4/20... Training Step: 15217... Training loss: 1.5849... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15218... Training loss: 1.6568... 0.0643 sec/batch
Epoch: 4/20... Training Step: 15219... Training loss: 1.6271... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15220... Training loss: 1.7427... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15221... Training loss: 1.6891... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15222... Training loss: 1.6208... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15223... Training loss: 1.6247... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15224... Training loss: 1.7869... 0.0593 sec/batch
Epoch: 4/20... Training Step: 15225... Training loss: 1.7103... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15226... Training loss: 1.7541... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15227... Training loss: 1.6913... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15228... Training loss: 1.7381... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15229... Training loss: 1.6866... 0.0670 sec/batch
Epoch: 4/20... Training Step: 15230... Training loss: 1.6629... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15231... Training loss: 1.6367... 0.0572 sec/batch
Epoch: 4/20... Training Step: 15232... Training loss: 1.6143... 0.0537 sec/batch
Epoch: 4/20... Training Step: 15233... Training loss: 1.7035... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15234... Training loss: 1.5992... 0.0591 sec/batch
Epoch: 4/20... Training Step: 15235... Training loss: 1.7863... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15236... Training loss: 1.8654... 0.0559 sec/batch
Epoch: 4/20... Training Step: 15237... Training loss: 1.7149... 0.0580 sec/batch
Epoch: 4/20... Training Step: 15238... Training loss: 1.6610... 0.0526 sec/batch
Epoch: 4/20... Training Step: 15239... Training loss: 1.8317... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15240... Training loss: 1.7828... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15241... Training loss: 1.5002... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15242... Training loss: 1.8105... 0.0631 sec/batch
Epoch: 4/20... Training Step: 15243... Training loss: 1.6685... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15244... Training loss: 1.6122... 0.0580 sec/batch
Epoch: 4/20... Training Step: 15245... Training loss: 1.7632... 0.0661 sec/batch
Epoch: 4/20... Training Step: 15246... Training loss: 1.8333... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15247... Training loss: 1.7293... 0.0584 sec/batch
Epoch: 4/20... Training Step: 15248... Training loss: 1.7251... 0.0528 sec/batch
Epoch: 4/20... Training Step: 15249... Training loss: 1.5925... 0.0757 sec/batch
Epoch: 4/20... Training Step: 15250... Training loss: 1.6164... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15251... Training loss: 1.7463... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15252... Training loss: 1.6216... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15253... Training loss: 1.6866... 0.0582 sec/batch
Epoch: 4/20... Training Step: 15254... Training loss: 1.6630... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15255... Training loss: 1.6876... 0.0558 sec/batch
Epoch: 4/20... Training Step: 15256... Training loss: 1.7751... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15257... Training loss: 1.8122... 0.0660 sec/batch
Epoch: 4/20... Training Step: 15258... Training loss: 1.6865... 0.0551 sec/batch
Epoch: 4/20... Training Step: 15259... Training loss: 1.8212... 0.0669 sec/batch
Epoch: 4/20... Training Step: 15260... Training loss: 1.5632... 0.0650 sec/batch
Epoch: 4/20... Training Step: 15261... Training loss: 1.7457... 0.0658 sec/batch
Epoch: 4/20... Training Step: 15262... Training loss: 1.5074... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15263... Training loss: 1.5669... 0.0721 sec/batch
Epoch: 4/20... Training Step: 15264... Training loss: 1.5676... 0.0712 sec/batch
Epoch: 4/20... Training Step: 15265... Training loss: 1.6380... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15266... Training loss: 1.6512... 0.0647 sec/batch
Epoch: 4/20... Training Step: 15267... Training loss: 1.6473... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15268... Training loss: 1.7311... 0.0678 sec/batch
Epoch: 4/20... Training Step: 15269... Training loss: 1.7698... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15270... Training loss: 1.6077... 0.0523 sec/batch
Epoch: 4/20... Training Step: 15271... Training loss: 1.6341... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15272... Training loss: 1.5368... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15273... Training loss: 1.6179... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15274... Training loss: 1.5682... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15275... Training loss: 1.6944... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15276... Training loss: 1.6086... 0.0697 sec/batch
Epoch: 4/20... Training Step: 15277... Training loss: 1.6729... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15278... Training loss: 1.7848... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15279... Training loss: 1.6669... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15280... Training loss: 1.5798... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15281... Training loss: 1.7884... 0.0630 sec/batch
Epoch: 4/20... Training Step: 15282... Training loss: 1.7813... 0.0574 sec/batch
Epoch: 4/20... Training Step: 15283... Training loss: 1.6280... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15284... Training loss: 1.6386... 0.0668 sec/batch
Epoch: 4/20... Training Step: 15285... Training loss: 1.6443... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15286... Training loss: 1.6823... 0.0566 sec/batch
Epoch: 4/20... Training Step: 15287... Training loss: 1.7408... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15288... Training loss: 1.5866... 0.0673 sec/batch
Epoch: 4/20... Training Step: 15289... Training loss: 1.5032... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15290... Training loss: 1.5442... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15291... Training loss: 1.5760... 0.0593 sec/batch
Epoch: 4/20... Training Step: 15292... Training loss: 1.5684... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15293... Training loss: 1.5434... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15294... Training loss: 1.6363... 0.0589 sec/batch
Epoch: 4/20... Training Step: 15295... Training loss: 1.6004... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15296... Training loss: 1.5369... 0.0532 sec/batch
Epoch: 4/20... Training Step: 15297... Training loss: 1.7460... 0.0554 sec/batch
Epoch: 4/20... Training Step: 15298... Training loss: 1.6909... 0.0548 sec/batch
Epoch: 4/20... Training Step: 15299... Training loss: 1.6195... 0.0530 sec/batch
Epoch: 4/20... Training Step: 15300... Training loss: 1.6277... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15301... Training loss: 1.6360... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15302... Training loss: 1.6105... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15303... Training loss: 1.6822... 0.0553 sec/batch
Epoch: 4/20... Training Step: 15304... Training loss: 1.7198... 0.0589 sec/batch
Epoch: 4/20... Training Step: 15305... Training loss: 1.5186... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15306... Training loss: 1.7904... 0.0562 sec/batch
Epoch: 4/20... Training Step: 15307... Training loss: 1.7371... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15308... Training loss: 1.6522... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15309... Training loss: 1.7169... 0.0604 sec/batch
Epoch: 4/20... Training Step: 15310... Training loss: 1.6216... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15311... Training loss: 1.6196... 0.0551 sec/batch
Epoch: 4/20... Training Step: 15312... Training loss: 1.7339... 0.0660 sec/batch
Epoch: 4/20... Training Step: 15313... Training loss: 1.6045... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15314... Training loss: 1.5903... 0.0562 sec/batch
Epoch: 4/20... Training Step: 15315... Training loss: 1.6996... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15316... Training loss: 1.6196... 0.0661 sec/batch
Epoch: 4/20... Training Step: 15317... Training loss: 1.4355... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15318... Training loss: 1.6249... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15319... Training loss: 1.6247... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15320... Training loss: 1.7061... 0.0685 sec/batch
Epoch: 4/20... Training Step: 15321... Training loss: 1.6239... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15322... Training loss: 1.5553... 0.0552 sec/batch
Epoch: 4/20... Training Step: 15323... Training loss: 1.5365... 0.0553 sec/batch
Epoch: 4/20... Training Step: 15324... Training loss: 1.5316... 0.0726 sec/batch
Epoch: 4/20... Training Step: 15325... Training loss: 1.6819... 0.0561 sec/batch
Epoch: 4/20... Training Step: 15326... Training loss: 1.5920... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15327... Training loss: 1.7513... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15328... Training loss: 1.7494... 0.0717 sec/batch
Epoch: 4/20... Training Step: 15329... Training loss: 1.6757... 0.0591 sec/batch
Epoch: 4/20... Training Step: 15330... Training loss: 1.8581... 0.0549 sec/batch
Epoch: 4/20... Training Step: 15331... Training loss: 1.7082... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15332... Training loss: 1.6703... 0.0674 sec/batch
Epoch: 4/20... Training Step: 15333... Training loss: 1.6777... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15334... Training loss: 1.7005... 0.0516 sec/batch
Epoch: 4/20... Training Step: 15335... Training loss: 1.7599... 0.0550 sec/batch
Epoch: 4/20... Training Step: 15336... Training loss: 1.5957... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15337... Training loss: 1.5581... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15338... Training loss: 1.8288... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15339... Training loss: 1.6877... 0.0611 sec/batch
Epoch: 4/20... Training Step: 15340... Training loss: 1.5337... 0.0538 sec/batch
Epoch: 4/20... Training Step: 15341... Training loss: 1.6511... 0.0620 sec/batch
Epoch: 4/20... Training Step: 15342... Training loss: 1.6919... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15343... Training loss: 1.6692... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15344... Training loss: 1.7295... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15345... Training loss: 1.6697... 0.0553 sec/batch
Epoch: 4/20... Training Step: 15346... Training loss: 1.6753... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15347... Training loss: 1.6343... 0.0523 sec/batch
Epoch: 4/20... Training Step: 15348... Training loss: 1.6308... 0.0622 sec/batch
Epoch: 4/20... Training Step: 15349... Training loss: 1.7728... 0.0653 sec/batch
Epoch: 4/20... Training Step: 15350... Training loss: 1.7470... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15351... Training loss: 1.6233... 0.0643 sec/batch
Epoch: 4/20... Training Step: 15352... Training loss: 1.6027... 0.0674 sec/batch
Epoch: 4/20... Training Step: 15353... Training loss: 1.7043... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15354... Training loss: 1.6850... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15355... Training loss: 1.6695... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15356... Training loss: 1.6807... 0.0709 sec/batch
Epoch: 4/20... Training Step: 15357... Training loss: 1.6378... 0.0527 sec/batch
Epoch: 4/20... Training Step: 15358... Training loss: 1.5902... 0.0547 sec/batch
Epoch: 4/20... Training Step: 15359... Training loss: 1.6482... 0.0595 sec/batch
Epoch: 4/20... Training Step: 15360... Training loss: 1.6409... 0.0569 sec/batch
Epoch: 4/20... Training Step: 15361... Training loss: 1.7297... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15362... Training loss: 1.7125... 0.0579 sec/batch
Epoch: 4/20... Training Step: 15363... Training loss: 1.6065... 0.0643 sec/batch
Epoch: 4/20... Training Step: 15364... Training loss: 1.5141... 0.0664 sec/batch
Epoch: 4/20... Training Step: 15365... Training loss: 1.5924... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15366... Training loss: 1.7351... 0.0551 sec/batch
Epoch: 4/20... Training Step: 15367... Training loss: 1.6067... 0.0526 sec/batch
Epoch: 4/20... Training Step: 15368... Training loss: 1.5810... 0.0787 sec/batch
Epoch: 4/20... Training Step: 15369... Training loss: 1.6394... 0.0627 sec/batch
Epoch: 4/20... Training Step: 15370... Training loss: 1.6738... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15371... Training loss: 1.6202... 0.0517 sec/batch
Epoch: 4/20... Training Step: 15372... Training loss: 1.6051... 0.0743 sec/batch
Epoch: 4/20... Training Step: 15373... Training loss: 1.8184... 0.0750 sec/batch
Epoch: 4/20... Training Step: 15374... Training loss: 1.6167... 0.0532 sec/batch
Epoch: 4/20... Training Step: 15375... Training loss: 1.6541... 0.0630 sec/batch
Epoch: 4/20... Training Step: 15376... Training loss: 1.4753... 0.0703 sec/batch
Epoch: 4/20... Training Step: 15377... Training loss: 1.5257... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15378... Training loss: 1.5644... 0.0658 sec/batch
Epoch: 4/20... Training Step: 15379... Training loss: 1.7492... 0.0656 sec/batch
Epoch: 4/20... Training Step: 15380... Training loss: 1.6813... 0.0708 sec/batch
Epoch: 4/20... Training Step: 15381... Training loss: 1.6338... 0.0592 sec/batch
Epoch: 4/20... Training Step: 15382... Training loss: 1.6412... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15383... Training loss: 1.7352... 0.0591 sec/batch
Epoch: 4/20... Training Step: 15384... Training loss: 1.6647... 0.0716 sec/batch
Epoch: 4/20... Training Step: 15385... Training loss: 1.6699... 0.0589 sec/batch
Epoch: 4/20... Training Step: 15386... Training loss: 1.4993... 0.0567 sec/batch
Epoch: 4/20... Training Step: 15387... Training loss: 1.7495... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15388... Training loss: 1.7099... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15389... Training loss: 1.7830... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15390... Training loss: 1.4839... 0.0601 sec/batch
Epoch: 4/20... Training Step: 15391... Training loss: 1.6649... 0.0570 sec/batch
Epoch: 4/20... Training Step: 15392... Training loss: 1.6579... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15393... Training loss: 1.7262... 0.0683 sec/batch
Epoch: 4/20... Training Step: 15394... Training loss: 1.7739... 0.0670 sec/batch
Epoch: 4/20... Training Step: 15395... Training loss: 1.6345... 0.0649 sec/batch
Epoch: 4/20... Training Step: 15396... Training loss: 1.8655... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15397... Training loss: 1.7126... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15398... Training loss: 1.7978... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15399... Training loss: 1.6782... 0.0714 sec/batch
Epoch: 4/20... Training Step: 15400... Training loss: 1.8351... 0.0676 sec/batch
Epoch: 4/20... Training Step: 15401... Training loss: 1.8339... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15402... Training loss: 1.6140... 0.0649 sec/batch
Epoch: 4/20... Training Step: 15403... Training loss: 1.5405... 0.0641 sec/batch
Epoch: 4/20... Training Step: 15404... Training loss: 1.7431... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15405... Training loss: 1.7310... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15406... Training loss: 1.5967... 0.0575 sec/batch
Epoch: 4/20... Training Step: 15407... Training loss: 1.6736... 0.0546 sec/batch
Epoch: 4/20... Training Step: 15408... Training loss: 1.6741... 0.0663 sec/batch
Epoch: 4/20... Training Step: 15409... Training loss: 1.6442... 0.0648 sec/batch
Epoch: 4/20... Training Step: 15410... Training loss: 1.6245... 0.0571 sec/batch
Epoch: 4/20... Training Step: 15411... Training loss: 1.5831... 0.0631 sec/batch
Epoch: 4/20... Training Step: 15412... Training loss: 1.6305... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15413... Training loss: 1.5888... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15414... Training loss: 1.6907... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15415... Training loss: 1.6740... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15416... Training loss: 1.6543... 0.0725 sec/batch
Epoch: 4/20... Training Step: 15417... Training loss: 1.6364... 0.0528 sec/batch
Epoch: 4/20... Training Step: 15418... Training loss: 1.5962... 0.0645 sec/batch
Epoch: 4/20... Training Step: 15419... Training loss: 1.6462... 0.0572 sec/batch
Epoch: 4/20... Training Step: 15420... Training loss: 1.6844... 0.0704 sec/batch
Epoch: 4/20... Training Step: 15421... Training loss: 1.7072... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15422... Training loss: 1.6040... 0.0574 sec/batch
Epoch: 4/20... Training Step: 15423... Training loss: 1.6638... 0.0541 sec/batch
Epoch: 4/20... Training Step: 15424... Training loss: 1.5094... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15425... Training loss: 1.5138... 0.0637 sec/batch
Epoch: 4/20... Training Step: 15426... Training loss: 1.6076... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15427... Training loss: 1.5970... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15428... Training loss: 1.6570... 0.0648 sec/batch
Epoch: 4/20... Training Step: 15429... Training loss: 1.7602... 0.0570 sec/batch
Epoch: 4/20... Training Step: 15430... Training loss: 1.5791... 0.0495 sec/batch
Epoch: 4/20... Training Step: 15431... Training loss: 1.7810... 0.0665 sec/batch
Epoch: 4/20... Training Step: 15432... Training loss: 1.7795... 0.0515 sec/batch
Epoch: 4/20... Training Step: 15433... Training loss: 1.7029... 0.0535 sec/batch
Epoch: 4/20... Training Step: 15434... Training loss: 1.6081... 0.0582 sec/batch
Epoch: 4/20... Training Step: 15435... Training loss: 1.4351... 0.0553 sec/batch
Epoch: 4/20... Training Step: 15436... Training loss: 1.6997... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15437... Training loss: 1.6182... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15438... Training loss: 1.6637... 0.0546 sec/batch
Epoch: 4/20... Training Step: 15439... Training loss: 1.5199... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15440... Training loss: 1.5322... 0.0665 sec/batch
Epoch: 4/20... Training Step: 15441... Training loss: 1.5049... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15442... Training loss: 1.6608... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15443... Training loss: 1.6919... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15444... Training loss: 1.6721... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15445... Training loss: 1.7238... 0.0561 sec/batch
Epoch: 4/20... Training Step: 15446... Training loss: 1.5610... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15447... Training loss: 1.6094... 0.0518 sec/batch
Epoch: 4/20... Training Step: 15448... Training loss: 1.7337... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15449... Training loss: 1.7131... 0.0539 sec/batch
Epoch: 4/20... Training Step: 15450... Training loss: 1.6342... 0.0576 sec/batch
Epoch: 4/20... Training Step: 15451... Training loss: 1.7325... 0.0569 sec/batch
Epoch: 4/20... Training Step: 15452... Training loss: 1.6167... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15453... Training loss: 1.5060... 0.0572 sec/batch
Epoch: 4/20... Training Step: 15454... Training loss: 1.5867... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15455... Training loss: 1.6991... 0.0548 sec/batch
Epoch: 4/20... Training Step: 15456... Training loss: 1.5536... 0.0721 sec/batch
Epoch: 4/20... Training Step: 15457... Training loss: 1.6499... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15458... Training loss: 1.5432... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15459... Training loss: 1.6018... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15460... Training loss: 1.7179... 0.0679 sec/batch
Epoch: 4/20... Training Step: 15461... Training loss: 1.5672... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15462... Training loss: 1.6528... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15463... Training loss: 1.7814... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15464... Training loss: 1.7385... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15465... Training loss: 1.5499... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15466... Training loss: 1.7482... 0.0510 sec/batch
Epoch: 4/20... Training Step: 15467... Training loss: 1.5801... 0.0541 sec/batch
Epoch: 4/20... Training Step: 15468... Training loss: 1.4841... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15469... Training loss: 1.7213... 0.0678 sec/batch
Epoch: 4/20... Training Step: 15470... Training loss: 1.6742... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15471... Training loss: 1.5888... 0.0547 sec/batch
Epoch: 4/20... Training Step: 15472... Training loss: 1.6086... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15473... Training loss: 1.6839... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15474... Training loss: 1.6330... 0.0604 sec/batch
Epoch: 4/20... Training Step: 15475... Training loss: 1.6217... 0.0568 sec/batch
Epoch: 4/20... Training Step: 15476... Training loss: 1.6261... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15477... Training loss: 1.6809... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15478... Training loss: 1.6010... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15479... Training loss: 1.6742... 0.0681 sec/batch
Epoch: 4/20... Training Step: 15480... Training loss: 1.4527... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15481... Training loss: 1.6519... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15482... Training loss: 1.7899... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15483... Training loss: 1.6194... 0.0666 sec/batch
Epoch: 4/20... Training Step: 15484... Training loss: 1.5946... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15485... Training loss: 1.4762... 0.0637 sec/batch
Epoch: 4/20... Training Step: 15486... Training loss: 1.5163... 0.0667 sec/batch
Epoch: 4/20... Training Step: 15487... Training loss: 1.6247... 0.0574 sec/batch
Epoch: 4/20... Training Step: 15488... Training loss: 1.5274... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15489... Training loss: 1.5593... 0.0649 sec/batch
Epoch: 4/20... Training Step: 15490... Training loss: 1.5401... 0.0639 sec/batch
Epoch: 4/20... Training Step: 15491... Training loss: 1.5776... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15492... Training loss: 1.7336... 0.0602 sec/batch
Epoch: 4/20... Training Step: 15493... Training loss: 1.5830... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15494... Training loss: 1.6848... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15495... Training loss: 1.6594... 0.0599 sec/batch
Epoch: 4/20... Training Step: 15496... Training loss: 1.7380... 0.0584 sec/batch
Epoch: 4/20... Training Step: 15497... Training loss: 1.8011... 0.0595 sec/batch
Epoch: 4/20... Training Step: 15498... Training loss: 1.5748... 0.0558 sec/batch
Epoch: 4/20... Training Step: 15499... Training loss: 1.6140... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15500... Training loss: 1.5583... 0.0575 sec/batch
Epoch: 4/20... Training Step: 15501... Training loss: 1.5958... 0.0673 sec/batch
Epoch: 4/20... Training Step: 15502... Training loss: 1.5735... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15503... Training loss: 1.6402... 0.0625 sec/batch
Epoch: 4/20... Training Step: 15504... Training loss: 2.1837... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15505... Training loss: 1.8119... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15506... Training loss: 1.6186... 0.0635 sec/batch
Epoch: 4/20... Training Step: 15507... Training loss: 1.8263... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15508... Training loss: 1.7850... 0.0673 sec/batch
Epoch: 4/20... Training Step: 15509... Training loss: 1.7207... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15510... Training loss: 1.6941... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15511... Training loss: 1.7683... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15512... Training loss: 1.6733... 0.0725 sec/batch
Epoch: 4/20... Training Step: 15513... Training loss: 1.7709... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15514... Training loss: 1.7293... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15515... Training loss: 1.6887... 0.0601 sec/batch
Epoch: 4/20... Training Step: 15516... Training loss: 1.6330... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15517... Training loss: 1.6723... 0.0658 sec/batch
Epoch: 4/20... Training Step: 15518... Training loss: 1.6718... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15519... Training loss: 1.8041... 0.0622 sec/batch
Epoch: 4/20... Training Step: 15520... Training loss: 1.7945... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15521... Training loss: 1.7065... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15522... Training loss: 1.7180... 0.0630 sec/batch
Epoch: 4/20... Training Step: 15523... Training loss: 1.6952... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15524... Training loss: 1.6937... 0.0602 sec/batch
Epoch: 4/20... Training Step: 15525... Training loss: 1.6591... 0.0641 sec/batch
Epoch: 4/20... Training Step: 15526... Training loss: 1.6576... 0.0602 sec/batch
Epoch: 4/20... Training Step: 15527... Training loss: 1.6804... 0.0664 sec/batch
Epoch: 4/20... Training Step: 15528... Training loss: 1.7070... 0.0695 sec/batch
Epoch: 4/20... Training Step: 15529... Training loss: 1.6887... 0.0645 sec/batch
Epoch: 4/20... Training Step: 15530... Training loss: 1.6245... 0.0568 sec/batch
Epoch: 4/20... Training Step: 15531... Training loss: 1.6177... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15532... Training loss: 1.7042... 0.0613 sec/batch
Epoch: 4/20... Training Step: 15533... Training loss: 1.6244... 0.0667 sec/batch
Epoch: 4/20... Training Step: 15534... Training loss: 1.8684... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15535... Training loss: 1.6220... 0.0613 sec/batch
Epoch: 4/20... Training Step: 15536... Training loss: 1.6739... 0.0705 sec/batch
Epoch: 4/20... Training Step: 15537... Training loss: 1.8947... 0.0710 sec/batch
Epoch: 4/20... Training Step: 15538... Training loss: 1.6367... 0.0684 sec/batch
Epoch: 4/20... Training Step: 15539... Training loss: 1.7280... 0.0629 sec/batch
Epoch: 4/20... Training Step: 15540... Training loss: 1.6316... 0.0545 sec/batch
Epoch: 4/20... Training Step: 15541... Training loss: 1.7329... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15542... Training loss: 1.7668... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15543... Training loss: 1.7607... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15544... Training loss: 1.7669... 0.0592 sec/batch
Epoch: 4/20... Training Step: 15545... Training loss: 1.6929... 0.0601 sec/batch
Epoch: 4/20... Training Step: 15546... Training loss: 1.7818... 0.0542 sec/batch
Epoch: 4/20... Training Step: 15547... Training loss: 1.5860... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15548... Training loss: 1.7078... 0.0541 sec/batch
Epoch: 4/20... Training Step: 15549... Training loss: 1.7065... 0.0530 sec/batch
Epoch: 4/20... Training Step: 15550... Training loss: 1.6066... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15551... Training loss: 1.7988... 0.0616 sec/batch
Epoch: 4/20... Training Step: 15552... Training loss: 1.7910... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15553... Training loss: 1.8893... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15554... Training loss: 1.7914... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15555... Training loss: 1.7063... 0.0691 sec/batch
Epoch: 4/20... Training Step: 15556... Training loss: 1.5610... 0.0599 sec/batch
Epoch: 4/20... Training Step: 15557... Training loss: 1.7162... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15558... Training loss: 1.7743... 0.0537 sec/batch
Epoch: 4/20... Training Step: 15559... Training loss: 1.6252... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15560... Training loss: 1.7657... 0.0560 sec/batch
Epoch: 4/20... Training Step: 15561... Training loss: 1.7583... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15562... Training loss: 1.6791... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15563... Training loss: 1.6907... 0.0595 sec/batch
Epoch: 4/20... Training Step: 15564... Training loss: 1.6592... 0.0545 sec/batch
Epoch: 4/20... Training Step: 15565... Training loss: 1.8852... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15566... Training loss: 1.9393... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15567... Training loss: 1.7222... 0.0635 sec/batch
Epoch: 4/20... Training Step: 15568... Training loss: 1.8018... 0.0677 sec/batch
Epoch: 4/20... Training Step: 15569... Training loss: 1.6945... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15570... Training loss: 1.7515... 0.0570 sec/batch
Epoch: 4/20... Training Step: 15571... Training loss: 1.7592... 0.0661 sec/batch
Epoch: 4/20... Training Step: 15572... Training loss: 1.6629... 0.0510 sec/batch
Epoch: 4/20... Training Step: 15573... Training loss: 1.8533... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15574... Training loss: 1.8722... 0.0616 sec/batch
Epoch: 4/20... Training Step: 15575... Training loss: 1.6725... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15576... Training loss: 1.7458... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15577... Training loss: 1.6084... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15578... Training loss: 1.7990... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15579... Training loss: 1.7806... 0.0549 sec/batch
Epoch: 4/20... Training Step: 15580... Training loss: 1.7381... 0.0602 sec/batch
Epoch: 4/20... Training Step: 15581... Training loss: 1.6376... 0.0566 sec/batch
Epoch: 4/20... Training Step: 15582... Training loss: 1.7505... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15583... Training loss: 1.7842... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15584... Training loss: 1.6976... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15585... Training loss: 1.5919... 0.0515 sec/batch
Epoch: 4/20... Training Step: 15586... Training loss: 1.9004... 0.0542 sec/batch
Epoch: 4/20... Training Step: 15587... Training loss: 1.7621... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15588... Training loss: 1.6810... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15589... Training loss: 1.7519... 0.0590 sec/batch
Epoch: 4/20... Training Step: 15590... Training loss: 1.8348... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15591... Training loss: 1.8018... 0.0639 sec/batch
Epoch: 4/20... Training Step: 15592... Training loss: 1.7263... 0.0527 sec/batch
Epoch: 4/20... Training Step: 15593... Training loss: 1.6476... 0.0547 sec/batch
Epoch: 4/20... Training Step: 15594... Training loss: 1.6143... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15595... Training loss: 1.6327... 0.0599 sec/batch
Epoch: 4/20... Training Step: 15596... Training loss: 1.7669... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15597... Training loss: 1.7282... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15598... Training loss: 1.6344... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15599... Training loss: 1.6390... 0.0637 sec/batch
Epoch: 4/20... Training Step: 15600... Training loss: 1.5916... 0.0583 sec/batch
Epoch: 4/20... Training Step: 15601... Training loss: 1.7170... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15602... Training loss: 1.7038... 0.0668 sec/batch
Epoch: 4/20... Training Step: 15603... Training loss: 1.8369... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15604... Training loss: 1.7278... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15605... Training loss: 1.6846... 0.0542 sec/batch
Epoch: 4/20... Training Step: 15606... Training loss: 1.7994... 0.0732 sec/batch
Epoch: 4/20... Training Step: 15607... Training loss: 1.7493... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15608... Training loss: 1.7839... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15609... Training loss: 1.7097... 0.0623 sec/batch
Epoch: 4/20... Training Step: 15610... Training loss: 1.5744... 0.0690 sec/batch
Epoch: 4/20... Training Step: 15611... Training loss: 1.8246... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15612... Training loss: 1.7004... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15613... Training loss: 1.5725... 0.0622 sec/batch
Epoch: 4/20... Training Step: 15614... Training loss: 1.8188... 0.0678 sec/batch
Epoch: 4/20... Training Step: 15615... Training loss: 1.6581... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15616... Training loss: 1.5756... 0.0668 sec/batch
Epoch: 4/20... Training Step: 15617... Training loss: 1.6365... 0.0676 sec/batch
Epoch: 4/20... Training Step: 15618... Training loss: 1.7573... 0.0662 sec/batch
Epoch: 4/20... Training Step: 15619... Training loss: 1.6888... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15620... Training loss: 1.6587... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15621... Training loss: 1.6925... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15622... Training loss: 1.8216... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15623... Training loss: 1.6678... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15624... Training loss: 1.7572... 0.0603 sec/batch
Epoch: 4/20... Training Step: 15625... Training loss: 1.8927... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15626... Training loss: 1.6269... 0.0634 sec/batch
Epoch: 4/20... Training Step: 15627... Training loss: 1.7296... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15628... Training loss: 1.6701... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15629... Training loss: 1.7008... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15630... Training loss: 1.6163... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15631... Training loss: 1.6231... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15632... Training loss: 1.5517... 0.0660 sec/batch
Epoch: 4/20... Training Step: 15633... Training loss: 1.8195... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15634... Training loss: 1.7405... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15635... Training loss: 1.6677... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15636... Training loss: 1.8099... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15637... Training loss: 1.8934... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15638... Training loss: 1.8335... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15639... Training loss: 1.9312... 0.0581 sec/batch
Epoch: 4/20... Training Step: 15640... Training loss: 1.8011... 0.0683 sec/batch
Epoch: 4/20... Training Step: 15641... Training loss: 1.6802... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15642... Training loss: 1.7969... 0.0675 sec/batch
Epoch: 4/20... Training Step: 15643... Training loss: 1.6061... 0.0648 sec/batch
Epoch: 4/20... Training Step: 15644... Training loss: 1.5963... 0.0540 sec/batch
Epoch: 4/20... Training Step: 15645... Training loss: 1.7489... 0.0647 sec/batch
Epoch: 4/20... Training Step: 15646... Training loss: 1.7331... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15647... Training loss: 1.6710... 0.0598 sec/batch
Epoch: 4/20... Training Step: 15648... Training loss: 1.8943... 0.0579 sec/batch
Epoch: 4/20... Training Step: 15649... Training loss: 1.8638... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15650... Training loss: 1.7201... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15651... Training loss: 1.7078... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15652... Training loss: 1.8746... 0.0615 sec/batch
Epoch: 4/20... Training Step: 15653... Training loss: 1.7650... 0.0577 sec/batch
Epoch: 4/20... Training Step: 15654... Training loss: 1.8670... 0.0592 sec/batch
Epoch: 4/20... Training Step: 15655... Training loss: 1.5912... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15656... Training loss: 1.8601... 0.0593 sec/batch
Epoch: 4/20... Training Step: 15657... Training loss: 1.6811... 0.0650 sec/batch
Epoch: 4/20... Training Step: 15658... Training loss: 1.9286... 0.0794 sec/batch
Epoch: 4/20... Training Step: 15659... Training loss: 1.6081... 0.0591 sec/batch
Epoch: 4/20... Training Step: 15660... Training loss: 1.8057... 0.0596 sec/batch
Epoch: 4/20... Training Step: 15661... Training loss: 1.5952... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15662... Training loss: 1.9152... 0.0715 sec/batch
Epoch: 4/20... Training Step: 15663... Training loss: 1.7298... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15664... Training loss: 1.7055... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15665... Training loss: 1.7515... 0.0653 sec/batch
Epoch: 4/20... Training Step: 15666... Training loss: 1.8042... 0.0681 sec/batch
Epoch: 4/20... Training Step: 15667... Training loss: 1.7200... 0.0662 sec/batch
Epoch: 4/20... Training Step: 15668... Training loss: 1.9293... 0.0555 sec/batch
Epoch: 4/20... Training Step: 15669... Training loss: 1.7253... 0.0906 sec/batch
Epoch: 4/20... Training Step: 15670... Training loss: 1.6018... 0.0930 sec/batch
Epoch: 4/20... Training Step: 15671... Training loss: 1.6639... 0.0945 sec/batch
Epoch: 4/20... Training Step: 15672... Training loss: 1.6523... 0.0883 sec/batch
Epoch: 4/20... Training Step: 15673... Training loss: 1.6177... 0.0814 sec/batch
Epoch: 4/20... Training Step: 15674... Training loss: 1.6927... 0.0829 sec/batch
Epoch: 4/20... Training Step: 15675... Training loss: 1.6757... 0.0983 sec/batch
Epoch: 4/20... Training Step: 15676... Training loss: 1.7231... 0.0877 sec/batch
Epoch: 4/20... Training Step: 15677... Training loss: 1.6304... 0.0870 sec/batch
Epoch: 4/20... Training Step: 15678... Training loss: 1.7479... 0.0936 sec/batch
Epoch: 4/20... Training Step: 15679... Training loss: 1.8547... 0.0865 sec/batch
Epoch: 4/20... Training Step: 15680... Training loss: 1.6171... 0.0872 sec/batch
Epoch: 4/20... Training Step: 15681... Training loss: 1.6096... 0.0943 sec/batch
Epoch: 4/20... Training Step: 15682... Training loss: 1.5451... 0.0830 sec/batch
Epoch: 4/20... Training Step: 15683... Training loss: 1.7734... 0.0917 sec/batch
Epoch: 4/20... Training Step: 15684... Training loss: 1.7305... 0.0885 sec/batch
Epoch: 4/20... Training Step: 15685... Training loss: 1.7867... 0.0797 sec/batch
Epoch: 4/20... Training Step: 15686... Training loss: 1.7666... 0.0841 sec/batch
Epoch: 4/20... Training Step: 15687... Training loss: 1.6841... 0.1025 sec/batch
Epoch: 4/20... Training Step: 15688... Training loss: 1.7010... 0.0957 sec/batch
Epoch: 4/20... Training Step: 15689... Training loss: 1.6727... 0.0817 sec/batch
Epoch: 4/20... Training Step: 15690... Training loss: 1.8827... 0.1066 sec/batch
Epoch: 4/20... Training Step: 15691... Training loss: 1.6510... 0.0854 sec/batch
Epoch: 4/20... Training Step: 15692... Training loss: 1.5574... 0.0866 sec/batch
Epoch: 4/20... Training Step: 15693... Training loss: 1.6735... 0.0909 sec/batch
Epoch: 4/20... Training Step: 15694... Training loss: 1.6967... 0.0834 sec/batch
Epoch: 4/20... Training Step: 15695... Training loss: 1.5835... 0.0874 sec/batch
Epoch: 4/20... Training Step: 15696... Training loss: 1.5507... 0.0890 sec/batch
Epoch: 4/20... Training Step: 15697... Training loss: 1.6607... 0.0836 sec/batch
Epoch: 4/20... Training Step: 15698... Training loss: 1.6183... 0.0844 sec/batch
Epoch: 4/20... Training Step: 15699... Training loss: 1.5600... 0.0989 sec/batch
Epoch: 4/20... Training Step: 15700... Training loss: 1.6233... 0.0844 sec/batch
Epoch: 4/20... Training Step: 15701... Training loss: 1.5357... 0.0903 sec/batch
Epoch: 4/20... Training Step: 15702... Training loss: 1.7038... 0.0957 sec/batch
Epoch: 4/20... Training Step: 15703... Training loss: 1.7391... 0.0845 sec/batch
Epoch: 4/20... Training Step: 15704... Training loss: 1.8974... 0.0886 sec/batch
Epoch: 4/20... Training Step: 15705... Training loss: 1.8397... 0.0936 sec/batch
Epoch: 4/20... Training Step: 15706... Training loss: 1.8366... 0.0815 sec/batch
Epoch: 4/20... Training Step: 15707... Training loss: 1.8484... 0.0851 sec/batch
Epoch: 4/20... Training Step: 15708... Training loss: 1.9053... 0.0965 sec/batch
Epoch: 4/20... Training Step: 15709... Training loss: 1.7512... 0.0850 sec/batch
Epoch: 4/20... Training Step: 15710... Training loss: 1.6996... 0.0894 sec/batch
Epoch: 4/20... Training Step: 15711... Training loss: 1.8305... 0.1019 sec/batch
Epoch: 4/20... Training Step: 15712... Training loss: 1.6250... 0.0922 sec/batch
Epoch: 4/20... Training Step: 15713... Training loss: 1.6699... 0.0759 sec/batch
Epoch: 4/20... Training Step: 15714... Training loss: 1.5547... 0.0608 sec/batch
Epoch: 4/20... Training Step: 15715... Training loss: 1.6978... 0.0656 sec/batch
Epoch: 4/20... Training Step: 15716... Training loss: 1.5997... 0.0707 sec/batch
Epoch: 4/20... Training Step: 15717... Training loss: 1.7302... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15718... Training loss: 2.0471... 0.0732 sec/batch
Epoch: 4/20... Training Step: 15719... Training loss: 1.7219... 0.0705 sec/batch
Epoch: 4/20... Training Step: 15720... Training loss: 1.8050... 0.0753 sec/batch
Epoch: 4/20... Training Step: 15721... Training loss: 1.5722... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15722... Training loss: 1.8181... 0.0666 sec/batch
Epoch: 4/20... Training Step: 15723... Training loss: 1.5113... 0.0613 sec/batch
Epoch: 4/20... Training Step: 15724... Training loss: 1.6214... 0.0709 sec/batch
Epoch: 4/20... Training Step: 15725... Training loss: 1.8552... 0.0687 sec/batch
Epoch: 4/20... Training Step: 15726... Training loss: 2.0289... 0.0771 sec/batch
Epoch: 4/20... Training Step: 15727... Training loss: 2.0734... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15728... Training loss: 1.9997... 0.0739 sec/batch
Epoch: 4/20... Training Step: 15729... Training loss: 1.9818... 0.0754 sec/batch
Epoch: 4/20... Training Step: 15730... Training loss: 2.0172... 0.0753 sec/batch
Epoch: 4/20... Training Step: 15731... Training loss: 1.9335... 0.0698 sec/batch
Epoch: 4/20... Training Step: 15732... Training loss: 1.9749... 0.0734 sec/batch
Epoch: 4/20... Training Step: 15733... Training loss: 2.0890... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15734... Training loss: 1.8139... 0.0756 sec/batch
Epoch: 4/20... Training Step: 15735... Training loss: 1.6320... 0.0629 sec/batch
Epoch: 4/20... Training Step: 15736... Training loss: 1.6762... 0.0678 sec/batch
Epoch: 4/20... Training Step: 15737... Training loss: 1.6503... 0.0672 sec/batch
Epoch: 4/20... Training Step: 15738... Training loss: 1.7681... 0.0724 sec/batch
Epoch: 4/20... Training Step: 15739... Training loss: 1.6472... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15740... Training loss: 1.7250... 0.0711 sec/batch
Epoch: 4/20... Training Step: 15741... Training loss: 1.5746... 0.0614 sec/batch
Epoch: 4/20... Training Step: 15742... Training loss: 1.5642... 0.0617 sec/batch
Epoch: 4/20... Training Step: 15743... Training loss: 1.6321... 0.0719 sec/batch
Epoch: 4/20... Training Step: 15744... Training loss: 1.5746... 0.0659 sec/batch
Epoch: 4/20... Training Step: 15745... Training loss: 1.7617... 0.0674 sec/batch
Epoch: 4/20... Training Step: 15746... Training loss: 1.5707... 0.0645 sec/batch
Epoch: 4/20... Training Step: 15747... Training loss: 1.7407... 0.0810 sec/batch
Epoch: 4/20... Training Step: 15748... Training loss: 1.7410... 0.0563 sec/batch
Epoch: 4/20... Training Step: 15749... Training loss: 1.8102... 0.0745 sec/batch
Epoch: 4/20... Training Step: 15750... Training loss: 1.8226... 0.0542 sec/batch
Epoch: 4/20... Training Step: 15751... Training loss: 1.7387... 0.0738 sec/batch
Epoch: 4/20... Training Step: 15752... Training loss: 1.8365... 0.0634 sec/batch
Epoch: 4/20... Training Step: 15753... Training loss: 1.9639... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15754... Training loss: 1.9504... 0.0676 sec/batch
Epoch: 4/20... Training Step: 15755... Training loss: 1.7795... 0.0749 sec/batch
Epoch: 4/20... Training Step: 15756... Training loss: 1.7373... 0.0744 sec/batch
Epoch: 4/20... Training Step: 15757... Training loss: 1.7132... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15758... Training loss: 1.7850... 0.0562 sec/batch
Epoch: 4/20... Training Step: 15759... Training loss: 1.5649... 0.0878 sec/batch
Epoch: 4/20... Training Step: 15760... Training loss: 1.5800... 0.0740 sec/batch
Epoch: 4/20... Training Step: 15761... Training loss: 1.8687... 0.0639 sec/batch
Epoch: 4/20... Training Step: 15762... Training loss: 1.7758... 0.0597 sec/batch
Epoch: 4/20... Training Step: 15763... Training loss: 1.9547... 0.0720 sec/batch
Epoch: 4/20... Training Step: 15764... Training loss: 1.5673... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15765... Training loss: 1.7851... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15766... Training loss: 1.5790... 0.0638 sec/batch
Epoch: 4/20... Training Step: 15767... Training loss: 1.8558... 0.0739 sec/batch
Epoch: 4/20... Training Step: 15768... Training loss: 1.6039... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15769... Training loss: 1.7624... 0.0669 sec/batch
Epoch: 4/20... Training Step: 15770... Training loss: 1.7833... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15771... Training loss: 1.8143... 0.0753 sec/batch
Epoch: 4/20... Training Step: 15772... Training loss: 1.5557... 0.0605 sec/batch
Epoch: 4/20... Training Step: 15773... Training loss: 1.7210... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15774... Training loss: 1.6752... 0.0558 sec/batch
Epoch: 4/20... Training Step: 15775... Training loss: 1.7354... 0.0752 sec/batch
Epoch: 4/20... Training Step: 15776... Training loss: 1.7734... 0.0633 sec/batch
Epoch: 4/20... Training Step: 15777... Training loss: 1.8052... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15778... Training loss: 1.5941... 0.0675 sec/batch
Epoch: 4/20... Training Step: 15779... Training loss: 1.7397... 0.0666 sec/batch
Epoch: 4/20... Training Step: 15780... Training loss: 1.7960... 0.0594 sec/batch
Epoch: 4/20... Training Step: 15781... Training loss: 1.7548... 0.0622 sec/batch
Epoch: 4/20... Training Step: 15782... Training loss: 1.6855... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15783... Training loss: 1.7340... 0.0647 sec/batch
Epoch: 4/20... Training Step: 15784... Training loss: 1.6249... 0.0681 sec/batch
Epoch: 4/20... Training Step: 15785... Training loss: 1.6095... 0.0642 sec/batch
Epoch: 4/20... Training Step: 15786... Training loss: 1.6545... 0.0538 sec/batch
Epoch: 4/20... Training Step: 15787... Training loss: 1.7477... 0.0691 sec/batch
Epoch: 4/20... Training Step: 15788... Training loss: 1.6826... 0.0616 sec/batch
Epoch: 4/20... Training Step: 15789... Training loss: 1.7414... 0.0521 sec/batch
Epoch: 4/20... Training Step: 15790... Training loss: 1.6584... 0.0655 sec/batch
Epoch: 4/20... Training Step: 15791... Training loss: 1.7923... 0.0685 sec/batch
Epoch: 4/20... Training Step: 15792... Training loss: 1.8170... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15793... Training loss: 1.6983... 0.0647 sec/batch
Epoch: 4/20... Training Step: 15794... Training loss: 1.6653... 0.0606 sec/batch
Epoch: 4/20... Training Step: 15795... Training loss: 1.7216... 0.0691 sec/batch
Epoch: 4/20... Training Step: 15796... Training loss: 1.6270... 0.0612 sec/batch
Epoch: 4/20... Training Step: 15797... Training loss: 1.8071... 0.0685 sec/batch
Epoch: 4/20... Training Step: 15798... Training loss: 1.7822... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15799... Training loss: 1.7035... 0.0700 sec/batch
Epoch: 4/20... Training Step: 15800... Training loss: 1.8018... 0.0602 sec/batch
Epoch: 4/20... Training Step: 15801... Training loss: 1.7400... 0.0683 sec/batch
Epoch: 4/20... Training Step: 15802... Training loss: 1.6601... 0.0662 sec/batch
Epoch: 4/20... Training Step: 15803... Training loss: 1.6722... 0.0628 sec/batch
Epoch: 4/20... Training Step: 15804... Training loss: 1.7954... 0.0695 sec/batch
Epoch: 4/20... Training Step: 15805... Training loss: 1.7496... 0.0748 sec/batch
Epoch: 4/20... Training Step: 15806... Training loss: 1.9375... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15807... Training loss: 1.8308... 0.0541 sec/batch
Epoch: 4/20... Training Step: 15808... Training loss: 1.7621... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15809... Training loss: 1.7767... 0.0646 sec/batch
Epoch: 4/20... Training Step: 15810... Training loss: 1.8524... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15811... Training loss: 1.7673... 0.0584 sec/batch
Epoch: 4/20... Training Step: 15812... Training loss: 2.0464... 0.0656 sec/batch
Epoch: 4/20... Training Step: 15813... Training loss: 1.7803... 0.0572 sec/batch
Epoch: 4/20... Training Step: 15814... Training loss: 1.8164... 0.0613 sec/batch
Epoch: 4/20... Training Step: 15815... Training loss: 1.9396... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15816... Training loss: 1.9909... 0.0630 sec/batch
Epoch: 4/20... Training Step: 15817... Training loss: 1.8012... 0.0651 sec/batch
Epoch: 4/20... Training Step: 15818... Training loss: 1.7270... 0.0522 sec/batch
Epoch: 4/20... Training Step: 15819... Training loss: 1.9209... 0.0588 sec/batch
Epoch: 4/20... Training Step: 15820... Training loss: 1.8750... 0.0641 sec/batch
Epoch: 4/20... Training Step: 15821... Training loss: 1.9406... 0.0716 sec/batch
Epoch: 4/20... Training Step: 15822... Training loss: 1.7690... 0.0652 sec/batch
Epoch: 4/20... Training Step: 15823... Training loss: 1.7064... 0.0636 sec/batch
Epoch: 4/20... Training Step: 15824... Training loss: 1.7518... 0.0578 sec/batch
Epoch: 4/20... Training Step: 15825... Training loss: 1.6699... 0.0683 sec/batch
Epoch: 4/20... Training Step: 15826... Training loss: 1.6275... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15827... Training loss: 1.7654... 0.0593 sec/batch
Epoch: 4/20... Training Step: 15828... Training loss: 1.7066... 0.0607 sec/batch
Epoch: 4/20... Training Step: 15829... Training loss: 1.7115... 0.0587 sec/batch
Epoch: 4/20... Training Step: 15830... Training loss: 1.9878... 0.0644 sec/batch
Epoch: 4/20... Training Step: 15831... Training loss: 1.7053... 0.0619 sec/batch
Epoch: 4/20... Training Step: 15832... Training loss: 1.8311... 0.0564 sec/batch
Epoch: 4/20... Training Step: 15833... Training loss: 1.6467... 0.0689 sec/batch
Epoch: 4/20... Training Step: 15834... Training loss: 1.7491... 0.0634 sec/batch
Epoch: 4/20... Training Step: 15835... Training loss: 1.8188... 0.0600 sec/batch
Epoch: 4/20... Training Step: 15836... Training loss: 1.7560... 0.0513 sec/batch
Epoch: 4/20... Training Step: 15837... Training loss: 1.6466... 0.0654 sec/batch
Epoch: 4/20... Training Step: 15838... Training loss: 1.6708... 0.0531 sec/batch
Epoch: 4/20... Training Step: 15839... Training loss: 1.6761... 0.0518 sec/batch
Epoch: 4/20... Training Step: 15840... Training loss: 1.8206... 0.0550 sec/batch
Epoch: 4/20... Training Step: 15841... Training loss: 1.6884... 0.0618 sec/batch
Epoch: 4/20... Training Step: 15842... Training loss: 1.9476... 0.0534 sec/batch
Epoch: 4/20... Training Step: 15843... Training loss: 1.8757... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15844... Training loss: 1.6888... 0.0512 sec/batch
Epoch: 4/20... Training Step: 15845... Training loss: 1.6620... 0.0735 sec/batch
Epoch: 4/20... Training Step: 15846... Training loss: 1.6289... 0.0585 sec/batch
Epoch: 4/20... Training Step: 15847... Training loss: 1.8315... 0.0592 sec/batch
Epoch: 4/20... Training Step: 15848... Training loss: 1.6056... 0.0635 sec/batch
Epoch: 4/20... Training Step: 15849... Training loss: 1.9027... 0.0624 sec/batch
Epoch: 4/20... Training Step: 15850... Training loss: 1.7004... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15851... Training loss: 1.7529... 0.0623 sec/batch
Epoch: 4/20... Training Step: 15852... Training loss: 1.7370... 0.0529 sec/batch
Epoch: 4/20... Training Step: 15853... Training loss: 1.6760... 0.0530 sec/batch
Epoch: 4/20... Training Step: 15854... Training loss: 1.6050... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15855... Training loss: 1.6831... 0.0613 sec/batch
Epoch: 4/20... Training Step: 15856... Training loss: 1.9460... 0.0569 sec/batch
Epoch: 4/20... Training Step: 15857... Training loss: 1.7971... 0.0660 sec/batch
Epoch: 4/20... Training Step: 15858... Training loss: 1.7585... 0.0530 sec/batch
Epoch: 4/20... Training Step: 15859... Training loss: 1.7594... 0.0586 sec/batch
Epoch: 4/20... Training Step: 15860... Training loss: 1.6062... 0.0505 sec/batch
Epoch: 4/20... Training Step: 15861... Training loss: 1.6652... 0.0576 sec/batch
Epoch: 4/20... Training Step: 15862... Training loss: 1.5678... 0.0660 sec/batch
Epoch: 4/20... Training Step: 15863... Training loss: 1.6118... 0.0535 sec/batch
Epoch: 4/20... Training Step: 15864... Training loss: 1.6944... 0.0621 sec/batch
Epoch: 4/20... Training Step: 15865... Training loss: 1.5615... 0.0673 sec/batch
Epoch: 4/20... Training Step: 15866... Training loss: 1.8422... 0.0522 sec/batch
Epoch: 4/20... Training Step: 15867... Training loss: 1.7982... 0.0568 sec/batch
Epoch: 4/20... Training Step: 15868... Training loss: 1.7009... 0.0632 sec/batch
Epoch: 4/20... Training Step: 15869... Training loss: 1.5401... 0.0582 sec/batch
Epoch: 4/20... Training Step: 15870... Training loss: 1.7400... 0.0626 sec/batch
Epoch: 4/20... Training Step: 15871... Training loss: 1.6382... 0.0546 sec/batch
Epoch: 4/20... Training Step: 15872... Training loss: 1.8164... 0.0620 sec/batch
Epoch: 4/20... Training Step: 15873... Training loss: 1.7429... 0.0671 sec/batch
Epoch: 4/20... Training Step: 15874... Training loss: 1.6569... 0.0527 sec/batch
Epoch: 4/20... Training Step: 15875... Training loss: 1.7224... 0.0565 sec/batch
Epoch: 4/20... Training Step: 15876... Training loss: 1.6914... 0.0526 sec/batch
Epoch: 4/20... Training Step: 15877... Training loss: 1.6553... 0.0549 sec/batch
Epoch: 4/20... Training Step: 15878... Training loss: 1.7356... 0.0609 sec/batch
Epoch: 4/20... Training Step: 15879... Training loss: 1.7638... 0.0610 sec/batch
Epoch: 4/20... Training Step: 15880... Training loss: 1.7834... 0.0543 sec/batch
Epoch: 5/20... Training Step: 15881... Training loss: 1.6702... 0.0730 sec/batch
Epoch: 5/20... Training Step: 15882... Training loss: 1.6763... 0.0624 sec/batch
Epoch: 5/20... Training Step: 15883... Training loss: 1.5984... 0.0582 sec/batch
Epoch: 5/20... Training Step: 15884... Training loss: 1.6166... 0.0597 sec/batch
Epoch: 5/20... Training Step: 15885... Training loss: 1.7233... 0.0599 sec/batch
Epoch: 5/20... Training Step: 15886... Training loss: 1.6005... 0.0528 sec/batch
Epoch: 5/20... Training Step: 15887... Training loss: 1.5748... 0.0633 sec/batch
Epoch: 5/20... Training Step: 15888... Training loss: 1.6443... 0.0517 sec/batch
Epoch: 5/20... Training Step: 15889... Training loss: 1.7486... 0.0550 sec/batch
Epoch: 5/20... Training Step: 15890... Training loss: 1.7211... 0.0561 sec/batch
Epoch: 5/20... Training Step: 15891... Training loss: 1.6281... 0.0614 sec/batch
Epoch: 5/20... Training Step: 15892... Training loss: 1.5329... 0.0633 sec/batch
Epoch: 5/20... Training Step: 15893... Training loss: 1.6005... 0.0723 sec/batch
Epoch: 5/20... Training Step: 15894... Training loss: 1.6082... 0.0606 sec/batch
Epoch: 5/20... Training Step: 15895... Training loss: 1.5460... 0.0561 sec/batch
Epoch: 5/20... Training Step: 15896... Training loss: 1.6606... 0.0527 sec/batch
Epoch: 5/20... Training Step: 15897... Training loss: 1.6632... 0.0601 sec/batch
Epoch: 5/20... Training Step: 15898... Training loss: 1.5560... 0.0661 sec/batch
Epoch: 5/20... Training Step: 15899... Training loss: 1.7622... 0.0595 sec/batch
Epoch: 5/20... Training Step: 15900... Training loss: 1.8256... 0.0574 sec/batch
Epoch: 5/20... Training Step: 15901... Training loss: 1.6041... 0.0657 sec/batch
Epoch: 5/20... Training Step: 15902... Training loss: 1.6762... 0.0625 sec/batch
Epoch: 5/20... Training Step: 15903... Training loss: 1.6159... 0.0597 sec/batch
Epoch: 5/20... Training Step: 15904... Training loss: 1.5909... 0.0598 sec/batch
Epoch: 5/20... Training Step: 15905... Training loss: 1.6085... 0.0663 sec/batch
Epoch: 5/20... Training Step: 15906... Training loss: 1.6857... 0.0611 sec/batch
Epoch: 5/20... Training Step: 15907... Training loss: 1.8070... 0.0628 sec/batch
Epoch: 5/20... Training Step: 15908... Training loss: 1.8727... 0.0625 sec/batch
Epoch: 5/20... Training Step: 15909... Training loss: 1.8197... 0.0624 sec/batch
Epoch: 5/20... Training Step: 15910... Training loss: 1.6216... 0.0625 sec/batch
Epoch: 5/20... Training Step: 15911... Training loss: 1.7617... 0.0641 sec/batch
Epoch: 5/20... Training Step: 15912... Training loss: 1.8969... 0.0684 sec/batch
Epoch: 5/20... Training Step: 15913... Training loss: 1.7385... 0.0639 sec/batch
Epoch: 5/20... Training Step: 15914... Training loss: 1.6533... 0.0581 sec/batch
Epoch: 5/20... Training Step: 15915... Training loss: 1.6315... 0.0635 sec/batch
Epoch: 5/20... Training Step: 15916... Training loss: 1.5185... 0.0603 sec/batch
Epoch: 5/20... Training Step: 15917... Training loss: 1.6705... 0.0687 sec/batch
Epoch: 5/20... Training Step: 15918... Training loss: 1.7503... 0.0602 sec/batch
Epoch: 5/20... Training Step: 15919... Training loss: 1.7073... 0.0605 sec/batch
Epoch: 5/20... Training Step: 15920... Training loss: 1.7370... 0.0624 sec/batch
Epoch: 5/20... Training Step: 15921... Training loss: 1.7025... 0.0733 sec/batch
Epoch: 5/20... Training Step: 15922... Training loss: 1.6691... 0.0644 sec/batch
Epoch: 5/20... Training Step: 15923... Training loss: 1.6121... 0.0610 sec/batch
Epoch: 5/20... Training Step: 15924... Training loss: 1.6371... 0.0589 sec/batch
Epoch: 5/20... Training Step: 15925... Training loss: 1.6320... 0.0692 sec/batch
Epoch: 5/20... Training Step: 15926... Training loss: 1.8337... 0.0603 sec/batch
Epoch: 5/20... Training Step: 15927... Training loss: 1.6900... 0.0597 sec/batch
Epoch: 5/20... Training Step: 15928... Training loss: 1.6434... 0.0598 sec/batch
Epoch: 5/20... Training Step: 15929... Training loss: 1.7609... 0.0701 sec/batch
Epoch: 5/20... Training Step: 15930... Training loss: 1.7683... 0.0599 sec/batch
Epoch: 5/20... Training Step: 15931... Training loss: 1.7613... 0.0601 sec/batch
Epoch: 5/20... Training Step: 15932... Training loss: 1.6712... 0.0637 sec/batch
Epoch: 5/20... Training Step: 15933... Training loss: 1.8298... 0.0687 sec/batch
Epoch: 5/20... Training Step: 15934... Training loss: 1.8080... 0.0668 sec/batch
Epoch: 5/20... Training Step: 15935... Training loss: 1.6901... 0.0534 sec/batch
Epoch: 5/20... Training Step: 15936... Training loss: 1.6279... 0.0647 sec/batch
Epoch: 5/20... Training Step: 15937... Training loss: 1.7037... 0.0650 sec/batch
Epoch: 5/20... Training Step: 15938... Training loss: 1.6321... 0.0657 sec/batch
Epoch: 5/20... Training Step: 15939... Training loss: 1.6933... 0.0607 sec/batch
Epoch: 5/20... Training Step: 15940... Training loss: 1.7292... 0.0603 sec/batch
Epoch: 5/20... Training Step: 15941... Training loss: 1.8479... 0.0709 sec/batch
Epoch: 5/20... Training Step: 15942... Training loss: 1.8387... 0.0555 sec/batch
Epoch: 5/20... Training Step: 15943... Training loss: 1.7327... 0.0624 sec/batch
Epoch: 5/20... Training Step: 15944... Training loss: 1.5354... 0.0622 sec/batch
Epoch: 5/20... Training Step: 15945... Training loss: 1.8694... 0.0675 sec/batch
Epoch: 5/20... Training Step: 15946... Training loss: 1.7518... 0.0570 sec/batch
Epoch: 5/20... Training Step: 15947... Training loss: 1.6957... 0.0653 sec/batch
Epoch: 5/20... Training Step: 15948... Training loss: 1.8868... 0.0513 sec/batch
Epoch: 5/20... Training Step: 15949... Training loss: 1.8968... 0.0744 sec/batch
Epoch: 5/20... Training Step: 15950... Training loss: 1.6997... 0.0564 sec/batch
Epoch: 5/20... Training Step: 15951... Training loss: 1.6986... 0.0644 sec/batch
Epoch: 5/20... Training Step: 15952... Training loss: 1.6836... 0.0671 sec/batch
Epoch: 5/20... Training Step: 15953... Training loss: 1.5989... 0.0625 sec/batch
Epoch: 5/20... Training Step: 15954... Training loss: 1.7496... 0.0646 sec/batch
Epoch: 5/20... Training Step: 15955... Training loss: 1.5790... 0.0627 sec/batch
Epoch: 5/20... Training Step: 15956... Training loss: 1.6225... 0.0628 sec/batch
Epoch: 5/20... Training Step: 15957... Training loss: 1.6244... 0.0682 sec/batch
Epoch: 5/20... Training Step: 15958... Training loss: 1.6021... 0.0534 sec/batch
Epoch: 5/20... Training Step: 15959... Training loss: 1.6306... 0.0566 sec/batch
Epoch: 5/20... Training Step: 15960... Training loss: 1.6571... 0.0621 sec/batch
Epoch: 5/20... Training Step: 15961... Training loss: 1.5981... 0.0736 sec/batch
Epoch: 5/20... Training Step: 15962... Training loss: 1.5913... 0.0636 sec/batch
Epoch: 5/20... Training Step: 15963... Training loss: 1.6611... 0.0601 sec/batch
Epoch: 5/20... Training Step: 15964... Training loss: 1.6161... 0.0557 sec/batch
Epoch: 5/20... Training Step: 15965... Training loss: 1.5328... 0.0547 sec/batch
Epoch: 5/20... Training Step: 15966... Training loss: 1.6048... 0.0578 sec/batch
Epoch: 5/20... Training Step: 15967... Training loss: 1.7037... 0.0576 sec/batch
Epoch: 5/20... Training Step: 15968... Training loss: 1.6148... 0.0596 sec/batch
Epoch: 5/20... Training Step: 15969... Training loss: 1.5803... 0.0665 sec/batch
Epoch: 5/20... Training Step: 15970... Training loss: 1.7096... 0.0605 sec/batch
Epoch: 5/20... Training Step: 15971... Training loss: 1.6095... 0.0515 sec/batch
Epoch: 5/20... Training Step: 15972... Training loss: 1.5988... 0.0564 sec/batch
Epoch: 5/20... Training Step: 15973... Training loss: 1.6494... 0.0611 sec/batch
Epoch: 5/20... Training Step: 15974... Training loss: 1.7070... 0.0568 sec/batch
Epoch: 5/20... Training Step: 15975... Training loss: 1.5972... 0.0543 sec/batch
Epoch: 5/20... Training Step: 15976... Training loss: 1.6364... 0.0542 sec/batch
Epoch: 5/20... Training Step: 15977... Training loss: 1.7174... 0.0598 sec/batch
Epoch: 5/20... Training Step: 15978... Training loss: 1.7724... 0.0585 sec/batch
Epoch: 5/20... Training Step: 15979... Training loss: 1.6266... 0.0523 sec/batch
Epoch: 5/20... Training Step: 15980... Training loss: 1.6211... 0.0625 sec/batch
Epoch: 5/20... Training Step: 15981... Training loss: 1.7694... 0.0686 sec/batch
Epoch: 5/20... Training Step: 15982... Training loss: 1.6109... 0.0584 sec/batch
Epoch: 5/20... Training Step: 15983... Training loss: 1.6680... 0.0602 sec/batch
Epoch: 5/20... Training Step: 15984... Training loss: 1.6328... 0.0621 sec/batch
Epoch: 5/20... Training Step: 15985... Training loss: 1.6787... 0.0653 sec/batch
Epoch: 5/20... Training Step: 15986... Training loss: 1.6344... 0.0640 sec/batch
Epoch: 5/20... Training Step: 15987... Training loss: 1.5523... 0.0613 sec/batch
Epoch: 5/20... Training Step: 15988... Training loss: 1.5737... 0.0578 sec/batch
Epoch: 5/20... Training Step: 15989... Training loss: 1.5847... 0.0641 sec/batch
Epoch: 5/20... Training Step: 15990... Training loss: 1.6699... 0.0583 sec/batch
Epoch: 5/20... Training Step: 15991... Training loss: 1.5255... 0.0604 sec/batch
Epoch: 5/20... Training Step: 15992... Training loss: 1.6665... 0.0586 sec/batch
Epoch: 5/20... Training Step: 15993... Training loss: 1.5673... 0.0624 sec/batch
Epoch: 5/20... Training Step: 15994... Training loss: 1.6667... 0.0496 sec/batch
Epoch: 5/20... Training Step: 15995... Training loss: 1.6776... 0.0576 sec/batch
Epoch: 5/20... Training Step: 15996... Training loss: 1.6891... 0.0632 sec/batch
Epoch: 5/20... Training Step: 15997... Training loss: 1.5576... 0.0535 sec/batch
Epoch: 5/20... Training Step: 15998... Training loss: 1.4867... 0.0549 sec/batch
Epoch: 5/20... Training Step: 15999... Training loss: 1.5742... 0.0547 sec/batch
Epoch: 5/20... Training Step: 16000... Training loss: 1.6764... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16001... Training loss: 1.5215... 0.0511 sec/batch
Epoch: 5/20... Training Step: 16002... Training loss: 1.5794... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16003... Training loss: 1.6658... 0.0512 sec/batch
Epoch: 5/20... Training Step: 16004... Training loss: 1.6352... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16005... Training loss: 1.6115... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16006... Training loss: 1.7191... 0.0526 sec/batch
Epoch: 5/20... Training Step: 16007... Training loss: 1.5501... 0.0535 sec/batch
Epoch: 5/20... Training Step: 16008... Training loss: 1.5858... 0.0556 sec/batch
Epoch: 5/20... Training Step: 16009... Training loss: 1.7299... 0.0543 sec/batch
Epoch: 5/20... Training Step: 16010... Training loss: 1.6278... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16011... Training loss: 1.7279... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16012... Training loss: 1.6408... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16013... Training loss: 1.7432... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16014... Training loss: 1.6181... 0.0588 sec/batch
Epoch: 5/20... Training Step: 16015... Training loss: 1.6864... 0.0645 sec/batch
Epoch: 5/20... Training Step: 16016... Training loss: 1.5637... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16017... Training loss: 1.6759... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16018... Training loss: 1.5646... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16019... Training loss: 1.6816... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16020... Training loss: 1.6319... 0.0523 sec/batch
Epoch: 5/20... Training Step: 16021... Training loss: 1.6965... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16022... Training loss: 1.5952... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16023... Training loss: 1.6468... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16024... Training loss: 1.7467... 0.0568 sec/batch
Epoch: 5/20... Training Step: 16025... Training loss: 1.6157... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16026... Training loss: 1.6593... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16027... Training loss: 1.6840... 0.0513 sec/batch
Epoch: 5/20... Training Step: 16028... Training loss: 1.5437... 0.0548 sec/batch
Epoch: 5/20... Training Step: 16029... Training loss: 1.5485... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16030... Training loss: 1.4711... 0.0556 sec/batch
Epoch: 5/20... Training Step: 16031... Training loss: 1.7161... 0.0539 sec/batch
Epoch: 5/20... Training Step: 16032... Training loss: 1.6354... 0.0546 sec/batch
Epoch: 5/20... Training Step: 16033... Training loss: 1.5690... 0.0549 sec/batch
Epoch: 5/20... Training Step: 16034... Training loss: 1.6382... 0.0568 sec/batch
Epoch: 5/20... Training Step: 16035... Training loss: 1.8831... 0.0575 sec/batch
Epoch: 5/20... Training Step: 16036... Training loss: 1.6465... 0.0609 sec/batch
Epoch: 5/20... Training Step: 16037... Training loss: 1.4615... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16038... Training loss: 1.7696... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16039... Training loss: 1.8577... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16040... Training loss: 1.8394... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16041... Training loss: 1.6748... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16042... Training loss: 1.6544... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16043... Training loss: 1.6877... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16044... Training loss: 1.6907... 0.0528 sec/batch
Epoch: 5/20... Training Step: 16045... Training loss: 1.6859... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16046... Training loss: 1.6982... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16047... Training loss: 1.6586... 0.0516 sec/batch
Epoch: 5/20... Training Step: 16048... Training loss: 1.8139... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16049... Training loss: 1.6416... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16050... Training loss: 1.7300... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16051... Training loss: 1.7144... 0.0546 sec/batch
Epoch: 5/20... Training Step: 16052... Training loss: 1.7212... 0.0584 sec/batch
Epoch: 5/20... Training Step: 16053... Training loss: 1.6971... 0.0748 sec/batch
Epoch: 5/20... Training Step: 16054... Training loss: 1.6635... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16055... Training loss: 1.6853... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16056... Training loss: 1.6440... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16057... Training loss: 1.6192... 0.0691 sec/batch
Epoch: 5/20... Training Step: 16058... Training loss: 1.5336... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16059... Training loss: 1.6231... 0.0637 sec/batch
Epoch: 5/20... Training Step: 16060... Training loss: 1.5720... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16061... Training loss: 1.6465... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16062... Training loss: 1.7385... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16063... Training loss: 1.6838... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16064... Training loss: 1.5912... 0.0674 sec/batch
Epoch: 5/20... Training Step: 16065... Training loss: 1.5989... 0.0646 sec/batch
Epoch: 5/20... Training Step: 16066... Training loss: 1.6864... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16067... Training loss: 1.4998... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16068... Training loss: 1.5796... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16069... Training loss: 1.6427... 0.0664 sec/batch
Epoch: 5/20... Training Step: 16070... Training loss: 1.6853... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16071... Training loss: 1.6180... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16072... Training loss: 1.6210... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16073... Training loss: 1.6731... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16074... Training loss: 1.7968... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16075... Training loss: 1.6895... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16076... Training loss: 1.5491... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16077... Training loss: 1.5316... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16078... Training loss: 1.7092... 0.0571 sec/batch
Epoch: 5/20... Training Step: 16079... Training loss: 1.6777... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16080... Training loss: 1.6935... 0.0527 sec/batch
Epoch: 5/20... Training Step: 16081... Training loss: 1.5605... 0.0784 sec/batch
Epoch: 5/20... Training Step: 16082... Training loss: 1.7513... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16083... Training loss: 1.5873... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16084... Training loss: 1.7724... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16085... Training loss: 1.7276... 0.0580 sec/batch
Epoch: 5/20... Training Step: 16086... Training loss: 1.6819... 0.0627 sec/batch
Epoch: 5/20... Training Step: 16087... Training loss: 1.6296... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16088... Training loss: 1.5999... 0.0569 sec/batch
Epoch: 5/20... Training Step: 16089... Training loss: 1.7599... 0.0662 sec/batch
Epoch: 5/20... Training Step: 16090... Training loss: 1.6573... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16091... Training loss: 1.5341... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16092... Training loss: 1.7016... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16093... Training loss: 1.7285... 0.0737 sec/batch
Epoch: 5/20... Training Step: 16094... Training loss: 1.6997... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16095... Training loss: 1.6427... 0.0539 sec/batch
Epoch: 5/20... Training Step: 16096... Training loss: 1.6930... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16097... Training loss: 1.7149... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16098... Training loss: 1.7395... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16099... Training loss: 1.6885... 0.0652 sec/batch
Epoch: 5/20... Training Step: 16100... Training loss: 1.6393... 0.0518 sec/batch
Epoch: 5/20... Training Step: 16101... Training loss: 1.6352... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16102... Training loss: 1.6685... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16103... Training loss: 1.7640... 0.0539 sec/batch
Epoch: 5/20... Training Step: 16104... Training loss: 1.7741... 0.0549 sec/batch
Epoch: 5/20... Training Step: 16105... Training loss: 1.6929... 0.0692 sec/batch
Epoch: 5/20... Training Step: 16106... Training loss: 1.7265... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16107... Training loss: 1.8355... 0.0569 sec/batch
Epoch: 5/20... Training Step: 16108... Training loss: 1.7937... 0.0528 sec/batch
Epoch: 5/20... Training Step: 16109... Training loss: 1.6187... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16110... Training loss: 1.6065... 0.0537 sec/batch
Epoch: 5/20... Training Step: 16111... Training loss: 1.6835... 0.0581 sec/batch
Epoch: 5/20... Training Step: 16112... Training loss: 1.6396... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16113... Training loss: 1.6371... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16114... Training loss: 1.6869... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16115... Training loss: 1.6714... 0.0553 sec/batch
Epoch: 5/20... Training Step: 16116... Training loss: 1.8380... 0.0537 sec/batch
Epoch: 5/20... Training Step: 16117... Training loss: 1.8181... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16118... Training loss: 1.7933... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16119... Training loss: 1.7238... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16120... Training loss: 1.6836... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16121... Training loss: 1.7617... 0.0718 sec/batch
Epoch: 5/20... Training Step: 16122... Training loss: 1.8115... 0.0560 sec/batch
Epoch: 5/20... Training Step: 16123... Training loss: 1.6696... 0.0528 sec/batch
Epoch: 5/20... Training Step: 16124... Training loss: 1.7777... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16125... Training loss: 1.7822... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16126... Training loss: 1.8215... 0.0743 sec/batch
Epoch: 5/20... Training Step: 16127... Training loss: 1.6940... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16128... Training loss: 1.6788... 0.0633 sec/batch
Epoch: 5/20... Training Step: 16129... Training loss: 1.5161... 0.0672 sec/batch
Epoch: 5/20... Training Step: 16130... Training loss: 1.6521... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16131... Training loss: 1.7054... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16132... Training loss: 1.7970... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16133... Training loss: 1.7185... 0.0665 sec/batch
Epoch: 5/20... Training Step: 16134... Training loss: 1.6037... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16135... Training loss: 1.8140... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16136... Training loss: 1.6736... 0.0563 sec/batch
Epoch: 5/20... Training Step: 16137... Training loss: 1.6562... 0.0660 sec/batch
Epoch: 5/20... Training Step: 16138... Training loss: 1.6055... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16139... Training loss: 1.6325... 0.0519 sec/batch
Epoch: 5/20... Training Step: 16140... Training loss: 1.6284... 0.0534 sec/batch
Epoch: 5/20... Training Step: 16141... Training loss: 1.6306... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16142... Training loss: 1.5653... 0.0561 sec/batch
Epoch: 5/20... Training Step: 16143... Training loss: 1.6913... 0.0560 sec/batch
Epoch: 5/20... Training Step: 16144... Training loss: 1.6356... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16145... Training loss: 1.5455... 0.0590 sec/batch
Epoch: 5/20... Training Step: 16146... Training loss: 1.5749... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16147... Training loss: 1.5878... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16148... Training loss: 1.6517... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16149... Training loss: 1.6552... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16150... Training loss: 1.7409... 0.0546 sec/batch
Epoch: 5/20... Training Step: 16151... Training loss: 1.6709... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16152... Training loss: 1.6339... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16153... Training loss: 1.8743... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16154... Training loss: 1.7855... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16155... Training loss: 1.7059... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16156... Training loss: 1.6554... 0.0560 sec/batch
Epoch: 5/20... Training Step: 16157... Training loss: 1.6040... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16158... Training loss: 1.7335... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16159... Training loss: 1.6504... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16160... Training loss: 1.7185... 0.0505 sec/batch
Epoch: 5/20... Training Step: 16161... Training loss: 1.6257... 0.0532 sec/batch
Epoch: 5/20... Training Step: 16162... Training loss: 1.6083... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16163... Training loss: 1.5552... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16164... Training loss: 1.7526... 0.0544 sec/batch
Epoch: 5/20... Training Step: 16165... Training loss: 1.6027... 0.0793 sec/batch
Epoch: 5/20... Training Step: 16166... Training loss: 1.6362... 0.0684 sec/batch
Epoch: 5/20... Training Step: 16167... Training loss: 1.4936... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16168... Training loss: 1.6504... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16169... Training loss: 1.5852... 0.0655 sec/batch
Epoch: 5/20... Training Step: 16170... Training loss: 1.5890... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16171... Training loss: 1.6455... 0.0642 sec/batch
Epoch: 5/20... Training Step: 16172... Training loss: 1.5847... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16173... Training loss: 1.5762... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16174... Training loss: 1.5962... 0.0636 sec/batch
Epoch: 5/20... Training Step: 16175... Training loss: 1.6871... 0.0652 sec/batch
Epoch: 5/20... Training Step: 16176... Training loss: 1.6482... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16177... Training loss: 1.5554... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16178... Training loss: 1.6977... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16179... Training loss: 1.5395... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16180... Training loss: 1.5582... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16181... Training loss: 1.7406... 0.0702 sec/batch
Epoch: 5/20... Training Step: 16182... Training loss: 1.6367... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16183... Training loss: 1.6377... 0.0643 sec/batch
Epoch: 5/20... Training Step: 16184... Training loss: 1.6569... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16185... Training loss: 1.6966... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16186... Training loss: 1.5331... 0.0645 sec/batch
Epoch: 5/20... Training Step: 16187... Training loss: 1.7950... 0.0633 sec/batch
Epoch: 5/20... Training Step: 16188... Training loss: 1.7170... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16189... Training loss: 1.6905... 0.0730 sec/batch
Epoch: 5/20... Training Step: 16190... Training loss: 1.7680... 0.0503 sec/batch
Epoch: 5/20... Training Step: 16191... Training loss: 1.6810... 0.0735 sec/batch
Epoch: 5/20... Training Step: 16192... Training loss: 1.6782... 0.0642 sec/batch
Epoch: 5/20... Training Step: 16193... Training loss: 1.7044... 0.0560 sec/batch
Epoch: 5/20... Training Step: 16194... Training loss: 1.7290... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16195... Training loss: 1.7002... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16196... Training loss: 1.7972... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16197... Training loss: 1.6323... 0.0710 sec/batch
Epoch: 5/20... Training Step: 16198... Training loss: 1.6142... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16199... Training loss: 1.7221... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16200... Training loss: 1.8192... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16201... Training loss: 1.6691... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16202... Training loss: 1.5107... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16203... Training loss: 1.7888... 0.0636 sec/batch
Epoch: 5/20... Training Step: 16204... Training loss: 1.6709... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16205... Training loss: 1.6353... 0.0654 sec/batch
Epoch: 5/20... Training Step: 16206... Training loss: 1.7386... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16207... Training loss: 1.7212... 0.0688 sec/batch
Epoch: 5/20... Training Step: 16208... Training loss: 1.7449... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16209... Training loss: 1.7393... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16210... Training loss: 1.7132... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16211... Training loss: 1.8145... 0.0715 sec/batch
Epoch: 5/20... Training Step: 16212... Training loss: 1.7461... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16213... Training loss: 1.7003... 0.0656 sec/batch
Epoch: 5/20... Training Step: 16214... Training loss: 1.7931... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16215... Training loss: 1.8450... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16216... Training loss: 1.7715... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16217... Training loss: 1.5933... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16218... Training loss: 1.6076... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16219... Training loss: 1.6404... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16220... Training loss: 1.6525... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16221... Training loss: 1.9265... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16222... Training loss: 1.9530... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16223... Training loss: 1.8151... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16224... Training loss: 1.6638... 0.0523 sec/batch
Epoch: 5/20... Training Step: 16225... Training loss: 1.6951... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16226... Training loss: 1.8149... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16227... Training loss: 1.7983... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16228... Training loss: 1.7691... 0.0707 sec/batch
Epoch: 5/20... Training Step: 16229... Training loss: 1.6991... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16230... Training loss: 1.6605... 0.0581 sec/batch
Epoch: 5/20... Training Step: 16231... Training loss: 1.6920... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16232... Training loss: 1.5664... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16233... Training loss: 1.6325... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16234... Training loss: 1.7137... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16235... Training loss: 1.7006... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16236... Training loss: 1.6106... 0.0698 sec/batch
Epoch: 5/20... Training Step: 16237... Training loss: 1.7347... 0.0644 sec/batch
Epoch: 5/20... Training Step: 16238... Training loss: 1.7286... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16239... Training loss: 1.5572... 0.0556 sec/batch
Epoch: 5/20... Training Step: 16240... Training loss: 1.7421... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16241... Training loss: 1.6871... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16242... Training loss: 1.7000... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16243... Training loss: 1.7668... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16244... Training loss: 1.6118... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16245... Training loss: 1.7402... 0.0663 sec/batch
Epoch: 5/20... Training Step: 16246... Training loss: 1.6414... 0.0543 sec/batch
Epoch: 5/20... Training Step: 16247... Training loss: 1.7359... 0.0552 sec/batch
Epoch: 5/20... Training Step: 16248... Training loss: 1.6827... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16249... Training loss: 1.6953... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16250... Training loss: 1.7545... 0.0573 sec/batch
Epoch: 5/20... Training Step: 16251... Training loss: 1.5901... 0.0505 sec/batch
Epoch: 5/20... Training Step: 16252... Training loss: 1.6691... 0.0563 sec/batch
Epoch: 5/20... Training Step: 16253... Training loss: 1.6905... 0.0568 sec/batch
Epoch: 5/20... Training Step: 16254... Training loss: 1.5715... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16255... Training loss: 1.7851... 0.0659 sec/batch
Epoch: 5/20... Training Step: 16256... Training loss: 1.6041... 0.0498 sec/batch
Epoch: 5/20... Training Step: 16257... Training loss: 1.7519... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16258... Training loss: 1.7201... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16259... Training loss: 1.7132... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16260... Training loss: 1.7417... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16261... Training loss: 1.6839... 0.0642 sec/batch
Epoch: 5/20... Training Step: 16262... Training loss: 1.7601... 0.0503 sec/batch
Epoch: 5/20... Training Step: 16263... Training loss: 1.8481... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16264... Training loss: 1.8713... 0.0636 sec/batch
Epoch: 5/20... Training Step: 16265... Training loss: 1.7277... 0.0648 sec/batch
Epoch: 5/20... Training Step: 16266... Training loss: 1.6753... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16267... Training loss: 1.7036... 0.0577 sec/batch
Epoch: 5/20... Training Step: 16268... Training loss: 1.7919... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16269... Training loss: 1.5907... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16270... Training loss: 1.6947... 0.0606 sec/batch
Epoch: 5/20... Training Step: 16271... Training loss: 1.6497... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16272... Training loss: 1.8282... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16273... Training loss: 1.6476... 0.0668 sec/batch
Epoch: 5/20... Training Step: 16274... Training loss: 1.5762... 0.0557 sec/batch
Epoch: 5/20... Training Step: 16275... Training loss: 1.6802... 0.0575 sec/batch
Epoch: 5/20... Training Step: 16276... Training loss: 1.7062... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16277... Training loss: 1.6345... 0.0660 sec/batch
Epoch: 5/20... Training Step: 16278... Training loss: 1.6294... 0.0577 sec/batch
Epoch: 5/20... Training Step: 16279... Training loss: 1.7088... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16280... Training loss: 1.6579... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16281... Training loss: 1.5701... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16282... Training loss: 1.6885... 0.0569 sec/batch
Epoch: 5/20... Training Step: 16283... Training loss: 1.5996... 0.0522 sec/batch
Epoch: 5/20... Training Step: 16284... Training loss: 1.6421... 0.0539 sec/batch
Epoch: 5/20... Training Step: 16285... Training loss: 1.7866... 0.0510 sec/batch
Epoch: 5/20... Training Step: 16286... Training loss: 1.6463... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16287... Training loss: 1.6760... 0.0569 sec/batch
Epoch: 5/20... Training Step: 16288... Training loss: 1.7258... 0.0548 sec/batch
Epoch: 5/20... Training Step: 16289... Training loss: 1.6336... 0.0744 sec/batch
Epoch: 5/20... Training Step: 16290... Training loss: 1.6893... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16291... Training loss: 1.6896... 0.0626 sec/batch
Epoch: 5/20... Training Step: 16292... Training loss: 1.6584... 0.0659 sec/batch
Epoch: 5/20... Training Step: 16293... Training loss: 1.9421... 0.0656 sec/batch
Epoch: 5/20... Training Step: 16294... Training loss: 1.6532... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16295... Training loss: 1.6842... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16296... Training loss: 1.7363... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16297... Training loss: 1.8357... 0.0721 sec/batch
Epoch: 5/20... Training Step: 16298... Training loss: 1.5802... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16299... Training loss: 1.6229... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16300... Training loss: 1.7279... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16301... Training loss: 1.6114... 0.0736 sec/batch
Epoch: 5/20... Training Step: 16302... Training loss: 1.6840... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16303... Training loss: 1.6202... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16304... Training loss: 1.6988... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16305... Training loss: 1.6947... 0.0735 sec/batch
Epoch: 5/20... Training Step: 16306... Training loss: 1.6951... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16307... Training loss: 1.7047... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16308... Training loss: 1.6941... 0.0585 sec/batch
Epoch: 5/20... Training Step: 16309... Training loss: 1.6892... 0.0689 sec/batch
Epoch: 5/20... Training Step: 16310... Training loss: 1.6795... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16311... Training loss: 1.5764... 0.0609 sec/batch
Epoch: 5/20... Training Step: 16312... Training loss: 1.6799... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16313... Training loss: 1.6323... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16314... Training loss: 1.6045... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16315... Training loss: 1.5810... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16316... Training loss: 1.6527... 0.0525 sec/batch
Epoch: 5/20... Training Step: 16317... Training loss: 1.6996... 0.0633 sec/batch
Epoch: 5/20... Training Step: 16318... Training loss: 1.6666... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16319... Training loss: 1.6342... 0.0517 sec/batch
Epoch: 5/20... Training Step: 16320... Training loss: 1.6968... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16321... Training loss: 1.7957... 0.0782 sec/batch
Epoch: 5/20... Training Step: 16322... Training loss: 1.6004... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16323... Training loss: 1.7951... 0.0608 sec/batch
Epoch: 5/20... Training Step: 16324... Training loss: 1.6102... 0.0644 sec/batch
Epoch: 5/20... Training Step: 16325... Training loss: 1.6984... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16326... Training loss: 1.7699... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16327... Training loss: 1.8130... 0.0681 sec/batch
Epoch: 5/20... Training Step: 16328... Training loss: 1.5276... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16329... Training loss: 1.5848... 0.0634 sec/batch
Epoch: 5/20... Training Step: 16330... Training loss: 1.7638... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16331... Training loss: 1.6503... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16332... Training loss: 1.6915... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16333... Training loss: 1.5557... 0.0772 sec/batch
Epoch: 5/20... Training Step: 16334... Training loss: 1.7170... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16335... Training loss: 1.5997... 0.0614 sec/batch
Epoch: 5/20... Training Step: 16336... Training loss: 1.6289... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16337... Training loss: 1.6212... 0.0666 sec/batch
Epoch: 5/20... Training Step: 16338... Training loss: 1.7012... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16339... Training loss: 1.7710... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16340... Training loss: 1.8244... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16341... Training loss: 1.6676... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16342... Training loss: 1.7492... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16343... Training loss: 1.6823... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16344... Training loss: 1.5996... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16345... Training loss: 1.6536... 0.0766 sec/batch
Epoch: 5/20... Training Step: 16346... Training loss: 1.6428... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16347... Training loss: 1.6996... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16348... Training loss: 1.6135... 0.0654 sec/batch
Epoch: 5/20... Training Step: 16349... Training loss: 1.5704... 0.0709 sec/batch
Epoch: 5/20... Training Step: 16350... Training loss: 1.6514... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16351... Training loss: 1.6299... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16352... Training loss: 1.5902... 0.0557 sec/batch
Epoch: 5/20... Training Step: 16353... Training loss: 1.7263... 0.0669 sec/batch
Epoch: 5/20... Training Step: 16354... Training loss: 1.7500... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16355... Training loss: 1.6567... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16356... Training loss: 1.5230... 0.0519 sec/batch
Epoch: 5/20... Training Step: 16357... Training loss: 1.5648... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16358... Training loss: 1.5602... 0.0574 sec/batch
Epoch: 5/20... Training Step: 16359... Training loss: 1.5791... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16360... Training loss: 1.7388... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16361... Training loss: 1.5822... 0.0532 sec/batch
Epoch: 5/20... Training Step: 16362... Training loss: 1.5891... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16363... Training loss: 1.7178... 0.0543 sec/batch
Epoch: 5/20... Training Step: 16364... Training loss: 1.6555... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16365... Training loss: 1.7001... 0.0548 sec/batch
Epoch: 5/20... Training Step: 16366... Training loss: 1.7632... 0.0563 sec/batch
Epoch: 5/20... Training Step: 16367... Training loss: 1.7816... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16368... Training loss: 1.6848... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16369... Training loss: 1.7187... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16370... Training loss: 1.7551... 0.0565 sec/batch
Epoch: 5/20... Training Step: 16371... Training loss: 1.6550... 0.0505 sec/batch
Epoch: 5/20... Training Step: 16372... Training loss: 1.7172... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16373... Training loss: 1.7608... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16374... Training loss: 1.7149... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16375... Training loss: 1.5394... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16376... Training loss: 1.6050... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16377... Training loss: 1.5089... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16378... Training loss: 1.6549... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16379... Training loss: 1.5538... 0.0624 sec/batch
Epoch: 5/20... Training Step: 16380... Training loss: 1.8341... 0.0536 sec/batch
Epoch: 5/20... Training Step: 16381... Training loss: 1.6327... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16382... Training loss: 1.7251... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16383... Training loss: 1.7061... 0.0517 sec/batch
Epoch: 5/20... Training Step: 16384... Training loss: 1.6774... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16385... Training loss: 1.5995... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16386... Training loss: 1.6280... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16387... Training loss: 1.7140... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16388... Training loss: 1.7787... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16389... Training loss: 1.7122... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16390... Training loss: 1.7106... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16391... Training loss: 1.7054... 0.0581 sec/batch
Epoch: 5/20... Training Step: 16392... Training loss: 1.8827... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16393... Training loss: 1.6228... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16394... Training loss: 1.5811... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16395... Training loss: 1.8220... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16396... Training loss: 1.7569... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16397... Training loss: 1.7642... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16398... Training loss: 1.6725... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16399... Training loss: 1.6367... 0.0533 sec/batch
Epoch: 5/20... Training Step: 16400... Training loss: 1.5729... 0.0568 sec/batch
Epoch: 5/20... Training Step: 16401... Training loss: 1.6874... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16402... Training loss: 1.7194... 0.0624 sec/batch
Epoch: 5/20... Training Step: 16403... Training loss: 1.6884... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16404... Training loss: 1.5928... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16405... Training loss: 1.7098... 0.0656 sec/batch
Epoch: 5/20... Training Step: 16406... Training loss: 1.6159... 0.0529 sec/batch
Epoch: 5/20... Training Step: 16407... Training loss: 1.6196... 0.0498 sec/batch
Epoch: 5/20... Training Step: 16408... Training loss: 1.6604... 0.0571 sec/batch
Epoch: 5/20... Training Step: 16409... Training loss: 1.7056... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16410... Training loss: 1.6961... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16411... Training loss: 1.7619... 0.0546 sec/batch
Epoch: 5/20... Training Step: 16412... Training loss: 1.6523... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16413... Training loss: 1.5737... 0.0476 sec/batch
Epoch: 5/20... Training Step: 16414... Training loss: 1.5971... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16415... Training loss: 1.6964... 0.0627 sec/batch
Epoch: 5/20... Training Step: 16416... Training loss: 1.6207... 0.0527 sec/batch
Epoch: 5/20... Training Step: 16417... Training loss: 1.5016... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16418... Training loss: 1.5316... 0.0731 sec/batch
Epoch: 5/20... Training Step: 16419... Training loss: 1.5205... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16420... Training loss: 1.5287... 0.0645 sec/batch
Epoch: 5/20... Training Step: 16421... Training loss: 1.5491... 0.0669 sec/batch
Epoch: 5/20... Training Step: 16422... Training loss: 1.6017... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16423... Training loss: 1.6769... 0.0653 sec/batch
Epoch: 5/20... Training Step: 16424... Training loss: 1.5453... 0.0649 sec/batch
Epoch: 5/20... Training Step: 16425... Training loss: 1.7304... 0.0684 sec/batch
Epoch: 5/20... Training Step: 16426... Training loss: 1.6299... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16427... Training loss: 1.6720... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16428... Training loss: 1.5605... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16429... Training loss: 1.6174... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16430... Training loss: 1.5173... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16431... Training loss: 1.5784... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16432... Training loss: 1.6935... 0.0642 sec/batch
Epoch: 5/20... Training Step: 16433... Training loss: 1.5528... 0.0685 sec/batch
Epoch: 5/20... Training Step: 16434... Training loss: 1.6065... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16435... Training loss: 1.6838... 0.0544 sec/batch
Epoch: 5/20... Training Step: 16436... Training loss: 1.6499... 0.0577 sec/batch
Epoch: 5/20... Training Step: 16437... Training loss: 1.6288... 0.0742 sec/batch
Epoch: 5/20... Training Step: 16438... Training loss: 1.5906... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16439... Training loss: 1.6998... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16440... Training loss: 1.6632... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16441... Training loss: 1.5830... 0.0774 sec/batch
Epoch: 5/20... Training Step: 16442... Training loss: 1.7271... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16443... Training loss: 1.6014... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16444... Training loss: 1.5796... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16445... Training loss: 1.5837... 0.0749 sec/batch
Epoch: 5/20... Training Step: 16446... Training loss: 1.6485... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16447... Training loss: 1.4905... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16448... Training loss: 1.7401... 0.0697 sec/batch
Epoch: 5/20... Training Step: 16449... Training loss: 1.6891... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16450... Training loss: 1.7208... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16451... Training loss: 1.6545... 0.0650 sec/batch
Epoch: 5/20... Training Step: 16452... Training loss: 1.6227... 0.0624 sec/batch
Epoch: 5/20... Training Step: 16453... Training loss: 1.4578... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16454... Training loss: 1.6016... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16455... Training loss: 1.5241... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16456... Training loss: 1.5616... 0.0609 sec/batch
Epoch: 5/20... Training Step: 16457... Training loss: 1.7147... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16458... Training loss: 1.7281... 0.0685 sec/batch
Epoch: 5/20... Training Step: 16459... Training loss: 1.6822... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16460... Training loss: 1.8427... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16461... Training loss: 1.7631... 0.0659 sec/batch
Epoch: 5/20... Training Step: 16462... Training loss: 1.5889... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16463... Training loss: 1.6762... 0.0547 sec/batch
Epoch: 5/20... Training Step: 16464... Training loss: 1.7457... 0.0655 sec/batch
Epoch: 5/20... Training Step: 16465... Training loss: 1.5475... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16466... Training loss: 1.7306... 0.0653 sec/batch
Epoch: 5/20... Training Step: 16467... Training loss: 1.5072... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16468... Training loss: 1.6296... 0.0565 sec/batch
Epoch: 5/20... Training Step: 16469... Training loss: 1.9634... 0.0702 sec/batch
Epoch: 5/20... Training Step: 16470... Training loss: 1.8097... 0.0513 sec/batch
Epoch: 5/20... Training Step: 16471... Training loss: 1.7671... 0.0577 sec/batch
Epoch: 5/20... Training Step: 16472... Training loss: 1.6226... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16473... Training loss: 1.5687... 0.0584 sec/batch
Epoch: 5/20... Training Step: 16474... Training loss: 1.8345... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16475... Training loss: 1.8952... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16476... Training loss: 1.7708... 0.0564 sec/batch
Epoch: 5/20... Training Step: 16477... Training loss: 1.5066... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16478... Training loss: 1.7298... 0.0606 sec/batch
Epoch: 5/20... Training Step: 16479... Training loss: 1.6459... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16480... Training loss: 1.5440... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16481... Training loss: 1.8927... 0.0749 sec/batch
Epoch: 5/20... Training Step: 16482... Training loss: 1.5877... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16483... Training loss: 1.6283... 0.0886 sec/batch
Epoch: 5/20... Training Step: 16484... Training loss: 1.5832... 0.0978 sec/batch
Epoch: 5/20... Training Step: 16485... Training loss: 1.6324... 0.0899 sec/batch
Epoch: 5/20... Training Step: 16486... Training loss: 1.6429... 0.0876 sec/batch
Epoch: 5/20... Training Step: 16487... Training loss: 1.5595... 0.0969 sec/batch
Epoch: 5/20... Training Step: 16488... Training loss: 1.7186... 0.0822 sec/batch
Epoch: 5/20... Training Step: 16489... Training loss: 1.7239... 0.0853 sec/batch
Epoch: 5/20... Training Step: 16490... Training loss: 1.6833... 0.1021 sec/batch
Epoch: 5/20... Training Step: 16491... Training loss: 1.7186... 0.0982 sec/batch
Epoch: 5/20... Training Step: 16492... Training loss: 1.6462... 0.0840 sec/batch
Epoch: 5/20... Training Step: 16493... Training loss: 1.6707... 0.0910 sec/batch
Epoch: 5/20... Training Step: 16494... Training loss: 1.7142... 0.0779 sec/batch
Epoch: 5/20... Training Step: 16495... Training loss: 1.5836... 0.0943 sec/batch
Epoch: 5/20... Training Step: 16496... Training loss: 1.6992... 0.0944 sec/batch
Epoch: 5/20... Training Step: 16497... Training loss: 1.5694... 0.0818 sec/batch
Epoch: 5/20... Training Step: 16498... Training loss: 1.5572... 0.0767 sec/batch
Epoch: 5/20... Training Step: 16499... Training loss: 1.5583... 0.1016 sec/batch
Epoch: 5/20... Training Step: 16500... Training loss: 1.5139... 0.0882 sec/batch
Epoch: 5/20... Training Step: 16501... Training loss: 1.5541... 0.0786 sec/batch
Epoch: 5/20... Training Step: 16502... Training loss: 1.6136... 0.1312 sec/batch
Epoch: 5/20... Training Step: 16503... Training loss: 1.6953... 0.0993 sec/batch
Epoch: 5/20... Training Step: 16504... Training loss: 1.6794... 0.0910 sec/batch
Epoch: 5/20... Training Step: 16505... Training loss: 1.5164... 0.0985 sec/batch
Epoch: 5/20... Training Step: 16506... Training loss: 1.5954... 0.0995 sec/batch
Epoch: 5/20... Training Step: 16507... Training loss: 1.5932... 0.0859 sec/batch
Epoch: 5/20... Training Step: 16508... Training loss: 1.5361... 0.0872 sec/batch
Epoch: 5/20... Training Step: 16509... Training loss: 1.6274... 0.0872 sec/batch
Epoch: 5/20... Training Step: 16510... Training loss: 1.7741... 0.0863 sec/batch
Epoch: 5/20... Training Step: 16511... Training loss: 1.6298... 0.0913 sec/batch
Epoch: 5/20... Training Step: 16512... Training loss: 1.5754... 0.0856 sec/batch
Epoch: 5/20... Training Step: 16513... Training loss: 1.5738... 0.0806 sec/batch
Epoch: 5/20... Training Step: 16514... Training loss: 1.6484... 0.0915 sec/batch
Epoch: 5/20... Training Step: 16515... Training loss: 1.7858... 0.0902 sec/batch
Epoch: 5/20... Training Step: 16516... Training loss: 1.7601... 0.0906 sec/batch
Epoch: 5/20... Training Step: 16517... Training loss: 1.6787... 0.0983 sec/batch
Epoch: 5/20... Training Step: 16518... Training loss: 1.6499... 0.0847 sec/batch
Epoch: 5/20... Training Step: 16519... Training loss: 1.6147... 0.0886 sec/batch
Epoch: 5/20... Training Step: 16520... Training loss: 1.5602... 0.0988 sec/batch
Epoch: 5/20... Training Step: 16521... Training loss: 1.6589... 0.0915 sec/batch
Epoch: 5/20... Training Step: 16522... Training loss: 1.6514... 0.0804 sec/batch
Epoch: 5/20... Training Step: 16523... Training loss: 1.5700... 0.0922 sec/batch
Epoch: 5/20... Training Step: 16524... Training loss: 1.5712... 0.0919 sec/batch
Epoch: 5/20... Training Step: 16525... Training loss: 1.6488... 0.0872 sec/batch
Epoch: 5/20... Training Step: 16526... Training loss: 1.7622... 0.0901 sec/batch
Epoch: 5/20... Training Step: 16527... Training loss: 1.6779... 0.0657 sec/batch
Epoch: 5/20... Training Step: 16528... Training loss: 1.7713... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16529... Training loss: 1.6927... 0.0702 sec/batch
Epoch: 5/20... Training Step: 16530... Training loss: 1.5626... 0.0723 sec/batch
Epoch: 5/20... Training Step: 16531... Training loss: 1.7620... 0.0580 sec/batch
Epoch: 5/20... Training Step: 16532... Training loss: 1.7687... 0.0718 sec/batch
Epoch: 5/20... Training Step: 16533... Training loss: 1.8284... 0.0682 sec/batch
Epoch: 5/20... Training Step: 16534... Training loss: 1.6082... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16535... Training loss: 1.6594... 0.0637 sec/batch
Epoch: 5/20... Training Step: 16536... Training loss: 1.7024... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16537... Training loss: 1.6165... 0.0779 sec/batch
Epoch: 5/20... Training Step: 16538... Training loss: 1.7116... 0.0685 sec/batch
Epoch: 5/20... Training Step: 16539... Training loss: 1.6929... 0.0664 sec/batch
Epoch: 5/20... Training Step: 16540... Training loss: 1.7082... 0.0651 sec/batch
Epoch: 5/20... Training Step: 16541... Training loss: 1.7072... 0.0653 sec/batch
Epoch: 5/20... Training Step: 16542... Training loss: 1.6114... 0.0670 sec/batch
Epoch: 5/20... Training Step: 16543... Training loss: 1.7310... 0.0662 sec/batch
Epoch: 5/20... Training Step: 16544... Training loss: 1.7511... 0.0683 sec/batch
Epoch: 5/20... Training Step: 16545... Training loss: 1.7431... 0.0677 sec/batch
Epoch: 5/20... Training Step: 16546... Training loss: 1.7269... 0.0633 sec/batch
Epoch: 5/20... Training Step: 16547... Training loss: 1.7043... 0.0649 sec/batch
Epoch: 5/20... Training Step: 16548... Training loss: 1.5700... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16549... Training loss: 1.5563... 0.0684 sec/batch
Epoch: 5/20... Training Step: 16550... Training loss: 1.6474... 0.0681 sec/batch
Epoch: 5/20... Training Step: 16551... Training loss: 1.6298... 0.0717 sec/batch
Epoch: 5/20... Training Step: 16552... Training loss: 1.6434... 0.0659 sec/batch
Epoch: 5/20... Training Step: 16553... Training loss: 1.7490... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16554... Training loss: 1.7329... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16555... Training loss: 1.6665... 0.0684 sec/batch
Epoch: 5/20... Training Step: 16556... Training loss: 1.6391... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16557... Training loss: 1.8835... 0.0657 sec/batch
Epoch: 5/20... Training Step: 16558... Training loss: 1.6733... 0.0679 sec/batch
Epoch: 5/20... Training Step: 16559... Training loss: 1.6952... 0.0660 sec/batch
Epoch: 5/20... Training Step: 16560... Training loss: 1.7756... 0.0642 sec/batch
Epoch: 5/20... Training Step: 16561... Training loss: 1.6758... 0.0590 sec/batch
Epoch: 5/20... Training Step: 16562... Training loss: 1.6649... 0.0682 sec/batch
Epoch: 5/20... Training Step: 16563... Training loss: 1.6673... 0.0702 sec/batch
Epoch: 5/20... Training Step: 16564... Training loss: 1.6271... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16565... Training loss: 1.6185... 0.0657 sec/batch
Epoch: 5/20... Training Step: 16566... Training loss: 1.7067... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16567... Training loss: 1.7810... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16568... Training loss: 1.5647... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16569... Training loss: 1.5561... 0.0672 sec/batch
Epoch: 5/20... Training Step: 16570... Training loss: 1.6352... 0.0687 sec/batch
Epoch: 5/20... Training Step: 16571... Training loss: 1.7625... 0.0811 sec/batch
Epoch: 5/20... Training Step: 16572... Training loss: 1.7145... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16573... Training loss: 1.6772... 0.0662 sec/batch
Epoch: 5/20... Training Step: 16574... Training loss: 1.7035... 0.0660 sec/batch
Epoch: 5/20... Training Step: 16575... Training loss: 1.6242... 0.0751 sec/batch
Epoch: 5/20... Training Step: 16576... Training loss: 1.7828... 0.0655 sec/batch
Epoch: 5/20... Training Step: 16577... Training loss: 1.7108... 0.0721 sec/batch
Epoch: 5/20... Training Step: 16578... Training loss: 1.6757... 0.0670 sec/batch
Epoch: 5/20... Training Step: 16579... Training loss: 1.7547... 0.0682 sec/batch
Epoch: 5/20... Training Step: 16580... Training loss: 1.5985... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16581... Training loss: 1.6710... 0.0689 sec/batch
Epoch: 5/20... Training Step: 16582... Training loss: 1.6664... 0.0658 sec/batch
Epoch: 5/20... Training Step: 16583... Training loss: 1.6502... 0.0664 sec/batch
Epoch: 5/20... Training Step: 16584... Training loss: 1.7159... 0.0760 sec/batch
Epoch: 5/20... Training Step: 16585... Training loss: 1.6399... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16586... Training loss: 1.6853... 0.0721 sec/batch
Epoch: 5/20... Training Step: 16587... Training loss: 1.7487... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16588... Training loss: 1.5999... 0.0548 sec/batch
Epoch: 5/20... Training Step: 16589... Training loss: 1.6584... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16590... Training loss: 1.7305... 0.0650 sec/batch
Epoch: 5/20... Training Step: 16591... Training loss: 1.6014... 0.0674 sec/batch
Epoch: 5/20... Training Step: 16592... Training loss: 1.7101... 0.0637 sec/batch
Epoch: 5/20... Training Step: 16593... Training loss: 1.8079... 0.0689 sec/batch
Epoch: 5/20... Training Step: 16594... Training loss: 1.5780... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16595... Training loss: 1.6241... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16596... Training loss: 1.5579... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16597... Training loss: 1.6892... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16598... Training loss: 1.5691... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16599... Training loss: 1.6163... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16600... Training loss: 1.7387... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16601... Training loss: 1.6958... 0.0668 sec/batch
Epoch: 5/20... Training Step: 16602... Training loss: 1.4836... 0.0585 sec/batch
Epoch: 5/20... Training Step: 16603... Training loss: 1.6022... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16604... Training loss: 1.7079... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16605... Training loss: 1.6374... 0.0691 sec/batch
Epoch: 5/20... Training Step: 16606... Training loss: 1.5785... 0.0666 sec/batch
Epoch: 5/20... Training Step: 16607... Training loss: 1.6429... 0.0614 sec/batch
Epoch: 5/20... Training Step: 16608... Training loss: 1.6362... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16609... Training loss: 1.6004... 0.0657 sec/batch
Epoch: 5/20... Training Step: 16610... Training loss: 1.5822... 0.0608 sec/batch
Epoch: 5/20... Training Step: 16611... Training loss: 1.6833... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16612... Training loss: 1.6163... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16613... Training loss: 1.6705... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16614... Training loss: 1.5391... 0.0580 sec/batch
Epoch: 5/20... Training Step: 16615... Training loss: 1.5301... 0.0633 sec/batch
Epoch: 5/20... Training Step: 16616... Training loss: 1.7038... 0.0670 sec/batch
Epoch: 5/20... Training Step: 16617... Training loss: 1.6467... 0.0709 sec/batch
Epoch: 5/20... Training Step: 16618... Training loss: 1.6718... 0.0679 sec/batch
Epoch: 5/20... Training Step: 16619... Training loss: 1.6848... 0.0609 sec/batch
Epoch: 5/20... Training Step: 16620... Training loss: 1.5902... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16621... Training loss: 1.6308... 0.0668 sec/batch
Epoch: 5/20... Training Step: 16622... Training loss: 1.5033... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16623... Training loss: 1.5227... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16624... Training loss: 1.6455... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16625... Training loss: 1.6666... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16626... Training loss: 1.7022... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16627... Training loss: 1.7300... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16628... Training loss: 1.6532... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16629... Training loss: 1.5525... 0.0584 sec/batch
Epoch: 5/20... Training Step: 16630... Training loss: 1.6696... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16631... Training loss: 1.5657... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16632... Training loss: 1.9027... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16633... Training loss: 1.7051... 0.0681 sec/batch
Epoch: 5/20... Training Step: 16634... Training loss: 1.5224... 0.0692 sec/batch
Epoch: 5/20... Training Step: 16635... Training loss: 1.6141... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16636... Training loss: 1.5755... 0.0687 sec/batch
Epoch: 5/20... Training Step: 16637... Training loss: 1.6318... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16638... Training loss: 1.6015... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16639... Training loss: 1.6592... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16640... Training loss: 1.5684... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16641... Training loss: 1.4927... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16642... Training loss: 1.7408... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16643... Training loss: 1.6849... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16644... Training loss: 1.6714... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16645... Training loss: 1.6424... 0.0692 sec/batch
Epoch: 5/20... Training Step: 16646... Training loss: 1.7172... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16647... Training loss: 1.7127... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16648... Training loss: 1.6845... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16649... Training loss: 1.7408... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16650... Training loss: 1.6548... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16651... Training loss: 1.6233... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16652... Training loss: 1.6302... 0.0578 sec/batch
Epoch: 5/20... Training Step: 16653... Training loss: 1.6617... 0.0665 sec/batch
Epoch: 5/20... Training Step: 16654... Training loss: 1.7692... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16655... Training loss: 1.6606... 0.0516 sec/batch
Epoch: 5/20... Training Step: 16656... Training loss: 1.7445... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16657... Training loss: 1.7694... 0.0573 sec/batch
Epoch: 5/20... Training Step: 16658... Training loss: 1.7925... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16659... Training loss: 1.6503... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16660... Training loss: 1.7389... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16661... Training loss: 1.6544... 0.0669 sec/batch
Epoch: 5/20... Training Step: 16662... Training loss: 1.6707... 0.0530 sec/batch
Epoch: 5/20... Training Step: 16663... Training loss: 1.6165... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16664... Training loss: 1.5578... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16665... Training loss: 1.6439... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16666... Training loss: 1.7274... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16667... Training loss: 1.6352... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16668... Training loss: 1.4467... 0.0512 sec/batch
Epoch: 5/20... Training Step: 16669... Training loss: 1.6904... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16670... Training loss: 1.6361... 0.0558 sec/batch
Epoch: 5/20... Training Step: 16671... Training loss: 1.6697... 0.0515 sec/batch
Epoch: 5/20... Training Step: 16672... Training loss: 1.4942... 0.0575 sec/batch
Epoch: 5/20... Training Step: 16673... Training loss: 1.7450... 0.0551 sec/batch
Epoch: 5/20... Training Step: 16674... Training loss: 1.6070... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16675... Training loss: 1.5641... 0.0582 sec/batch
Epoch: 5/20... Training Step: 16676... Training loss: 1.5569... 0.0540 sec/batch
Epoch: 5/20... Training Step: 16677... Training loss: 1.7063... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16678... Training loss: 1.5103... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16679... Training loss: 1.6434... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16680... Training loss: 1.6601... 0.0648 sec/batch
Epoch: 5/20... Training Step: 16681... Training loss: 1.6016... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16682... Training loss: 1.5357... 0.0571 sec/batch
Epoch: 5/20... Training Step: 16683... Training loss: 1.6872... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16684... Training loss: 1.8025... 0.0597 sec/batch
Epoch: 5/20... Training Step: 16685... Training loss: 1.7865... 0.0553 sec/batch
Epoch: 5/20... Training Step: 16686... Training loss: 1.7087... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16687... Training loss: 1.7162... 0.0578 sec/batch
Epoch: 5/20... Training Step: 16688... Training loss: 1.6198... 0.0519 sec/batch
Epoch: 5/20... Training Step: 16689... Training loss: 1.5801... 0.0531 sec/batch
Epoch: 5/20... Training Step: 16690... Training loss: 1.5471... 0.0493 sec/batch
Epoch: 5/20... Training Step: 16691... Training loss: 1.5908... 0.0584 sec/batch
Epoch: 5/20... Training Step: 16692... Training loss: 1.6304... 0.0575 sec/batch
Epoch: 5/20... Training Step: 16693... Training loss: 1.5870... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16694... Training loss: 1.6325... 0.0611 sec/batch
Epoch: 5/20... Training Step: 16695... Training loss: 1.6379... 0.0608 sec/batch
Epoch: 5/20... Training Step: 16696... Training loss: 1.6383... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16697... Training loss: 1.5284... 0.0653 sec/batch
Epoch: 5/20... Training Step: 16698... Training loss: 1.5994... 0.0493 sec/batch
Epoch: 5/20... Training Step: 16699... Training loss: 1.5804... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16700... Training loss: 1.6939... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16701... Training loss: 1.7255... 0.0568 sec/batch
Epoch: 5/20... Training Step: 16702... Training loss: 1.7337... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16703... Training loss: 1.6213... 0.0584 sec/batch
Epoch: 5/20... Training Step: 16704... Training loss: 1.5582... 0.0575 sec/batch
Epoch: 5/20... Training Step: 16705... Training loss: 1.5108... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16706... Training loss: 1.6465... 0.0586 sec/batch
Epoch: 5/20... Training Step: 16707... Training loss: 1.6573... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16708... Training loss: 1.6938... 0.0564 sec/batch
Epoch: 5/20... Training Step: 16709... Training loss: 1.6197... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16710... Training loss: 1.6314... 0.0634 sec/batch
Epoch: 5/20... Training Step: 16711... Training loss: 1.7533... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16712... Training loss: 1.6076... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16713... Training loss: 1.5524... 0.0690 sec/batch
Epoch: 5/20... Training Step: 16714... Training loss: 1.6088... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16715... Training loss: 1.7264... 0.0585 sec/batch
Epoch: 5/20... Training Step: 16716... Training loss: 1.4797... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16717... Training loss: 1.6344... 0.0686 sec/batch
Epoch: 5/20... Training Step: 16718... Training loss: 1.5294... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16719... Training loss: 1.6692... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16720... Training loss: 1.5843... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16721... Training loss: 1.6530... 0.0697 sec/batch
Epoch: 5/20... Training Step: 16722... Training loss: 1.6258... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16723... Training loss: 1.6607... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16724... Training loss: 1.7293... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16725... Training loss: 1.6414... 0.0665 sec/batch
Epoch: 5/20... Training Step: 16726... Training loss: 1.6503... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16727... Training loss: 1.6759... 0.0643 sec/batch
Epoch: 5/20... Training Step: 16728... Training loss: 1.6338... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16729... Training loss: 1.6590... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16730... Training loss: 1.7837... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16731... Training loss: 1.7582... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16732... Training loss: 1.4989... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16733... Training loss: 1.6651... 0.0704 sec/batch
Epoch: 5/20... Training Step: 16734... Training loss: 1.5200... 0.0672 sec/batch
Epoch: 5/20... Training Step: 16735... Training loss: 1.7625... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16736... Training loss: 1.7598... 0.0543 sec/batch
Epoch: 5/20... Training Step: 16737... Training loss: 1.7696... 0.0678 sec/batch
Epoch: 5/20... Training Step: 16738... Training loss: 1.6358... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16739... Training loss: 1.6455... 0.0596 sec/batch
Epoch: 5/20... Training Step: 16740... Training loss: 1.8196... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16741... Training loss: 1.6474... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16742... Training loss: 1.7158... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16743... Training loss: 1.7238... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16744... Training loss: 1.6601... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16745... Training loss: 1.7541... 0.0729 sec/batch
Epoch: 5/20... Training Step: 16746... Training loss: 1.8753... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16747... Training loss: 1.7591... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16748... Training loss: 1.7530... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16749... Training loss: 1.7160... 0.0743 sec/batch
Epoch: 5/20... Training Step: 16750... Training loss: 1.7036... 0.0751 sec/batch
Epoch: 5/20... Training Step: 16751... Training loss: 1.5606... 0.0625 sec/batch
Epoch: 5/20... Training Step: 16752... Training loss: 1.5621... 0.0646 sec/batch
Epoch: 5/20... Training Step: 16753... Training loss: 1.6490... 0.0676 sec/batch
Epoch: 5/20... Training Step: 16754... Training loss: 1.6612... 0.0534 sec/batch
Epoch: 5/20... Training Step: 16755... Training loss: 1.7704... 0.0652 sec/batch
Epoch: 5/20... Training Step: 16756... Training loss: 1.6520... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16757... Training loss: 1.6068... 0.0573 sec/batch
Epoch: 5/20... Training Step: 16758... Training loss: 1.7921... 0.0672 sec/batch
Epoch: 5/20... Training Step: 16759... Training loss: 1.6008... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16760... Training loss: 1.5317... 0.0747 sec/batch
Epoch: 5/20... Training Step: 16761... Training loss: 1.9503... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16762... Training loss: 1.8938... 0.0629 sec/batch
Epoch: 5/20... Training Step: 16763... Training loss: 1.6677... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16764... Training loss: 1.6926... 0.0698 sec/batch
Epoch: 5/20... Training Step: 16765... Training loss: 1.5754... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16766... Training loss: 1.6301... 0.0664 sec/batch
Epoch: 5/20... Training Step: 16767... Training loss: 1.7096... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16768... Training loss: 1.5380... 0.0665 sec/batch
Epoch: 5/20... Training Step: 16769... Training loss: 1.6179... 0.0673 sec/batch
Epoch: 5/20... Training Step: 16770... Training loss: 1.6792... 0.0672 sec/batch
Epoch: 5/20... Training Step: 16771... Training loss: 1.6358... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16772... Training loss: 1.7186... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16773... Training loss: 1.7876... 0.0637 sec/batch
Epoch: 5/20... Training Step: 16774... Training loss: 1.7480... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16775... Training loss: 1.6993... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16776... Training loss: 1.6027... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16777... Training loss: 1.8082... 0.0626 sec/batch
Epoch: 5/20... Training Step: 16778... Training loss: 1.5437... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16779... Training loss: 1.6474... 0.0638 sec/batch
Epoch: 5/20... Training Step: 16780... Training loss: 1.6594... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16781... Training loss: 1.5372... 0.0585 sec/batch
Epoch: 5/20... Training Step: 16782... Training loss: 1.5604... 0.0527 sec/batch
Epoch: 5/20... Training Step: 16783... Training loss: 1.5405... 0.0547 sec/batch
Epoch: 5/20... Training Step: 16784... Training loss: 1.6072... 0.0572 sec/batch
Epoch: 5/20... Training Step: 16785... Training loss: 1.5964... 0.0490 sec/batch
Epoch: 5/20... Training Step: 16786... Training loss: 1.7520... 0.0608 sec/batch
Epoch: 5/20... Training Step: 16787... Training loss: 1.8345... 0.0549 sec/batch
Epoch: 5/20... Training Step: 16788... Training loss: 1.7300... 0.0601 sec/batch
Epoch: 5/20... Training Step: 16789... Training loss: 1.8603... 0.0522 sec/batch
Epoch: 5/20... Training Step: 16790... Training loss: 1.7408... 0.0529 sec/batch
Epoch: 5/20... Training Step: 16791... Training loss: 1.6227... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16792... Training loss: 1.7449... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16793... Training loss: 1.7481... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16794... Training loss: 1.6513... 0.0557 sec/batch
Epoch: 5/20... Training Step: 16795... Training loss: 1.6272... 0.0709 sec/batch
Epoch: 5/20... Training Step: 16796... Training loss: 1.6720... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16797... Training loss: 1.7391... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16798... Training loss: 1.5171... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16799... Training loss: 1.5706... 0.0700 sec/batch
Epoch: 5/20... Training Step: 16800... Training loss: 1.5594... 0.0610 sec/batch
Epoch: 5/20... Training Step: 16801... Training loss: 1.5674... 0.0503 sec/batch
Epoch: 5/20... Training Step: 16802... Training loss: 1.7117... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16803... Training loss: 1.7615... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16804... Training loss: 1.6324... 0.0596 sec/batch
Epoch: 5/20... Training Step: 16805... Training loss: 1.7094... 0.0578 sec/batch
Epoch: 5/20... Training Step: 16806... Training loss: 1.7147... 0.0691 sec/batch
Epoch: 5/20... Training Step: 16807... Training loss: 1.6319... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16808... Training loss: 1.5598... 0.0590 sec/batch
Epoch: 5/20... Training Step: 16809... Training loss: 1.5749... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16810... Training loss: 1.6820... 0.0750 sec/batch
Epoch: 5/20... Training Step: 16811... Training loss: 1.7021... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16812... Training loss: 1.5611... 0.0580 sec/batch
Epoch: 5/20... Training Step: 16813... Training loss: 1.6392... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16814... Training loss: 1.6647... 0.0669 sec/batch
Epoch: 5/20... Training Step: 16815... Training loss: 1.6072... 0.0615 sec/batch
Epoch: 5/20... Training Step: 16816... Training loss: 1.6112... 0.0788 sec/batch
Epoch: 5/20... Training Step: 16817... Training loss: 1.5384... 0.0769 sec/batch
Epoch: 5/20... Training Step: 16818... Training loss: 1.5972... 0.0893 sec/batch
Epoch: 5/20... Training Step: 16819... Training loss: 1.6158... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16820... Training loss: 1.6004... 0.0780 sec/batch
Epoch: 5/20... Training Step: 16821... Training loss: 1.5971... 0.0968 sec/batch
Epoch: 5/20... Training Step: 16822... Training loss: 1.5224... 0.0683 sec/batch
Epoch: 5/20... Training Step: 16823... Training loss: 1.5746... 0.0740 sec/batch
Epoch: 5/20... Training Step: 16824... Training loss: 1.6770... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16825... Training loss: 1.6472... 0.0503 sec/batch
Epoch: 5/20... Training Step: 16826... Training loss: 1.6312... 0.0553 sec/batch
Epoch: 5/20... Training Step: 16827... Training loss: 1.5714... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16828... Training loss: 1.6684... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16829... Training loss: 1.6709... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16830... Training loss: 1.7100... 0.0609 sec/batch
Epoch: 5/20... Training Step: 16831... Training loss: 1.6735... 0.0562 sec/batch
Epoch: 5/20... Training Step: 16832... Training loss: 1.7027... 0.0533 sec/batch
Epoch: 5/20... Training Step: 16833... Training loss: 1.6994... 0.0600 sec/batch
Epoch: 5/20... Training Step: 16834... Training loss: 1.6540... 0.0581 sec/batch
Epoch: 5/20... Training Step: 16835... Training loss: 1.6252... 0.0757 sec/batch
Epoch: 5/20... Training Step: 16836... Training loss: 1.5455... 0.0588 sec/batch
Epoch: 5/20... Training Step: 16837... Training loss: 1.7647... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16838... Training loss: 1.7498... 0.0606 sec/batch
Epoch: 5/20... Training Step: 16839... Training loss: 1.5493... 0.0684 sec/batch
Epoch: 5/20... Training Step: 16840... Training loss: 1.7455... 0.0593 sec/batch
Epoch: 5/20... Training Step: 16841... Training loss: 1.5723... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16842... Training loss: 1.5961... 0.0585 sec/batch
Epoch: 5/20... Training Step: 16843... Training loss: 1.5074... 0.0523 sec/batch
Epoch: 5/20... Training Step: 16844... Training loss: 1.5689... 0.0648 sec/batch
Epoch: 5/20... Training Step: 16845... Training loss: 1.6986... 0.0639 sec/batch
Epoch: 5/20... Training Step: 16846... Training loss: 1.6853... 0.0577 sec/batch
Epoch: 5/20... Training Step: 16847... Training loss: 1.5872... 0.0654 sec/batch
Epoch: 5/20... Training Step: 16848... Training loss: 1.6894... 0.0596 sec/batch
Epoch: 5/20... Training Step: 16849... Training loss: 1.6806... 0.0699 sec/batch
Epoch: 5/20... Training Step: 16850... Training loss: 1.5934... 0.0647 sec/batch
Epoch: 5/20... Training Step: 16851... Training loss: 1.5337... 0.0681 sec/batch
Epoch: 5/20... Training Step: 16852... Training loss: 1.6636... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16853... Training loss: 1.6521... 0.0535 sec/batch
Epoch: 5/20... Training Step: 16854... Training loss: 1.5766... 0.0558 sec/batch
Epoch: 5/20... Training Step: 16855... Training loss: 1.5714... 0.0700 sec/batch
Epoch: 5/20... Training Step: 16856... Training loss: 1.7318... 0.0579 sec/batch
Epoch: 5/20... Training Step: 16857... Training loss: 1.7993... 0.0674 sec/batch
Epoch: 5/20... Training Step: 16858... Training loss: 1.5874... 0.0683 sec/batch
Epoch: 5/20... Training Step: 16859... Training loss: 1.7354... 0.0839 sec/batch
Epoch: 5/20... Training Step: 16860... Training loss: 1.5445... 0.0698 sec/batch
Epoch: 5/20... Training Step: 16861... Training loss: 1.7523... 0.0703 sec/batch
Epoch: 5/20... Training Step: 16862... Training loss: 1.5118... 0.0706 sec/batch
Epoch: 5/20... Training Step: 16863... Training loss: 1.7871... 0.1127 sec/batch
Epoch: 5/20... Training Step: 16864... Training loss: 1.6366... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16865... Training loss: 1.6045... 0.0757 sec/batch
Epoch: 5/20... Training Step: 16866... Training loss: 1.6096... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16867... Training loss: 1.5676... 0.0576 sec/batch
Epoch: 5/20... Training Step: 16868... Training loss: 1.5701... 0.0588 sec/batch
Epoch: 5/20... Training Step: 16869... Training loss: 1.5905... 0.0656 sec/batch
Epoch: 5/20... Training Step: 16870... Training loss: 1.5322... 0.0671 sec/batch
Epoch: 5/20... Training Step: 16871... Training loss: 1.5055... 0.0646 sec/batch
Epoch: 5/20... Training Step: 16872... Training loss: 1.6919... 0.0605 sec/batch
Epoch: 5/20... Training Step: 16873... Training loss: 1.7130... 0.0662 sec/batch
Epoch: 5/20... Training Step: 16874... Training loss: 1.7090... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16875... Training loss: 1.7193... 0.0578 sec/batch
Epoch: 5/20... Training Step: 16876... Training loss: 1.7478... 0.0656 sec/batch
Epoch: 5/20... Training Step: 16877... Training loss: 1.5952... 0.0696 sec/batch
Epoch: 5/20... Training Step: 16878... Training loss: 1.7648... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16879... Training loss: 1.5799... 0.0648 sec/batch
Epoch: 5/20... Training Step: 16880... Training loss: 1.6648... 0.0623 sec/batch
Epoch: 5/20... Training Step: 16881... Training loss: 1.6905... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16882... Training loss: 1.4956... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16883... Training loss: 1.5912... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16884... Training loss: 1.5340... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16885... Training loss: 1.5583... 0.0688 sec/batch
Epoch: 5/20... Training Step: 16886... Training loss: 1.5229... 0.0614 sec/batch
Epoch: 5/20... Training Step: 16887... Training loss: 1.6743... 0.0589 sec/batch
Epoch: 5/20... Training Step: 16888... Training loss: 1.7395... 0.0583 sec/batch
Epoch: 5/20... Training Step: 16889... Training loss: 1.6732... 0.0529 sec/batch
Epoch: 5/20... Training Step: 16890... Training loss: 1.7508... 0.0603 sec/batch
Epoch: 5/20... Training Step: 16891... Training loss: 1.6474... 0.0644 sec/batch
Epoch: 5/20... Training Step: 16892... Training loss: 1.6749... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16893... Training loss: 1.6846... 0.0711 sec/batch
Epoch: 5/20... Training Step: 16894... Training loss: 1.7294... 0.0612 sec/batch
Epoch: 5/20... Training Step: 16895... Training loss: 1.5301... 0.0662 sec/batch
Epoch: 5/20... Training Step: 16896... Training loss: 1.7847... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16897... Training loss: 1.7511... 0.0717 sec/batch
Epoch: 5/20... Training Step: 16898... Training loss: 1.7066... 0.0644 sec/batch
Epoch: 5/20... Training Step: 16899... Training loss: 1.6086... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16900... Training loss: 1.4849... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16901... Training loss: 1.5421... 0.0685 sec/batch
Epoch: 5/20... Training Step: 16902... Training loss: 1.6236... 0.0616 sec/batch
Epoch: 5/20... Training Step: 16903... Training loss: 1.5363... 0.0590 sec/batch
Epoch: 5/20... Training Step: 16904... Training loss: 1.6711... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16905... Training loss: 1.6318... 0.0729 sec/batch
Epoch: 5/20... Training Step: 16906... Training loss: 1.7027... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16907... Training loss: 1.7382... 0.0654 sec/batch
Epoch: 5/20... Training Step: 16908... Training loss: 1.6426... 0.0509 sec/batch
Epoch: 5/20... Training Step: 16909... Training loss: 1.6885... 0.0602 sec/batch
Epoch: 5/20... Training Step: 16910... Training loss: 1.5996... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16911... Training loss: 1.6005... 0.0640 sec/batch
Epoch: 5/20... Training Step: 16912... Training loss: 1.6003... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16913... Training loss: 1.5672... 0.0679 sec/batch
Epoch: 5/20... Training Step: 16914... Training loss: 1.6606... 0.0517 sec/batch
Epoch: 5/20... Training Step: 16915... Training loss: 1.6955... 0.0666 sec/batch
Epoch: 5/20... Training Step: 16916... Training loss: 1.7342... 0.0652 sec/batch
Epoch: 5/20... Training Step: 16917... Training loss: 1.5549... 0.0636 sec/batch
Epoch: 5/20... Training Step: 16918... Training loss: 1.5504... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16919... Training loss: 1.7161... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16920... Training loss: 1.6410... 0.0569 sec/batch
Epoch: 5/20... Training Step: 16921... Training loss: 1.6012... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16922... Training loss: 1.5765... 0.0570 sec/batch
Epoch: 5/20... Training Step: 16923... Training loss: 1.7433... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16924... Training loss: 1.7115... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16925... Training loss: 1.7008... 0.0559 sec/batch
Epoch: 5/20... Training Step: 16926... Training loss: 1.6876... 0.0557 sec/batch
Epoch: 5/20... Training Step: 16927... Training loss: 1.5931... 0.0630 sec/batch
Epoch: 5/20... Training Step: 16928... Training loss: 1.5792... 0.0529 sec/batch
Epoch: 5/20... Training Step: 16929... Training loss: 1.5194... 0.0596 sec/batch
Epoch: 5/20... Training Step: 16930... Training loss: 1.6522... 0.0588 sec/batch
Epoch: 5/20... Training Step: 16931... Training loss: 1.6190... 0.0509 sec/batch
Epoch: 5/20... Training Step: 16932... Training loss: 1.6938... 0.0604 sec/batch
Epoch: 5/20... Training Step: 16933... Training loss: 1.6284... 0.0558 sec/batch
Epoch: 5/20... Training Step: 16934... Training loss: 1.5975... 0.0608 sec/batch
Epoch: 5/20... Training Step: 16935... Training loss: 1.6141... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16936... Training loss: 1.5478... 0.0555 sec/batch
Epoch: 5/20... Training Step: 16937... Training loss: 1.5066... 0.0668 sec/batch
Epoch: 5/20... Training Step: 16938... Training loss: 1.6627... 0.0619 sec/batch
Epoch: 5/20... Training Step: 16939... Training loss: 1.6707... 0.0541 sec/batch
Epoch: 5/20... Training Step: 16940... Training loss: 1.6712... 0.0542 sec/batch
Epoch: 5/20... Training Step: 16941... Training loss: 1.6454... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16942... Training loss: 1.6084... 0.0632 sec/batch
Epoch: 5/20... Training Step: 16943... Training loss: 1.5930... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16944... Training loss: 1.6404... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16945... Training loss: 1.6092... 0.0594 sec/batch
Epoch: 5/20... Training Step: 16946... Training loss: 1.7059... 0.0631 sec/batch
Epoch: 5/20... Training Step: 16947... Training loss: 1.5432... 0.0626 sec/batch
Epoch: 5/20... Training Step: 16948... Training loss: 1.7172... 0.0509 sec/batch
Epoch: 5/20... Training Step: 16949... Training loss: 1.5816... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16950... Training loss: 1.5639... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16951... Training loss: 1.6575... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16952... Training loss: 1.5090... 0.0567 sec/batch
Epoch: 5/20... Training Step: 16953... Training loss: 1.4718... 0.0667 sec/batch
Epoch: 5/20... Training Step: 16954... Training loss: 1.5370... 0.0574 sec/batch
Epoch: 5/20... Training Step: 16955... Training loss: 1.6045... 0.0552 sec/batch
Epoch: 5/20... Training Step: 16956... Training loss: 1.5329... 0.0550 sec/batch
Epoch: 5/20... Training Step: 16957... Training loss: 1.6528... 0.0712 sec/batch
Epoch: 5/20... Training Step: 16958... Training loss: 1.7216... 0.0587 sec/batch
Epoch: 5/20... Training Step: 16959... Training loss: 1.5787... 0.0620 sec/batch
Epoch: 5/20... Training Step: 16960... Training loss: 1.4960... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16961... Training loss: 1.6187... 0.0675 sec/batch
Epoch: 5/20... Training Step: 16962... Training loss: 1.6129... 0.0540 sec/batch
Epoch: 5/20... Training Step: 16963... Training loss: 1.7319... 0.0552 sec/batch
Epoch: 5/20... Training Step: 16964... Training loss: 1.3558... 0.0532 sec/batch
Epoch: 5/20... Training Step: 16965... Training loss: 1.5597... 0.0641 sec/batch
Epoch: 5/20... Training Step: 16966... Training loss: 1.5782... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16967... Training loss: 1.5630... 0.0617 sec/batch
Epoch: 5/20... Training Step: 16968... Training loss: 1.5126... 0.0622 sec/batch
Epoch: 5/20... Training Step: 16969... Training loss: 1.5679... 0.0695 sec/batch
Epoch: 5/20... Training Step: 16970... Training loss: 1.8145... 0.0522 sec/batch
Epoch: 5/20... Training Step: 16971... Training loss: 1.6971... 0.0627 sec/batch
Epoch: 5/20... Training Step: 16972... Training loss: 1.7511... 0.0554 sec/batch
Epoch: 5/20... Training Step: 16973... Training loss: 1.6295... 0.0701 sec/batch
Epoch: 5/20... Training Step: 16974... Training loss: 1.6521... 0.0613 sec/batch
Epoch: 5/20... Training Step: 16975... Training loss: 1.5776... 0.0654 sec/batch
Epoch: 5/20... Training Step: 16976... Training loss: 1.6796... 0.0661 sec/batch
Epoch: 5/20... Training Step: 16977... Training loss: 1.5691... 0.0700 sec/batch
Epoch: 5/20... Training Step: 16978... Training loss: 1.4374... 0.0598 sec/batch
Epoch: 5/20... Training Step: 16979... Training loss: 1.6711... 0.0607 sec/batch
Epoch: 5/20... Training Step: 16980... Training loss: 1.6487... 0.0599 sec/batch
Epoch: 5/20... Training Step: 16981... Training loss: 1.5271... 0.0731 sec/batch
Epoch: 5/20... Training Step: 16982... Training loss: 1.5351... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16983... Training loss: 1.4962... 0.0606 sec/batch
Epoch: 5/20... Training Step: 16984... Training loss: 1.7070... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16985... Training loss: 1.6466... 0.0628 sec/batch
Epoch: 5/20... Training Step: 16986... Training loss: 1.5611... 0.0566 sec/batch
Epoch: 5/20... Training Step: 16987... Training loss: 1.6401... 0.0621 sec/batch
Epoch: 5/20... Training Step: 16988... Training loss: 1.6799... 0.0592 sec/batch
Epoch: 5/20... Training Step: 16989... Training loss: 1.4863... 0.0690 sec/batch
Epoch: 5/20... Training Step: 16990... Training loss: 1.5498... 0.0528 sec/batch
Epoch: 5/20... Training Step: 16991... Training loss: 1.6711... 0.0563 sec/batch
Epoch: 5/20... Training Step: 16992... Training loss: 1.6502... 0.0618 sec/batch
Epoch: 5/20... Training Step: 16993... Training loss: 1.4779... 0.0750 sec/batch
Epoch: 5/20... Training Step: 16994... Training loss: 1.7008... 0.0668 sec/batch
Epoch: 5/20... Training Step: 16995... Training loss: 1.6587... 0.0595 sec/batch
Epoch: 5/20... Training Step: 16996... Training loss: 1.7706... 0.0591 sec/batch
Epoch: 5/20... Training Step: 16997... Training loss: 1.6215... 0.0771 sec/batch
Epoch: 5/20... Training Step: 16998... Training loss: 1.5915... 0.0634 sec/batch
Epoch: 5/20... Training Step: 16999... Training loss: 1.4913... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17000... Training loss: 1.4223... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17001... Training loss: 1.7048... 0.0660 sec/batch
Epoch: 5/20... Training Step: 17002... Training loss: 1.6814... 0.0655 sec/batch
Epoch: 5/20... Training Step: 17003... Training loss: 1.6105... 0.0577 sec/batch
Epoch: 5/20... Training Step: 17004... Training loss: 1.5655... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17005... Training loss: 1.6615... 0.0675 sec/batch
Epoch: 5/20... Training Step: 17006... Training loss: 1.6161... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17007... Training loss: 1.6445... 0.0695 sec/batch
Epoch: 5/20... Training Step: 17008... Training loss: 1.4761... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17009... Training loss: 1.4613... 0.0862 sec/batch
Epoch: 5/20... Training Step: 17010... Training loss: 1.5367... 0.0576 sec/batch
Epoch: 5/20... Training Step: 17011... Training loss: 1.6217... 0.0595 sec/batch
Epoch: 5/20... Training Step: 17012... Training loss: 1.4217... 0.0722 sec/batch
Epoch: 5/20... Training Step: 17013... Training loss: 1.5194... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17014... Training loss: 1.7391... 0.0586 sec/batch
Epoch: 5/20... Training Step: 17015... Training loss: 1.5136... 0.0667 sec/batch
Epoch: 5/20... Training Step: 17016... Training loss: 1.6785... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17017... Training loss: 1.6338... 0.0619 sec/batch
Epoch: 5/20... Training Step: 17018... Training loss: 1.7631... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17019... Training loss: 1.6787... 0.0705 sec/batch
Epoch: 5/20... Training Step: 17020... Training loss: 1.7703... 0.0634 sec/batch
Epoch: 5/20... Training Step: 17021... Training loss: 1.5750... 0.0656 sec/batch
Epoch: 5/20... Training Step: 17022... Training loss: 1.5580... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17023... Training loss: 1.5380... 0.0759 sec/batch
Epoch: 5/20... Training Step: 17024... Training loss: 1.8853... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17025... Training loss: 1.8048... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17026... Training loss: 1.5354... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17027... Training loss: 1.7630... 0.0628 sec/batch
Epoch: 5/20... Training Step: 17028... Training loss: 1.7514... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17029... Training loss: 1.8454... 0.0713 sec/batch
Epoch: 5/20... Training Step: 17030... Training loss: 1.5984... 0.1011 sec/batch
Epoch: 5/20... Training Step: 17031... Training loss: 1.6610... 0.0870 sec/batch
Epoch: 5/20... Training Step: 17032... Training loss: 1.7294... 0.0825 sec/batch
Epoch: 5/20... Training Step: 17033... Training loss: 1.6295... 0.0937 sec/batch
Epoch: 5/20... Training Step: 17034... Training loss: 1.7129... 0.1011 sec/batch
Epoch: 5/20... Training Step: 17035... Training loss: 1.6557... 0.0830 sec/batch
Epoch: 5/20... Training Step: 17036... Training loss: 1.5776... 0.0952 sec/batch
Epoch: 5/20... Training Step: 17037... Training loss: 1.6452... 0.1012 sec/batch
Epoch: 5/20... Training Step: 17038... Training loss: 1.5915... 0.0897 sec/batch
Epoch: 5/20... Training Step: 17039... Training loss: 1.6787... 0.1239 sec/batch
Epoch: 5/20... Training Step: 17040... Training loss: 1.5354... 0.0785 sec/batch
Epoch: 5/20... Training Step: 17041... Training loss: 1.6281... 0.0971 sec/batch
Epoch: 5/20... Training Step: 17042... Training loss: 1.7236... 0.1126 sec/batch
Epoch: 5/20... Training Step: 17043... Training loss: 1.6765... 0.0973 sec/batch
Epoch: 5/20... Training Step: 17044... Training loss: 1.6867... 0.0931 sec/batch
Epoch: 5/20... Training Step: 17045... Training loss: 1.7275... 0.1101 sec/batch
Epoch: 5/20... Training Step: 17046... Training loss: 1.7413... 0.0890 sec/batch
Epoch: 5/20... Training Step: 17047... Training loss: 1.6208... 0.0891 sec/batch
Epoch: 5/20... Training Step: 17048... Training loss: 1.7399... 0.1188 sec/batch
Epoch: 5/20... Training Step: 17049... Training loss: 1.6617... 0.0906 sec/batch
Epoch: 5/20... Training Step: 17050... Training loss: 1.6230... 0.0962 sec/batch
Epoch: 5/20... Training Step: 17051... Training loss: 1.5067... 0.0946 sec/batch
Epoch: 5/20... Training Step: 17052... Training loss: 1.7066... 0.0831 sec/batch
Epoch: 5/20... Training Step: 17053... Training loss: 1.6566... 0.0848 sec/batch
Epoch: 5/20... Training Step: 17054... Training loss: 1.5168... 0.1195 sec/batch
Epoch: 5/20... Training Step: 17055... Training loss: 1.6545... 0.0895 sec/batch
Epoch: 5/20... Training Step: 17056... Training loss: 1.6321... 0.0957 sec/batch
Epoch: 5/20... Training Step: 17057... Training loss: 1.6258... 0.0830 sec/batch
Epoch: 5/20... Training Step: 17058... Training loss: 1.5012... 0.0870 sec/batch
Epoch: 5/20... Training Step: 17059... Training loss: 1.5026... 0.0876 sec/batch
Epoch: 5/20... Training Step: 17060... Training loss: 1.6467... 0.0937 sec/batch
Epoch: 5/20... Training Step: 17061... Training loss: 1.6753... 0.0876 sec/batch
Epoch: 5/20... Training Step: 17062... Training loss: 1.6655... 0.0914 sec/batch
Epoch: 5/20... Training Step: 17063... Training loss: 1.6895... 0.1070 sec/batch
Epoch: 5/20... Training Step: 17064... Training loss: 1.7802... 0.0768 sec/batch
Epoch: 5/20... Training Step: 17065... Training loss: 1.6024... 0.0884 sec/batch
Epoch: 5/20... Training Step: 17066... Training loss: 1.6179... 0.1163 sec/batch
Epoch: 5/20... Training Step: 17067... Training loss: 1.7909... 0.0788 sec/batch
Epoch: 5/20... Training Step: 17068... Training loss: 1.7487... 0.0919 sec/batch
Epoch: 5/20... Training Step: 17069... Training loss: 1.5313... 0.1060 sec/batch
Epoch: 5/20... Training Step: 17070... Training loss: 1.6211... 0.0838 sec/batch
Epoch: 5/20... Training Step: 17071... Training loss: 1.6093... 0.0834 sec/batch
Epoch: 5/20... Training Step: 17072... Training loss: 1.6880... 0.0930 sec/batch
Epoch: 5/20... Training Step: 17073... Training loss: 1.6772... 0.0729 sec/batch
Epoch: 5/20... Training Step: 17074... Training loss: 1.6052... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17075... Training loss: 1.7826... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17076... Training loss: 1.6794... 0.0771 sec/batch
Epoch: 5/20... Training Step: 17077... Training loss: 1.4962... 0.0703 sec/batch
Epoch: 5/20... Training Step: 17078... Training loss: 1.6141... 0.0662 sec/batch
Epoch: 5/20... Training Step: 17079... Training loss: 1.5980... 0.0683 sec/batch
Epoch: 5/20... Training Step: 17080... Training loss: 1.5398... 0.0696 sec/batch
Epoch: 5/20... Training Step: 17081... Training loss: 1.5883... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17082... Training loss: 1.6301... 0.0688 sec/batch
Epoch: 5/20... Training Step: 17083... Training loss: 1.6160... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17084... Training loss: 1.5301... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17085... Training loss: 1.6977... 0.0754 sec/batch
Epoch: 5/20... Training Step: 17086... Training loss: 1.5974... 0.0680 sec/batch
Epoch: 5/20... Training Step: 17087... Training loss: 1.5786... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17088... Training loss: 1.6998... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17089... Training loss: 1.6817... 0.0756 sec/batch
Epoch: 5/20... Training Step: 17090... Training loss: 1.6413... 0.0661 sec/batch
Epoch: 5/20... Training Step: 17091... Training loss: 1.5940... 0.0649 sec/batch
Epoch: 5/20... Training Step: 17092... Training loss: 1.8104... 0.0745 sec/batch
Epoch: 5/20... Training Step: 17093... Training loss: 1.8900... 0.0698 sec/batch
Epoch: 5/20... Training Step: 17094... Training loss: 1.5962... 0.0744 sec/batch
Epoch: 5/20... Training Step: 17095... Training loss: 1.5401... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17096... Training loss: 1.7840... 0.0660 sec/batch
Epoch: 5/20... Training Step: 17097... Training loss: 1.8476... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17098... Training loss: 1.7441... 0.0665 sec/batch
Epoch: 5/20... Training Step: 17099... Training loss: 1.6045... 0.0705 sec/batch
Epoch: 5/20... Training Step: 17100... Training loss: 1.6225... 0.0650 sec/batch
Epoch: 5/20... Training Step: 17101... Training loss: 1.7090... 0.0880 sec/batch
Epoch: 5/20... Training Step: 17102... Training loss: 1.7363... 0.0714 sec/batch
Epoch: 5/20... Training Step: 17103... Training loss: 1.5023... 0.0741 sec/batch
Epoch: 5/20... Training Step: 17104... Training loss: 1.5545... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17105... Training loss: 1.6724... 0.0702 sec/batch
Epoch: 5/20... Training Step: 17106... Training loss: 1.6055... 0.0656 sec/batch
Epoch: 5/20... Training Step: 17107... Training loss: 1.7878... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17108... Training loss: 1.7007... 0.0667 sec/batch
Epoch: 5/20... Training Step: 17109... Training loss: 1.5847... 0.0696 sec/batch
Epoch: 5/20... Training Step: 17110... Training loss: 1.7690... 0.0639 sec/batch
Epoch: 5/20... Training Step: 17111... Training loss: 1.5211... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17112... Training loss: 1.7506... 0.0684 sec/batch
Epoch: 5/20... Training Step: 17113... Training loss: 1.6268... 0.0693 sec/batch
Epoch: 5/20... Training Step: 17114... Training loss: 1.7419... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17115... Training loss: 1.5356... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17116... Training loss: 1.6144... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17117... Training loss: 1.7016... 0.0799 sec/batch
Epoch: 5/20... Training Step: 17118... Training loss: 1.4516... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17119... Training loss: 1.5069... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17120... Training loss: 1.6244... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17121... Training loss: 1.4749... 0.0746 sec/batch
Epoch: 5/20... Training Step: 17122... Training loss: 1.5781... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17123... Training loss: 1.5294... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17124... Training loss: 1.7397... 0.0726 sec/batch
Epoch: 5/20... Training Step: 17125... Training loss: 1.5713... 0.0769 sec/batch
Epoch: 5/20... Training Step: 17126... Training loss: 1.6318... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17127... Training loss: 1.6872... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17128... Training loss: 1.6139... 0.0761 sec/batch
Epoch: 5/20... Training Step: 17129... Training loss: 1.6652... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17130... Training loss: 1.4994... 0.0574 sec/batch
Epoch: 5/20... Training Step: 17131... Training loss: 1.6692... 0.0621 sec/batch
Epoch: 5/20... Training Step: 17132... Training loss: 1.6135... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17133... Training loss: 1.5444... 0.0604 sec/batch
Epoch: 5/20... Training Step: 17134... Training loss: 1.6753... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17135... Training loss: 1.5873... 0.0711 sec/batch
Epoch: 5/20... Training Step: 17136... Training loss: 1.6394... 0.0582 sec/batch
Epoch: 5/20... Training Step: 17137... Training loss: 1.6873... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17138... Training loss: 1.5412... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17139... Training loss: 1.6479... 0.0582 sec/batch
Epoch: 5/20... Training Step: 17140... Training loss: 1.5713... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17141... Training loss: 1.7606... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17142... Training loss: 1.7069... 0.0575 sec/batch
Epoch: 5/20... Training Step: 17143... Training loss: 1.6075... 0.0763 sec/batch
Epoch: 5/20... Training Step: 17144... Training loss: 1.4588... 0.0673 sec/batch
Epoch: 5/20... Training Step: 17145... Training loss: 1.6108... 0.0605 sec/batch
Epoch: 5/20... Training Step: 17146... Training loss: 1.5901... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17147... Training loss: 1.7425... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17148... Training loss: 1.6156... 0.0582 sec/batch
Epoch: 5/20... Training Step: 17149... Training loss: 1.6897... 0.0525 sec/batch
Epoch: 5/20... Training Step: 17150... Training loss: 1.6592... 0.0670 sec/batch
Epoch: 5/20... Training Step: 17151... Training loss: 1.5623... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17152... Training loss: 1.6629... 0.0525 sec/batch
Epoch: 5/20... Training Step: 17153... Training loss: 1.6883... 0.0621 sec/batch
Epoch: 5/20... Training Step: 17154... Training loss: 1.5522... 0.0523 sec/batch
Epoch: 5/20... Training Step: 17155... Training loss: 1.6417... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17156... Training loss: 1.5650... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17157... Training loss: 1.6879... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17158... Training loss: 1.7899... 0.0604 sec/batch
Epoch: 5/20... Training Step: 17159... Training loss: 1.7291... 0.0759 sec/batch
Epoch: 5/20... Training Step: 17160... Training loss: 1.5765... 0.0660 sec/batch
Epoch: 5/20... Training Step: 17161... Training loss: 1.6341... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17162... Training loss: 1.7769... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17163... Training loss: 1.6455... 0.0657 sec/batch
Epoch: 5/20... Training Step: 17164... Training loss: 1.7714... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17165... Training loss: 1.8229... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17166... Training loss: 1.5514... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17167... Training loss: 1.7101... 0.0703 sec/batch
Epoch: 5/20... Training Step: 17168... Training loss: 1.6096... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17169... Training loss: 1.6616... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17170... Training loss: 1.7756... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17171... Training loss: 1.8240... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17172... Training loss: 1.7923... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17173... Training loss: 1.7380... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17174... Training loss: 1.6483... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17175... Training loss: 1.7109... 0.0731 sec/batch
Epoch: 5/20... Training Step: 17176... Training loss: 1.7743... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17177... Training loss: 1.5858... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17178... Training loss: 1.4933... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17179... Training loss: 1.5577... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17180... Training loss: 1.5810... 0.0640 sec/batch
Epoch: 5/20... Training Step: 17181... Training loss: 1.6653... 0.0533 sec/batch
Epoch: 5/20... Training Step: 17182... Training loss: 1.7393... 0.0688 sec/batch
Epoch: 5/20... Training Step: 17183... Training loss: 1.6376... 0.0649 sec/batch
Epoch: 5/20... Training Step: 17184... Training loss: 1.6841... 0.0648 sec/batch
Epoch: 5/20... Training Step: 17185... Training loss: 1.5850... 0.0677 sec/batch
Epoch: 5/20... Training Step: 17186... Training loss: 1.8447... 0.0573 sec/batch
Epoch: 5/20... Training Step: 17187... Training loss: 1.8602... 0.0747 sec/batch
Epoch: 5/20... Training Step: 17188... Training loss: 1.8030... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17189... Training loss: 1.7160... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17190... Training loss: 1.7293... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17191... Training loss: 1.6219... 0.0678 sec/batch
Epoch: 5/20... Training Step: 17192... Training loss: 1.6329... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17193... Training loss: 1.5980... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17194... Training loss: 1.7318... 0.0583 sec/batch
Epoch: 5/20... Training Step: 17195... Training loss: 1.7363... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17196... Training loss: 1.6559... 0.0553 sec/batch
Epoch: 5/20... Training Step: 17197... Training loss: 1.6880... 0.0591 sec/batch
Epoch: 5/20... Training Step: 17198... Training loss: 1.5964... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17199... Training loss: 1.5903... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17200... Training loss: 1.7269... 0.0520 sec/batch
Epoch: 5/20... Training Step: 17201... Training loss: 1.7440... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17202... Training loss: 1.6527... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17203... Training loss: 1.7946... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17204... Training loss: 1.6552... 0.0544 sec/batch
Epoch: 5/20... Training Step: 17205... Training loss: 1.6864... 0.0541 sec/batch
Epoch: 5/20... Training Step: 17206... Training loss: 1.6387... 0.0724 sec/batch
Epoch: 5/20... Training Step: 17207... Training loss: 1.6773... 0.0588 sec/batch
Epoch: 5/20... Training Step: 17208... Training loss: 1.6564... 0.0558 sec/batch
Epoch: 5/20... Training Step: 17209... Training loss: 1.6585... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17210... Training loss: 1.6719... 0.0700 sec/batch
Epoch: 5/20... Training Step: 17211... Training loss: 1.6703... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17212... Training loss: 1.6823... 0.0546 sec/batch
Epoch: 5/20... Training Step: 17213... Training loss: 1.6209... 0.0520 sec/batch
Epoch: 5/20... Training Step: 17214... Training loss: 1.7027... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17215... Training loss: 1.6639... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17216... Training loss: 1.6021... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17217... Training loss: 1.5315... 0.0591 sec/batch
Epoch: 5/20... Training Step: 17218... Training loss: 1.6427... 0.0667 sec/batch
Epoch: 5/20... Training Step: 17219... Training loss: 1.7274... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17220... Training loss: 1.6432... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17221... Training loss: 1.4699... 0.0523 sec/batch
Epoch: 5/20... Training Step: 17222... Training loss: 1.5239... 0.0797 sec/batch
Epoch: 5/20... Training Step: 17223... Training loss: 1.7174... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17224... Training loss: 1.7106... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17225... Training loss: 1.7138... 0.0588 sec/batch
Epoch: 5/20... Training Step: 17226... Training loss: 1.6598... 0.0695 sec/batch
Epoch: 5/20... Training Step: 17227... Training loss: 1.6091... 0.0533 sec/batch
Epoch: 5/20... Training Step: 17228... Training loss: 1.6492... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17229... Training loss: 1.7430... 0.0578 sec/batch
Epoch: 5/20... Training Step: 17230... Training loss: 1.5218... 0.0576 sec/batch
Epoch: 5/20... Training Step: 17231... Training loss: 1.4979... 0.0566 sec/batch
Epoch: 5/20... Training Step: 17232... Training loss: 1.5564... 0.0532 sec/batch
Epoch: 5/20... Training Step: 17233... Training loss: 1.6068... 0.0595 sec/batch
Epoch: 5/20... Training Step: 17234... Training loss: 1.6102... 0.0519 sec/batch
Epoch: 5/20... Training Step: 17235... Training loss: 1.5894... 0.0535 sec/batch
Epoch: 5/20... Training Step: 17236... Training loss: 1.6231... 0.0563 sec/batch
Epoch: 5/20... Training Step: 17237... Training loss: 1.6742... 0.0554 sec/batch
Epoch: 5/20... Training Step: 17238... Training loss: 1.5435... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17239... Training loss: 1.7694... 0.0595 sec/batch
Epoch: 5/20... Training Step: 17240... Training loss: 1.6162... 0.0589 sec/batch
Epoch: 5/20... Training Step: 17241... Training loss: 1.6410... 0.0580 sec/batch
Epoch: 5/20... Training Step: 17242... Training loss: 1.5738... 0.0577 sec/batch
Epoch: 5/20... Training Step: 17243... Training loss: 1.4951... 0.0583 sec/batch
Epoch: 5/20... Training Step: 17244... Training loss: 1.5901... 0.0580 sec/batch
Epoch: 5/20... Training Step: 17245... Training loss: 1.5755... 0.0514 sec/batch
Epoch: 5/20... Training Step: 17246... Training loss: 1.6312... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17247... Training loss: 1.5608... 0.0548 sec/batch
Epoch: 5/20... Training Step: 17248... Training loss: 1.6470... 0.0540 sec/batch
Epoch: 5/20... Training Step: 17249... Training loss: 1.7922... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17250... Training loss: 1.6810... 0.0677 sec/batch
Epoch: 5/20... Training Step: 17251... Training loss: 1.7068... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17252... Training loss: 1.5572... 0.0588 sec/batch
Epoch: 5/20... Training Step: 17253... Training loss: 1.6109... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17254... Training loss: 1.6879... 0.0588 sec/batch
Epoch: 5/20... Training Step: 17255... Training loss: 1.5680... 0.0534 sec/batch
Epoch: 5/20... Training Step: 17256... Training loss: 1.6427... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17257... Training loss: 1.7623... 0.0563 sec/batch
Epoch: 5/20... Training Step: 17258... Training loss: 1.6019... 0.0560 sec/batch
Epoch: 5/20... Training Step: 17259... Training loss: 1.5127... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17260... Training loss: 1.5933... 0.0688 sec/batch
Epoch: 5/20... Training Step: 17261... Training loss: 1.7061... 0.0544 sec/batch
Epoch: 5/20... Training Step: 17262... Training loss: 1.7327... 0.0686 sec/batch
Epoch: 5/20... Training Step: 17263... Training loss: 1.8150... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17264... Training loss: 1.6633... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17265... Training loss: 1.7179... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17266... Training loss: 1.9021... 0.0677 sec/batch
Epoch: 5/20... Training Step: 17267... Training loss: 1.6654... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17268... Training loss: 1.6868... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17269... Training loss: 1.6963... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17270... Training loss: 1.6364... 0.0709 sec/batch
Epoch: 5/20... Training Step: 17271... Training loss: 1.6837... 0.0705 sec/batch
Epoch: 5/20... Training Step: 17272... Training loss: 1.6680... 0.0620 sec/batch
Epoch: 5/20... Training Step: 17273... Training loss: 1.6744... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17274... Training loss: 1.6462... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17275... Training loss: 1.6850... 0.0574 sec/batch
Epoch: 5/20... Training Step: 17276... Training loss: 1.5885... 0.0650 sec/batch
Epoch: 5/20... Training Step: 17277... Training loss: 1.6320... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17278... Training loss: 1.5552... 0.0690 sec/batch
Epoch: 5/20... Training Step: 17279... Training loss: 1.7461... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17280... Training loss: 1.7685... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17281... Training loss: 1.5187... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17282... Training loss: 1.7305... 0.0681 sec/batch
Epoch: 5/20... Training Step: 17283... Training loss: 1.6040... 0.0586 sec/batch
Epoch: 5/20... Training Step: 17284... Training loss: 1.7213... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17285... Training loss: 1.7064... 0.0671 sec/batch
Epoch: 5/20... Training Step: 17286... Training loss: 1.5721... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17287... Training loss: 1.7019... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17288... Training loss: 1.6971... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17289... Training loss: 1.6997... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17290... Training loss: 1.6364... 0.0677 sec/batch
Epoch: 5/20... Training Step: 17291... Training loss: 1.6057... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17292... Training loss: 1.6798... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17293... Training loss: 1.6821... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17294... Training loss: 1.9303... 0.0684 sec/batch
Epoch: 5/20... Training Step: 17295... Training loss: 1.6731... 0.0530 sec/batch
Epoch: 5/20... Training Step: 17296... Training loss: 1.7006... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17297... Training loss: 1.7770... 0.0538 sec/batch
Epoch: 5/20... Training Step: 17298... Training loss: 1.7219... 0.0600 sec/batch
Epoch: 5/20... Training Step: 17299... Training loss: 1.6094... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17300... Training loss: 1.7377... 0.0579 sec/batch
Epoch: 5/20... Training Step: 17301... Training loss: 1.8401... 0.0594 sec/batch
Epoch: 5/20... Training Step: 17302... Training loss: 1.6818... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17303... Training loss: 1.7349... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17304... Training loss: 1.6130... 0.0549 sec/batch
Epoch: 5/20... Training Step: 17305... Training loss: 1.6221... 0.0566 sec/batch
Epoch: 5/20... Training Step: 17306... Training loss: 1.6809... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17307... Training loss: 1.5623... 0.0553 sec/batch
Epoch: 5/20... Training Step: 17308... Training loss: 1.7329... 0.0550 sec/batch
Epoch: 5/20... Training Step: 17309... Training loss: 1.6863... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17310... Training loss: 1.6760... 0.0760 sec/batch
Epoch: 5/20... Training Step: 17311... Training loss: 1.6880... 0.0581 sec/batch
Epoch: 5/20... Training Step: 17312... Training loss: 1.7485... 0.0529 sec/batch
Epoch: 5/20... Training Step: 17313... Training loss: 1.8526... 0.0585 sec/batch
Epoch: 5/20... Training Step: 17314... Training loss: 1.6250... 0.0688 sec/batch
Epoch: 5/20... Training Step: 17315... Training loss: 1.6856... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17316... Training loss: 1.6660... 0.0675 sec/batch
Epoch: 5/20... Training Step: 17317... Training loss: 1.6577... 0.0661 sec/batch
Epoch: 5/20... Training Step: 17318... Training loss: 1.6553... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17319... Training loss: 1.8123... 0.0562 sec/batch
Epoch: 5/20... Training Step: 17320... Training loss: 1.5830... 0.0531 sec/batch
Epoch: 5/20... Training Step: 17321... Training loss: 1.5261... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17322... Training loss: 1.6740... 0.0742 sec/batch
Epoch: 5/20... Training Step: 17323... Training loss: 1.7267... 0.0534 sec/batch
Epoch: 5/20... Training Step: 17324... Training loss: 1.7074... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17325... Training loss: 1.7221... 0.0613 sec/batch
Epoch: 5/20... Training Step: 17326... Training loss: 1.7835... 0.0573 sec/batch
Epoch: 5/20... Training Step: 17327... Training loss: 1.7604... 0.0589 sec/batch
Epoch: 5/20... Training Step: 17328... Training loss: 1.5895... 0.0555 sec/batch
Epoch: 5/20... Training Step: 17329... Training loss: 1.7004... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17330... Training loss: 1.6654... 0.0659 sec/batch
Epoch: 5/20... Training Step: 17331... Training loss: 1.6364... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17332... Training loss: 1.6456... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17333... Training loss: 1.8302... 0.0551 sec/batch
Epoch: 5/20... Training Step: 17334... Training loss: 1.8126... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17335... Training loss: 1.7475... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17336... Training loss: 1.5060... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17337... Training loss: 1.6404... 0.0542 sec/batch
Epoch: 5/20... Training Step: 17338... Training loss: 1.6613... 0.0719 sec/batch
Epoch: 5/20... Training Step: 17339... Training loss: 1.6248... 0.0578 sec/batch
Epoch: 5/20... Training Step: 17340... Training loss: 1.6590... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17341... Training loss: 1.6455... 0.0589 sec/batch
Epoch: 5/20... Training Step: 17342... Training loss: 1.5985... 0.0540 sec/batch
Epoch: 5/20... Training Step: 17343... Training loss: 1.6161... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17344... Training loss: 1.6305... 0.0545 sec/batch
Epoch: 5/20... Training Step: 17345... Training loss: 1.6776... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17346... Training loss: 1.7027... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17347... Training loss: 1.5612... 0.0585 sec/batch
Epoch: 5/20... Training Step: 17348... Training loss: 1.6227... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17349... Training loss: 1.6202... 0.0579 sec/batch
Epoch: 5/20... Training Step: 17350... Training loss: 1.5883... 0.0693 sec/batch
Epoch: 5/20... Training Step: 17351... Training loss: 1.6518... 0.0522 sec/batch
Epoch: 5/20... Training Step: 17352... Training loss: 1.7680... 0.0588 sec/batch
Epoch: 5/20... Training Step: 17353... Training loss: 1.6626... 0.0519 sec/batch
Epoch: 5/20... Training Step: 17354... Training loss: 1.6948... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17355... Training loss: 1.5900... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17356... Training loss: 1.5519... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17357... Training loss: 1.7916... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17358... Training loss: 1.7189... 0.0700 sec/batch
Epoch: 5/20... Training Step: 17359... Training loss: 1.6725... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17360... Training loss: 1.6621... 0.0529 sec/batch
Epoch: 5/20... Training Step: 17361... Training loss: 1.6386... 0.0505 sec/batch
Epoch: 5/20... Training Step: 17362... Training loss: 1.5965... 0.0681 sec/batch
Epoch: 5/20... Training Step: 17363... Training loss: 1.6652... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17364... Training loss: 1.7344... 0.0601 sec/batch
Epoch: 5/20... Training Step: 17365... Training loss: 1.5597... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17366... Training loss: 1.7136... 0.0732 sec/batch
Epoch: 5/20... Training Step: 17367... Training loss: 1.7143... 0.0582 sec/batch
Epoch: 5/20... Training Step: 17368... Training loss: 1.6944... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17369... Training loss: 1.7584... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17370... Training loss: 1.7635... 0.0759 sec/batch
Epoch: 5/20... Training Step: 17371... Training loss: 1.7770... 0.0554 sec/batch
Epoch: 5/20... Training Step: 17372... Training loss: 1.5533... 0.0538 sec/batch
Epoch: 5/20... Training Step: 17373... Training loss: 1.8041... 0.0560 sec/batch
Epoch: 5/20... Training Step: 17374... Training loss: 1.5750... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17375... Training loss: 1.6374... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17376... Training loss: 1.5702... 0.0662 sec/batch
Epoch: 5/20... Training Step: 17377... Training loss: 1.5922... 0.0558 sec/batch
Epoch: 5/20... Training Step: 17378... Training loss: 1.6612... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17379... Training loss: 1.7387... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17380... Training loss: 1.6910... 0.0600 sec/batch
Epoch: 5/20... Training Step: 17381... Training loss: 1.6460... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17382... Training loss: 1.6572... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17383... Training loss: 1.8144... 0.0579 sec/batch
Epoch: 5/20... Training Step: 17384... Training loss: 1.8194... 0.0558 sec/batch
Epoch: 5/20... Training Step: 17385... Training loss: 1.5260... 0.0580 sec/batch
Epoch: 5/20... Training Step: 17386... Training loss: 1.6167... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17387... Training loss: 1.7382... 0.0656 sec/batch
Epoch: 5/20... Training Step: 17388... Training loss: 1.5802... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17389... Training loss: 1.6286... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17390... Training loss: 1.6039... 0.0700 sec/batch
Epoch: 5/20... Training Step: 17391... Training loss: 1.6056... 0.0684 sec/batch
Epoch: 5/20... Training Step: 17392... Training loss: 1.4947... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17393... Training loss: 1.5946... 0.0743 sec/batch
Epoch: 5/20... Training Step: 17394... Training loss: 1.5850... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17395... Training loss: 1.6189... 0.0727 sec/batch
Epoch: 5/20... Training Step: 17396... Training loss: 1.7357... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17397... Training loss: 1.6840... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17398... Training loss: 1.6181... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17399... Training loss: 1.7697... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17400... Training loss: 1.6126... 0.0699 sec/batch
Epoch: 5/20... Training Step: 17401... Training loss: 1.6396... 0.0576 sec/batch
Epoch: 5/20... Training Step: 17402... Training loss: 1.8234... 0.0604 sec/batch
Epoch: 5/20... Training Step: 17403... Training loss: 1.7040... 0.0607 sec/batch
Epoch: 5/20... Training Step: 17404... Training loss: 1.6984... 0.0627 sec/batch
Epoch: 5/20... Training Step: 17405... Training loss: 1.7586... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17406... Training loss: 1.6586... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17407... Training loss: 1.3909... 0.0570 sec/batch
Epoch: 5/20... Training Step: 17408... Training loss: 1.7151... 0.0649 sec/batch
Epoch: 5/20... Training Step: 17409... Training loss: 1.6724... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17410... Training loss: 1.6690... 0.0546 sec/batch
Epoch: 5/20... Training Step: 17411... Training loss: 1.7003... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17412... Training loss: 1.8241... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17413... Training loss: 1.6312... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17414... Training loss: 1.6048... 0.0666 sec/batch
Epoch: 5/20... Training Step: 17415... Training loss: 1.6341... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17416... Training loss: 1.5449... 0.0565 sec/batch
Epoch: 5/20... Training Step: 17417... Training loss: 1.5629... 0.0685 sec/batch
Epoch: 5/20... Training Step: 17418... Training loss: 1.6464... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17419... Training loss: 1.8255... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17420... Training loss: 1.7574... 0.0666 sec/batch
Epoch: 5/20... Training Step: 17421... Training loss: 1.7069... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17422... Training loss: 1.6500... 0.0671 sec/batch
Epoch: 5/20... Training Step: 17423... Training loss: 1.6920... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17424... Training loss: 1.8275... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17425... Training loss: 1.4625... 0.0700 sec/batch
Epoch: 5/20... Training Step: 17426... Training loss: 1.7557... 0.0611 sec/batch
Epoch: 5/20... Training Step: 17427... Training loss: 1.5569... 0.0705 sec/batch
Epoch: 5/20... Training Step: 17428... Training loss: 1.5055... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17429... Training loss: 1.4447... 0.0666 sec/batch
Epoch: 5/20... Training Step: 17430... Training loss: 1.5695... 0.0594 sec/batch
Epoch: 5/20... Training Step: 17431... Training loss: 1.5237... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17432... Training loss: 1.6901... 0.0720 sec/batch
Epoch: 5/20... Training Step: 17433... Training loss: 1.6655... 0.0605 sec/batch
Epoch: 5/20... Training Step: 17434... Training loss: 1.7278... 0.0595 sec/batch
Epoch: 5/20... Training Step: 17435... Training loss: 1.7479... 0.0667 sec/batch
Epoch: 5/20... Training Step: 17436... Training loss: 1.7165... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17437... Training loss: 1.7240... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17438... Training loss: 1.7584... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17439... Training loss: 1.6735... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17440... Training loss: 1.6642... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17441... Training loss: 1.5579... 0.0685 sec/batch
Epoch: 5/20... Training Step: 17442... Training loss: 1.7944... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17443... Training loss: 1.6435... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17444... Training loss: 1.6815... 0.0560 sec/batch
Epoch: 5/20... Training Step: 17445... Training loss: 1.7059... 0.0722 sec/batch
Epoch: 5/20... Training Step: 17446... Training loss: 1.8598... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17447... Training loss: 1.6355... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17448... Training loss: 1.6290... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17449... Training loss: 1.6500... 0.0695 sec/batch
Epoch: 5/20... Training Step: 17450... Training loss: 1.6474... 0.0542 sec/batch
Epoch: 5/20... Training Step: 17451... Training loss: 1.5869... 0.0655 sec/batch
Epoch: 5/20... Training Step: 17452... Training loss: 1.6142... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17453... Training loss: 1.7232... 0.0699 sec/batch
Epoch: 5/20... Training Step: 17454... Training loss: 1.6221... 0.0535 sec/batch
Epoch: 5/20... Training Step: 17455... Training loss: 1.7120... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17456... Training loss: 1.6457... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17457... Training loss: 1.6435... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17458... Training loss: 1.8429... 0.0552 sec/batch
Epoch: 5/20... Training Step: 17459... Training loss: 1.7830... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17460... Training loss: 1.6663... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17461... Training loss: 1.8560... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17462... Training loss: 1.7180... 0.0548 sec/batch
Epoch: 5/20... Training Step: 17463... Training loss: 1.7157... 0.0628 sec/batch
Epoch: 5/20... Training Step: 17464... Training loss: 1.5582... 0.0569 sec/batch
Epoch: 5/20... Training Step: 17465... Training loss: 1.6879... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17466... Training loss: 1.6285... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17467... Training loss: 1.6410... 0.0595 sec/batch
Epoch: 5/20... Training Step: 17468... Training loss: 1.6380... 0.0563 sec/batch
Epoch: 5/20... Training Step: 17469... Training loss: 1.7742... 0.0717 sec/batch
Epoch: 5/20... Training Step: 17470... Training loss: 1.6737... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17471... Training loss: 1.6080... 0.0650 sec/batch
Epoch: 5/20... Training Step: 17472... Training loss: 1.6707... 0.0581 sec/batch
Epoch: 5/20... Training Step: 17473... Training loss: 1.5643... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17474... Training loss: 1.6697... 0.0541 sec/batch
Epoch: 5/20... Training Step: 17475... Training loss: 1.6924... 0.0549 sec/batch
Epoch: 5/20... Training Step: 17476... Training loss: 1.6213... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17477... Training loss: 1.5797... 0.0601 sec/batch
Epoch: 5/20... Training Step: 17478... Training loss: 1.6611... 0.0619 sec/batch
Epoch: 5/20... Training Step: 17479... Training loss: 1.6700... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17480... Training loss: 1.6836... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17481... Training loss: 1.7232... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17482... Training loss: 1.6674... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17483... Training loss: 1.6933... 0.0656 sec/batch
Epoch: 5/20... Training Step: 17484... Training loss: 1.8231... 0.0579 sec/batch
Epoch: 5/20... Training Step: 17485... Training loss: 1.8529... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17486... Training loss: 1.7015... 0.0568 sec/batch
Epoch: 5/20... Training Step: 17487... Training loss: 1.8725... 0.0522 sec/batch
Epoch: 5/20... Training Step: 17488... Training loss: 1.8681... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17489... Training loss: 1.7497... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17490... Training loss: 1.7548... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17491... Training loss: 1.9288... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17492... Training loss: 1.6403... 0.0529 sec/batch
Epoch: 5/20... Training Step: 17493... Training loss: 1.7109... 0.0585 sec/batch
Epoch: 5/20... Training Step: 17494... Training loss: 1.7435... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17495... Training loss: 1.7469... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17496... Training loss: 1.6541... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17497... Training loss: 1.7698... 0.0639 sec/batch
Epoch: 5/20... Training Step: 17498... Training loss: 1.7971... 0.0604 sec/batch
Epoch: 5/20... Training Step: 17499... Training loss: 1.7785... 0.0673 sec/batch
Epoch: 5/20... Training Step: 17500... Training loss: 1.7662... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17501... Training loss: 1.7065... 0.0640 sec/batch
Epoch: 5/20... Training Step: 17502... Training loss: 1.6654... 0.0574 sec/batch
Epoch: 5/20... Training Step: 17503... Training loss: 1.6183... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17504... Training loss: 1.6013... 0.0520 sec/batch
Epoch: 5/20... Training Step: 17505... Training loss: 1.7603... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17506... Training loss: 1.7147... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17507... Training loss: 1.5263... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17508... Training loss: 1.7742... 0.0581 sec/batch
Epoch: 5/20... Training Step: 17509... Training loss: 1.7079... 0.0527 sec/batch
Epoch: 5/20... Training Step: 17510... Training loss: 1.7244... 0.0561 sec/batch
Epoch: 5/20... Training Step: 17511... Training loss: 1.6541... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17512... Training loss: 1.4953... 0.0591 sec/batch
Epoch: 5/20... Training Step: 17513... Training loss: 1.6165... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17514... Training loss: 1.6067... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17515... Training loss: 1.6884... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17516... Training loss: 1.7032... 0.0582 sec/batch
Epoch: 5/20... Training Step: 17517... Training loss: 1.6908... 0.0570 sec/batch
Epoch: 5/20... Training Step: 17518... Training loss: 1.6251... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17519... Training loss: 1.6578... 0.0611 sec/batch
Epoch: 5/20... Training Step: 17520... Training loss: 1.7552... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17521... Training loss: 1.5657... 0.0769 sec/batch
Epoch: 5/20... Training Step: 17522... Training loss: 1.6449... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17523... Training loss: 1.5964... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17524... Training loss: 1.6126... 0.0655 sec/batch
Epoch: 5/20... Training Step: 17525... Training loss: 1.6531... 0.0733 sec/batch
Epoch: 5/20... Training Step: 17526... Training loss: 1.5938... 0.0516 sec/batch
Epoch: 5/20... Training Step: 17527... Training loss: 1.5413... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17528... Training loss: 1.7515... 0.0621 sec/batch
Epoch: 5/20... Training Step: 17529... Training loss: 1.5041... 0.0668 sec/batch
Epoch: 5/20... Training Step: 17530... Training loss: 1.6040... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17531... Training loss: 1.6400... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17532... Training loss: 1.8682... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17533... Training loss: 1.6906... 0.0634 sec/batch
Epoch: 5/20... Training Step: 17534... Training loss: 1.4898... 0.0639 sec/batch
Epoch: 5/20... Training Step: 17535... Training loss: 1.5740... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17536... Training loss: 1.4732... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17537... Training loss: 1.6644... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17538... Training loss: 1.7272... 0.0611 sec/batch
Epoch: 5/20... Training Step: 17539... Training loss: 1.7883... 0.0572 sec/batch
Epoch: 5/20... Training Step: 17540... Training loss: 1.5420... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17541... Training loss: 1.6621... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17542... Training loss: 1.6228... 0.0627 sec/batch
Epoch: 5/20... Training Step: 17543... Training loss: 1.6789... 0.0499 sec/batch
Epoch: 5/20... Training Step: 17544... Training loss: 1.7771... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17545... Training loss: 1.9523... 0.0706 sec/batch
Epoch: 5/20... Training Step: 17546... Training loss: 1.7259... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17547... Training loss: 1.6287... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17548... Training loss: 1.8080... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17549... Training loss: 1.6481... 0.0634 sec/batch
Epoch: 5/20... Training Step: 17550... Training loss: 1.7993... 0.0549 sec/batch
Epoch: 5/20... Training Step: 17551... Training loss: 1.6273... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17552... Training loss: 1.7236... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17553... Training loss: 1.6495... 0.0731 sec/batch
Epoch: 5/20... Training Step: 17554... Training loss: 1.7319... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17555... Training loss: 1.8425... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17556... Training loss: 1.7033... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17557... Training loss: 1.7669... 0.0690 sec/batch
Epoch: 5/20... Training Step: 17558... Training loss: 1.7539... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17559... Training loss: 1.6754... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17560... Training loss: 1.5869... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17561... Training loss: 1.7169... 0.0671 sec/batch
Epoch: 5/20... Training Step: 17562... Training loss: 1.7071... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17563... Training loss: 1.6996... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17564... Training loss: 1.5967... 0.0577 sec/batch
Epoch: 5/20... Training Step: 17565... Training loss: 1.5813... 0.0676 sec/batch
Epoch: 5/20... Training Step: 17566... Training loss: 1.6215... 0.0526 sec/batch
Epoch: 5/20... Training Step: 17567... Training loss: 1.6543... 0.0661 sec/batch
Epoch: 5/20... Training Step: 17568... Training loss: 1.8422... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17569... Training loss: 1.7109... 0.0703 sec/batch
Epoch: 5/20... Training Step: 17570... Training loss: 1.5841... 0.0585 sec/batch
Epoch: 5/20... Training Step: 17571... Training loss: 1.6641... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17572... Training loss: 1.6403... 0.0605 sec/batch
Epoch: 5/20... Training Step: 17573... Training loss: 1.5278... 0.0720 sec/batch
Epoch: 5/20... Training Step: 17574... Training loss: 1.6566... 0.0589 sec/batch
Epoch: 5/20... Training Step: 17575... Training loss: 1.6983... 0.0607 sec/batch
Epoch: 5/20... Training Step: 17576... Training loss: 1.6435... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17577... Training loss: 1.5134... 0.0691 sec/batch
Epoch: 5/20... Training Step: 17578... Training loss: 1.7282... 0.0893 sec/batch
Epoch: 5/20... Training Step: 17579... Training loss: 1.6653... 0.0867 sec/batch
Epoch: 5/20... Training Step: 17580... Training loss: 1.7353... 0.0972 sec/batch
Epoch: 5/20... Training Step: 17581... Training loss: 1.5883... 0.0923 sec/batch
Epoch: 5/20... Training Step: 17582... Training loss: 1.6736... 0.0810 sec/batch
Epoch: 5/20... Training Step: 17583... Training loss: 1.6609... 0.0865 sec/batch
Epoch: 5/20... Training Step: 17584... Training loss: 1.6858... 0.0831 sec/batch
Epoch: 5/20... Training Step: 17585... Training loss: 1.7933... 0.0914 sec/batch
Epoch: 5/20... Training Step: 17586... Training loss: 1.7714... 0.0918 sec/batch
Epoch: 5/20... Training Step: 17587... Training loss: 1.6438... 0.0862 sec/batch
Epoch: 5/20... Training Step: 17588... Training loss: 1.6120... 0.0919 sec/batch
Epoch: 5/20... Training Step: 17589... Training loss: 1.7505... 0.0912 sec/batch
Epoch: 5/20... Training Step: 17590... Training loss: 1.5882... 0.0799 sec/batch
Epoch: 5/20... Training Step: 17591... Training loss: 1.7110... 0.0952 sec/batch
Epoch: 5/20... Training Step: 17592... Training loss: 1.6379... 0.0913 sec/batch
Epoch: 5/20... Training Step: 17593... Training loss: 1.5245... 0.0882 sec/batch
Epoch: 5/20... Training Step: 17594... Training loss: 1.7496... 0.0854 sec/batch
Epoch: 5/20... Training Step: 17595... Training loss: 1.7337... 0.0905 sec/batch
Epoch: 5/20... Training Step: 17596... Training loss: 1.7157... 0.0817 sec/batch
Epoch: 5/20... Training Step: 17597... Training loss: 1.5096... 0.0832 sec/batch
Epoch: 5/20... Training Step: 17598... Training loss: 1.6335... 0.0952 sec/batch
Epoch: 5/20... Training Step: 17599... Training loss: 1.6665... 0.0911 sec/batch
Epoch: 5/20... Training Step: 17600... Training loss: 1.6098... 0.0885 sec/batch
Epoch: 5/20... Training Step: 17601... Training loss: 1.6400... 0.1011 sec/batch
Epoch: 5/20... Training Step: 17602... Training loss: 1.7113... 0.1018 sec/batch
Epoch: 5/20... Training Step: 17603... Training loss: 1.6411... 0.0928 sec/batch
Epoch: 5/20... Training Step: 17604... Training loss: 1.5397... 0.1000 sec/batch
Epoch: 5/20... Training Step: 17605... Training loss: 1.6175... 0.0807 sec/batch
Epoch: 5/20... Training Step: 17606... Training loss: 1.6524... 0.0934 sec/batch
Epoch: 5/20... Training Step: 17607... Training loss: 1.5598... 0.0814 sec/batch
Epoch: 5/20... Training Step: 17608... Training loss: 1.6722... 0.0866 sec/batch
Epoch: 5/20... Training Step: 17609... Training loss: 1.6329... 0.0779 sec/batch
Epoch: 5/20... Training Step: 17610... Training loss: 1.5151... 0.0869 sec/batch
Epoch: 5/20... Training Step: 17611... Training loss: 1.6855... 0.0906 sec/batch
Epoch: 5/20... Training Step: 17612... Training loss: 1.8480... 0.0850 sec/batch
Epoch: 5/20... Training Step: 17613... Training loss: 1.7459... 0.1054 sec/batch
Epoch: 5/20... Training Step: 17614... Training loss: 1.8154... 0.0975 sec/batch
Epoch: 5/20... Training Step: 17615... Training loss: 1.7279... 0.0864 sec/batch
Epoch: 5/20... Training Step: 17616... Training loss: 1.6070... 0.1045 sec/batch
Epoch: 5/20... Training Step: 17617... Training loss: 1.6173... 0.0938 sec/batch
Epoch: 5/20... Training Step: 17618... Training loss: 1.6288... 0.0900 sec/batch
Epoch: 5/20... Training Step: 17619... Training loss: 1.6965... 0.0941 sec/batch
Epoch: 5/20... Training Step: 17620... Training loss: 1.6669... 0.0769 sec/batch
Epoch: 5/20... Training Step: 17621... Training loss: 1.6892... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17622... Training loss: 1.6679... 0.0600 sec/batch
Epoch: 5/20... Training Step: 17623... Training loss: 1.7255... 0.0674 sec/batch
Epoch: 5/20... Training Step: 17624... Training loss: 1.6091... 0.0665 sec/batch
Epoch: 5/20... Training Step: 17625... Training loss: 1.6374... 0.0688 sec/batch
Epoch: 5/20... Training Step: 17626... Training loss: 1.5343... 0.0745 sec/batch
Epoch: 5/20... Training Step: 17627... Training loss: 1.6768... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17628... Training loss: 1.7822... 0.0701 sec/batch
Epoch: 5/20... Training Step: 17629... Training loss: 1.4971... 0.0680 sec/batch
Epoch: 5/20... Training Step: 17630... Training loss: 1.7270... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17631... Training loss: 1.6781... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17632... Training loss: 1.5741... 0.0714 sec/batch
Epoch: 5/20... Training Step: 17633... Training loss: 1.5177... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17634... Training loss: 1.6946... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17635... Training loss: 1.6735... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17636... Training loss: 1.4956... 0.0764 sec/batch
Epoch: 5/20... Training Step: 17637... Training loss: 1.5951... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17638... Training loss: 1.5673... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17639... Training loss: 1.7524... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17640... Training loss: 1.6022... 0.0839 sec/batch
Epoch: 5/20... Training Step: 17641... Training loss: 1.6828... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17642... Training loss: 1.7084... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17643... Training loss: 1.5652... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17644... Training loss: 1.6407... 0.0711 sec/batch
Epoch: 5/20... Training Step: 17645... Training loss: 1.7321... 0.0669 sec/batch
Epoch: 5/20... Training Step: 17646... Training loss: 1.7342... 0.0657 sec/batch
Epoch: 5/20... Training Step: 17647... Training loss: 1.4956... 0.0699 sec/batch
Epoch: 5/20... Training Step: 17648... Training loss: 1.6117... 0.0719 sec/batch
Epoch: 5/20... Training Step: 17649... Training loss: 1.7672... 0.0620 sec/batch
Epoch: 5/20... Training Step: 17650... Training loss: 1.7412... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17651... Training loss: 1.4692... 0.0708 sec/batch
Epoch: 5/20... Training Step: 17652... Training loss: 1.5976... 0.0648 sec/batch
Epoch: 5/20... Training Step: 17653... Training loss: 1.6142... 0.0677 sec/batch
Epoch: 5/20... Training Step: 17654... Training loss: 1.7813... 0.0687 sec/batch
Epoch: 5/20... Training Step: 17655... Training loss: 1.7712... 0.0674 sec/batch
Epoch: 5/20... Training Step: 17656... Training loss: 1.4548... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17657... Training loss: 1.7259... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17658... Training loss: 1.4917... 0.0724 sec/batch
Epoch: 5/20... Training Step: 17659... Training loss: 1.5987... 0.0724 sec/batch
Epoch: 5/20... Training Step: 17660... Training loss: 1.6867... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17661... Training loss: 1.6690... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17662... Training loss: 1.5915... 0.0744 sec/batch
Epoch: 5/20... Training Step: 17663... Training loss: 1.6435... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17664... Training loss: 1.6165... 0.0703 sec/batch
Epoch: 5/20... Training Step: 17665... Training loss: 1.6220... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17666... Training loss: 1.6755... 0.0665 sec/batch
Epoch: 5/20... Training Step: 17667... Training loss: 1.7272... 0.0678 sec/batch
Epoch: 5/20... Training Step: 17668... Training loss: 1.6358... 0.0634 sec/batch
Epoch: 5/20... Training Step: 17669... Training loss: 1.6573... 0.0697 sec/batch
Epoch: 5/20... Training Step: 17670... Training loss: 1.6034... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17671... Training loss: 1.6742... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17672... Training loss: 1.8209... 0.0695 sec/batch
Epoch: 5/20... Training Step: 17673... Training loss: 1.5936... 0.0620 sec/batch
Epoch: 5/20... Training Step: 17674... Training loss: 1.4567... 0.0755 sec/batch
Epoch: 5/20... Training Step: 17675... Training loss: 1.4100... 0.0667 sec/batch
Epoch: 5/20... Training Step: 17676... Training loss: 1.6157... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17677... Training loss: 1.4120... 0.0672 sec/batch
Epoch: 5/20... Training Step: 17678... Training loss: 1.5291... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17679... Training loss: 1.5752... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17680... Training loss: 1.6711... 0.0559 sec/batch
Epoch: 5/20... Training Step: 17681... Training loss: 1.6509... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17682... Training loss: 1.6635... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17683... Training loss: 1.6364... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17684... Training loss: 1.5473... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17685... Training loss: 1.5135... 0.0568 sec/batch
Epoch: 5/20... Training Step: 17686... Training loss: 1.5790... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17687... Training loss: 1.6369... 0.0731 sec/batch
Epoch: 5/20... Training Step: 17688... Training loss: 1.6736... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17689... Training loss: 1.5290... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17690... Training loss: 1.4631... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17691... Training loss: 1.5595... 0.0693 sec/batch
Epoch: 5/20... Training Step: 17692... Training loss: 1.6528... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17693... Training loss: 1.5081... 0.0640 sec/batch
Epoch: 5/20... Training Step: 17694... Training loss: 1.6076... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17695... Training loss: 1.6760... 0.0704 sec/batch
Epoch: 5/20... Training Step: 17696... Training loss: 1.7612... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17697... Training loss: 1.5790... 0.0594 sec/batch
Epoch: 5/20... Training Step: 17698... Training loss: 1.5778... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17699... Training loss: 1.6445... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17700... Training loss: 1.6672... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17701... Training loss: 1.7239... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17702... Training loss: 1.6712... 0.0537 sec/batch
Epoch: 5/20... Training Step: 17703... Training loss: 1.6423... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17704... Training loss: 1.6449... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17705... Training loss: 1.6559... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17706... Training loss: 1.6088... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17707... Training loss: 1.7090... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17708... Training loss: 1.6617... 0.0674 sec/batch
Epoch: 5/20... Training Step: 17709... Training loss: 1.7544... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17710... Training loss: 1.6015... 0.0570 sec/batch
Epoch: 5/20... Training Step: 17711... Training loss: 1.7049... 0.0662 sec/batch
Epoch: 5/20... Training Step: 17712... Training loss: 1.6677... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17713... Training loss: 1.6034... 0.0631 sec/batch
Epoch: 5/20... Training Step: 17714... Training loss: 1.7163... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17715... Training loss: 1.5631... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17716... Training loss: 1.6382... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17717... Training loss: 1.6911... 0.0600 sec/batch
Epoch: 5/20... Training Step: 17718... Training loss: 1.5899... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17719... Training loss: 1.5742... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17720... Training loss: 1.5996... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17721... Training loss: 1.5976... 0.0611 sec/batch
Epoch: 5/20... Training Step: 17722... Training loss: 1.6498... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17723... Training loss: 1.4238... 0.0661 sec/batch
Epoch: 5/20... Training Step: 17724... Training loss: 1.5052... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17725... Training loss: 1.6951... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17726... Training loss: 1.7094... 0.0625 sec/batch
Epoch: 5/20... Training Step: 17727... Training loss: 1.6234... 0.0706 sec/batch
Epoch: 5/20... Training Step: 17728... Training loss: 1.6673... 0.0691 sec/batch
Epoch: 5/20... Training Step: 17729... Training loss: 1.5929... 0.0680 sec/batch
Epoch: 5/20... Training Step: 17730... Training loss: 1.7777... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17731... Training loss: 1.6270... 0.0729 sec/batch
Epoch: 5/20... Training Step: 17732... Training loss: 1.8063... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17733... Training loss: 1.8179... 0.0720 sec/batch
Epoch: 5/20... Training Step: 17734... Training loss: 1.8041... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17735... Training loss: 1.7194... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17736... Training loss: 1.4482... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17737... Training loss: 1.8901... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17738... Training loss: 1.7451... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17739... Training loss: 1.5804... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17740... Training loss: 1.7216... 0.0650 sec/batch
Epoch: 5/20... Training Step: 17741... Training loss: 1.7197... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17742... Training loss: 1.6994... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17743... Training loss: 1.7047... 0.0734 sec/batch
Epoch: 5/20... Training Step: 17744... Training loss: 1.7759... 0.0586 sec/batch
Epoch: 5/20... Training Step: 17745... Training loss: 1.6275... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17746... Training loss: 1.4898... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17747... Training loss: 1.7599... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17748... Training loss: 1.6745... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17749... Training loss: 1.6663... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17750... Training loss: 1.5753... 0.0648 sec/batch
Epoch: 5/20... Training Step: 17751... Training loss: 1.6332... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17752... Training loss: 1.4894... 0.0618 sec/batch
Epoch: 5/20... Training Step: 17753... Training loss: 1.5113... 0.0693 sec/batch
Epoch: 5/20... Training Step: 17754... Training loss: 1.5663... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17755... Training loss: 1.6262... 0.0811 sec/batch
Epoch: 5/20... Training Step: 17756... Training loss: 1.7717... 0.0745 sec/batch
Epoch: 5/20... Training Step: 17757... Training loss: 1.6251... 0.0502 sec/batch
Epoch: 5/20... Training Step: 17758... Training loss: 1.6886... 0.0654 sec/batch
Epoch: 5/20... Training Step: 17759... Training loss: 1.5826... 0.0686 sec/batch
Epoch: 5/20... Training Step: 17760... Training loss: 1.7253... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17761... Training loss: 1.8052... 0.0608 sec/batch
Epoch: 5/20... Training Step: 17762... Training loss: 1.5029... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17763... Training loss: 1.7211... 0.0751 sec/batch
Epoch: 5/20... Training Step: 17764... Training loss: 1.7548... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17765... Training loss: 1.4852... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17766... Training loss: 1.6242... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17767... Training loss: 1.7619... 0.0787 sec/batch
Epoch: 5/20... Training Step: 17768... Training loss: 1.7286... 0.0696 sec/batch
Epoch: 5/20... Training Step: 17769... Training loss: 1.5797... 0.0601 sec/batch
Epoch: 5/20... Training Step: 17770... Training loss: 1.6455... 0.0552 sec/batch
Epoch: 5/20... Training Step: 17771... Training loss: 1.7547... 0.0786 sec/batch
Epoch: 5/20... Training Step: 17772... Training loss: 1.7005... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17773... Training loss: 1.6369... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17774... Training loss: 1.6840... 0.0744 sec/batch
Epoch: 5/20... Training Step: 17775... Training loss: 1.5504... 0.0772 sec/batch
Epoch: 5/20... Training Step: 17776... Training loss: 1.6172... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17777... Training loss: 1.7345... 0.0709 sec/batch
Epoch: 5/20... Training Step: 17778... Training loss: 1.7607... 0.0776 sec/batch
Epoch: 5/20... Training Step: 17779... Training loss: 1.6405... 0.0701 sec/batch
Epoch: 5/20... Training Step: 17780... Training loss: 1.4604... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17781... Training loss: 1.6683... 0.0670 sec/batch
Epoch: 5/20... Training Step: 17782... Training loss: 1.7447... 0.0671 sec/batch
Epoch: 5/20... Training Step: 17783... Training loss: 1.6117... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17784... Training loss: 1.4418... 0.0685 sec/batch
Epoch: 5/20... Training Step: 17785... Training loss: 1.5220... 0.0666 sec/batch
Epoch: 5/20... Training Step: 17786... Training loss: 1.7202... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17787... Training loss: 1.6262... 0.0592 sec/batch
Epoch: 5/20... Training Step: 17788... Training loss: 1.7539... 0.0826 sec/batch
Epoch: 5/20... Training Step: 17789... Training loss: 1.6522... 0.0675 sec/batch
Epoch: 5/20... Training Step: 17790... Training loss: 1.5435... 0.0692 sec/batch
Epoch: 5/20... Training Step: 17791... Training loss: 1.6128... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17792... Training loss: 1.6796... 0.0682 sec/batch
Epoch: 5/20... Training Step: 17793... Training loss: 1.6252... 0.0770 sec/batch
Epoch: 5/20... Training Step: 17794... Training loss: 1.6647... 0.0775 sec/batch
Epoch: 5/20... Training Step: 17795... Training loss: 1.6118... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17796... Training loss: 1.6606... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17797... Training loss: 1.5733... 0.0686 sec/batch
Epoch: 5/20... Training Step: 17798... Training loss: 1.7018... 0.0735 sec/batch
Epoch: 5/20... Training Step: 17799... Training loss: 1.5720... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17800... Training loss: 1.6130... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17801... Training loss: 1.7025... 0.0541 sec/batch
Epoch: 5/20... Training Step: 17802... Training loss: 1.4797... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17803... Training loss: 1.5683... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17804... Training loss: 1.5439... 0.0623 sec/batch
Epoch: 5/20... Training Step: 17805... Training loss: 1.6206... 0.0713 sec/batch
Epoch: 5/20... Training Step: 17806... Training loss: 1.5117... 0.0615 sec/batch
Epoch: 5/20... Training Step: 17807... Training loss: 1.6627... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17808... Training loss: 1.6912... 0.0572 sec/batch
Epoch: 5/20... Training Step: 17809... Training loss: 1.6802... 0.0594 sec/batch
Epoch: 5/20... Training Step: 17810... Training loss: 1.5982... 0.0579 sec/batch
Epoch: 5/20... Training Step: 17811... Training loss: 1.5693... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17812... Training loss: 1.6434... 0.0586 sec/batch
Epoch: 5/20... Training Step: 17813... Training loss: 1.6791... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17814... Training loss: 1.9319... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17815... Training loss: 1.6749... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17816... Training loss: 1.6761... 0.0632 sec/batch
Epoch: 5/20... Training Step: 17817... Training loss: 1.8027... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17818... Training loss: 2.0364... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17819... Training loss: 1.7707... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17820... Training loss: 1.8930... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17821... Training loss: 1.6479... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17822... Training loss: 1.6547... 0.0644 sec/batch
Epoch: 5/20... Training Step: 17823... Training loss: 1.6942... 0.0627 sec/batch
Epoch: 5/20... Training Step: 17824... Training loss: 1.6105... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17825... Training loss: 1.7431... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17826... Training loss: 1.5383... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17827... Training loss: 1.6241... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17828... Training loss: 1.6429... 0.0540 sec/batch
Epoch: 5/20... Training Step: 17829... Training loss: 1.7243... 0.0708 sec/batch
Epoch: 5/20... Training Step: 17830... Training loss: 1.5930... 0.0621 sec/batch
Epoch: 5/20... Training Step: 17831... Training loss: 1.6682... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17832... Training loss: 1.6729... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17833... Training loss: 1.6527... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17834... Training loss: 1.6111... 0.0639 sec/batch
Epoch: 5/20... Training Step: 17835... Training loss: 1.7810... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17836... Training loss: 1.7823... 0.0643 sec/batch
Epoch: 5/20... Training Step: 17837... Training loss: 1.6780... 0.0774 sec/batch
Epoch: 5/20... Training Step: 17838... Training loss: 1.6281... 0.0601 sec/batch
Epoch: 5/20... Training Step: 17839... Training loss: 1.6474... 0.0633 sec/batch
Epoch: 5/20... Training Step: 17840... Training loss: 1.6935... 0.0652 sec/batch
Epoch: 5/20... Training Step: 17841... Training loss: 1.6327... 0.0670 sec/batch
Epoch: 5/20... Training Step: 17842... Training loss: 1.6779... 0.0732 sec/batch
Epoch: 5/20... Training Step: 17843... Training loss: 1.5539... 0.0616 sec/batch
Epoch: 5/20... Training Step: 17844... Training loss: 1.4240... 0.0613 sec/batch
Epoch: 5/20... Training Step: 17845... Training loss: 1.6415... 0.0698 sec/batch
Epoch: 5/20... Training Step: 17846... Training loss: 1.5266... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17847... Training loss: 1.6909... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17848... Training loss: 1.7152... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17849... Training loss: 1.8059... 0.0813 sec/batch
Epoch: 5/20... Training Step: 17850... Training loss: 1.6535... 0.0675 sec/batch
Epoch: 5/20... Training Step: 17851... Training loss: 1.6357... 0.0680 sec/batch
Epoch: 5/20... Training Step: 17852... Training loss: 1.6565... 0.0651 sec/batch
Epoch: 5/20... Training Step: 17853... Training loss: 1.7040... 0.0895 sec/batch
Epoch: 5/20... Training Step: 17854... Training loss: 1.7458... 0.0672 sec/batch
Epoch: 5/20... Training Step: 17855... Training loss: 1.7763... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17856... Training loss: 1.5178... 0.0829 sec/batch
Epoch: 5/20... Training Step: 17857... Training loss: 1.6712... 0.0665 sec/batch
Epoch: 5/20... Training Step: 17858... Training loss: 1.5789... 0.0629 sec/batch
Epoch: 5/20... Training Step: 17859... Training loss: 1.6802... 0.0828 sec/batch
Epoch: 5/20... Training Step: 17860... Training loss: 1.5971... 0.0602 sec/batch
Epoch: 5/20... Training Step: 17861... Training loss: 1.7265... 0.0581 sec/batch
Epoch: 5/20... Training Step: 17862... Training loss: 1.6015... 0.0687 sec/batch
Epoch: 5/20... Training Step: 17863... Training loss: 1.6126... 0.0571 sec/batch
Epoch: 5/20... Training Step: 17864... Training loss: 1.8188... 0.0554 sec/batch
Epoch: 5/20... Training Step: 17865... Training loss: 1.9190... 0.0674 sec/batch
Epoch: 5/20... Training Step: 17866... Training loss: 1.6413... 0.0778 sec/batch
Epoch: 5/20... Training Step: 17867... Training loss: 1.6024... 0.0604 sec/batch
Epoch: 5/20... Training Step: 17868... Training loss: 1.6576... 0.0645 sec/batch
Epoch: 5/20... Training Step: 17869... Training loss: 1.5979... 0.0586 sec/batch
Epoch: 5/20... Training Step: 17870... Training loss: 1.6239... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17871... Training loss: 1.5016... 0.0505 sec/batch
Epoch: 5/20... Training Step: 17872... Training loss: 1.6941... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17873... Training loss: 1.8689... 0.0647 sec/batch
Epoch: 5/20... Training Step: 17874... Training loss: 1.7353... 0.0555 sec/batch
Epoch: 5/20... Training Step: 17875... Training loss: 1.6810... 0.0638 sec/batch
Epoch: 5/20... Training Step: 17876... Training loss: 1.7220... 0.0708 sec/batch
Epoch: 5/20... Training Step: 17877... Training loss: 1.6260... 0.0646 sec/batch
Epoch: 5/20... Training Step: 17878... Training loss: 1.6478... 0.0610 sec/batch
Epoch: 5/20... Training Step: 17879... Training loss: 1.6110... 0.0557 sec/batch
Epoch: 5/20... Training Step: 17880... Training loss: 1.6555... 0.0784 sec/batch
Epoch: 5/20... Training Step: 17881... Training loss: 1.5408... 0.0627 sec/batch
Epoch: 5/20... Training Step: 17882... Training loss: 1.4523... 0.0499 sec/batch
Epoch: 5/20... Training Step: 17883... Training loss: 1.4923... 0.0567 sec/batch
Epoch: 5/20... Training Step: 17884... Training loss: 1.5583... 0.0785 sec/batch
Epoch: 5/20... Training Step: 17885... Training loss: 1.6815... 0.0726 sec/batch
Epoch: 5/20... Training Step: 17886... Training loss: 1.5046... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17887... Training loss: 1.6174... 0.0620 sec/batch
Epoch: 5/20... Training Step: 17888... Training loss: 1.5844... 0.0765 sec/batch
Epoch: 5/20... Training Step: 17889... Training loss: 1.5242... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17890... Training loss: 1.7182... 0.0534 sec/batch
Epoch: 5/20... Training Step: 17891... Training loss: 1.5486... 0.0554 sec/batch
Epoch: 5/20... Training Step: 17892... Training loss: 1.6017... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17893... Training loss: 1.6293... 0.0642 sec/batch
Epoch: 5/20... Training Step: 17894... Training loss: 1.5207... 0.0558 sec/batch
Epoch: 5/20... Training Step: 17895... Training loss: 1.6510... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17896... Training loss: 1.5695... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17897... Training loss: 1.5545... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17898... Training loss: 1.4773... 0.0598 sec/batch
Epoch: 5/20... Training Step: 17899... Training loss: 1.4873... 0.0593 sec/batch
Epoch: 5/20... Training Step: 17900... Training loss: 1.5427... 0.0553 sec/batch
Epoch: 5/20... Training Step: 17901... Training loss: 1.7307... 0.0786 sec/batch
Epoch: 5/20... Training Step: 17902... Training loss: 1.5648... 0.0580 sec/batch
Epoch: 5/20... Training Step: 17903... Training loss: 1.6918... 0.0621 sec/batch
Epoch: 5/20... Training Step: 17904... Training loss: 1.5961... 0.0689 sec/batch
Epoch: 5/20... Training Step: 17905... Training loss: 1.5919... 0.0704 sec/batch
Epoch: 5/20... Training Step: 17906... Training loss: 1.4720... 0.0612 sec/batch
Epoch: 5/20... Training Step: 17907... Training loss: 1.6614... 0.0603 sec/batch
Epoch: 5/20... Training Step: 17908... Training loss: 1.4790... 0.0613 sec/batch
Epoch: 5/20... Training Step: 17909... Training loss: 1.6961... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17910... Training loss: 1.5406... 0.0590 sec/batch
Epoch: 5/20... Training Step: 17911... Training loss: 1.6216... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17912... Training loss: 1.6916... 0.0831 sec/batch
Epoch: 5/20... Training Step: 17913... Training loss: 1.6193... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17914... Training loss: 1.6735... 0.0574 sec/batch
Epoch: 5/20... Training Step: 17915... Training loss: 1.6114... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17916... Training loss: 1.6542... 0.0670 sec/batch
Epoch: 5/20... Training Step: 17917... Training loss: 1.5030... 0.0663 sec/batch
Epoch: 5/20... Training Step: 17918... Training loss: 1.5933... 0.0536 sec/batch
Epoch: 5/20... Training Step: 17919... Training loss: 1.5733... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17920... Training loss: 1.6655... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17921... Training loss: 1.5891... 0.0736 sec/batch
Epoch: 5/20... Training Step: 17922... Training loss: 1.7235... 0.0511 sec/batch
Epoch: 5/20... Training Step: 17923... Training loss: 1.6936... 0.0548 sec/batch
Epoch: 5/20... Training Step: 17924... Training loss: 1.6580... 0.0863 sec/batch
Epoch: 5/20... Training Step: 17925... Training loss: 1.6869... 0.0649 sec/batch
Epoch: 5/20... Training Step: 17926... Training loss: 1.6610... 0.0577 sec/batch
Epoch: 5/20... Training Step: 17927... Training loss: 1.6941... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17928... Training loss: 1.5440... 0.0650 sec/batch
Epoch: 5/20... Training Step: 17929... Training loss: 1.6276... 0.0635 sec/batch
Epoch: 5/20... Training Step: 17930... Training loss: 1.7995... 0.0628 sec/batch
Epoch: 5/20... Training Step: 17931... Training loss: 1.7392... 0.0626 sec/batch
Epoch: 5/20... Training Step: 17932... Training loss: 1.6159... 0.0694 sec/batch
Epoch: 5/20... Training Step: 17933... Training loss: 1.5594... 0.0747 sec/batch
Epoch: 5/20... Training Step: 17934... Training loss: 1.5229... 0.0671 sec/batch
Epoch: 5/20... Training Step: 17935... Training loss: 1.5498... 0.0648 sec/batch
Epoch: 5/20... Training Step: 17936... Training loss: 1.6642... 0.0686 sec/batch
Epoch: 5/20... Training Step: 17937... Training loss: 1.5478... 0.0609 sec/batch
Epoch: 5/20... Training Step: 17938... Training loss: 1.5256... 0.0637 sec/batch
Epoch: 5/20... Training Step: 17939... Training loss: 1.6239... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17940... Training loss: 1.5814... 0.0665 sec/batch
Epoch: 5/20... Training Step: 17941... Training loss: 1.6575... 0.0553 sec/batch
Epoch: 5/20... Training Step: 17942... Training loss: 1.6653... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17943... Training loss: 1.6450... 0.0762 sec/batch
Epoch: 5/20... Training Step: 17944... Training loss: 1.7373... 0.0770 sec/batch
Epoch: 5/20... Training Step: 17945... Training loss: 1.5822... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17946... Training loss: 1.6583... 0.0613 sec/batch
Epoch: 5/20... Training Step: 17947... Training loss: 1.5605... 0.0687 sec/batch
Epoch: 5/20... Training Step: 17948... Training loss: 1.6866... 0.0722 sec/batch
Epoch: 5/20... Training Step: 17949... Training loss: 1.7587... 0.0664 sec/batch
Epoch: 5/20... Training Step: 17950... Training loss: 1.5702... 0.0657 sec/batch
Epoch: 5/20... Training Step: 17951... Training loss: 1.6800... 0.0705 sec/batch
Epoch: 5/20... Training Step: 17952... Training loss: 1.8500... 0.0653 sec/batch
Epoch: 5/20... Training Step: 17953... Training loss: 1.5626... 0.0596 sec/batch
Epoch: 5/20... Training Step: 17954... Training loss: 1.9298... 0.0824 sec/batch
Epoch: 5/20... Training Step: 17955... Training loss: 1.6153... 0.0628 sec/batch
Epoch: 5/20... Training Step: 17956... Training loss: 1.6351... 0.0584 sec/batch
Epoch: 5/20... Training Step: 17957... Training loss: 1.5638... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17958... Training loss: 1.7044... 0.0636 sec/batch
Epoch: 5/20... Training Step: 17959... Training loss: 1.6697... 0.0762 sec/batch
Epoch: 5/20... Training Step: 17960... Training loss: 1.6202... 0.0606 sec/batch
Epoch: 5/20... Training Step: 17961... Training loss: 1.6800... 0.0587 sec/batch
Epoch: 5/20... Training Step: 17962... Training loss: 1.6415... 0.0903 sec/batch
Epoch: 5/20... Training Step: 17963... Training loss: 1.6599... 0.0662 sec/batch
Epoch: 5/20... Training Step: 17964... Training loss: 1.6888... 0.0605 sec/batch
Epoch: 5/20... Training Step: 17965... Training loss: 1.6247... 0.0624 sec/batch
Epoch: 5/20... Training Step: 17966... Training loss: 1.6358... 0.0855 sec/batch
Epoch: 5/20... Training Step: 17967... Training loss: 1.6894... 0.0673 sec/batch
Epoch: 5/20... Training Step: 17968... Training loss: 1.8131... 0.0580 sec/batch
Epoch: 5/20... Training Step: 17969... Training loss: 1.6516... 0.0614 sec/batch
Epoch: 5/20... Training Step: 17970... Training loss: 1.6052... 0.0599 sec/batch
Epoch: 5/20... Training Step: 17971... Training loss: 1.7448... 0.0715 sec/batch
Epoch: 5/20... Training Step: 17972... Training loss: 1.5863... 0.0622 sec/batch
Epoch: 5/20... Training Step: 17973... Training loss: 1.6151... 0.0597 sec/batch
Epoch: 5/20... Training Step: 17974... Training loss: 1.5678... 0.0876 sec/batch
Epoch: 5/20... Training Step: 17975... Training loss: 1.7046... 0.0630 sec/batch
Epoch: 5/20... Training Step: 17976... Training loss: 1.6681... 0.0658 sec/batch
Epoch: 5/20... Training Step: 17977... Training loss: 1.6419... 0.0627 sec/batch
Epoch: 5/20... Training Step: 17978... Training loss: 1.7019... 0.0792 sec/batch
Epoch: 5/20... Training Step: 17979... Training loss: 1.5087... 0.0679 sec/batch
Epoch: 5/20... Training Step: 17980... Training loss: 1.6233... 0.0617 sec/batch
Epoch: 5/20... Training Step: 17981... Training loss: 1.5421... 0.0605 sec/batch
Epoch: 5/20... Training Step: 17982... Training loss: 1.6573... 0.0827 sec/batch
Epoch: 5/20... Training Step: 17983... Training loss: 1.7745... 0.0641 sec/batch
Epoch: 5/20... Training Step: 17984... Training loss: 1.7821... 0.0874 sec/batch
Epoch: 5/20... Training Step: 17985... Training loss: 1.7021... 0.0953 sec/batch
Epoch: 5/20... Training Step: 17986... Training loss: 1.6977... 0.1140 sec/batch
Epoch: 5/20... Training Step: 17987... Training loss: 1.6911... 0.0801 sec/batch
Epoch: 5/20... Training Step: 17988... Training loss: 1.7147... 0.0953 sec/batch
Epoch: 5/20... Training Step: 17989... Training loss: 1.6558... 0.0899 sec/batch
Epoch: 5/20... Training Step: 17990... Training loss: 1.6099... 0.0912 sec/batch
Epoch: 5/20... Training Step: 17991... Training loss: 1.5014... 0.0905 sec/batch
Epoch: 5/20... Training Step: 17992... Training loss: 1.6139... 0.1124 sec/batch
Epoch: 5/20... Training Step: 17993... Training loss: 1.6462... 0.0816 sec/batch
Epoch: 5/20... Training Step: 17994... Training loss: 1.5877... 0.1260 sec/batch
Epoch: 5/20... Training Step: 17995... Training loss: 1.5471... 0.0963 sec/batch
Epoch: 5/20... Training Step: 17996... Training loss: 1.6486... 0.0951 sec/batch
Epoch: 5/20... Training Step: 17997... Training loss: 1.5628... 0.1052 sec/batch
Epoch: 5/20... Training Step: 17998... Training loss: 1.6234... 0.1056 sec/batch
Epoch: 5/20... Training Step: 17999... Training loss: 1.5279... 0.0843 sec/batch
Epoch: 5/20... Training Step: 18000... Training loss: 1.4912... 0.1292 sec/batch
Epoch: 5/20... Training Step: 18001... Training loss: 1.7297... 0.0886 sec/batch
Epoch: 5/20... Training Step: 18002... Training loss: 1.7866... 0.0959 sec/batch
Epoch: 5/20... Training Step: 18003... Training loss: 1.5691... 0.0858 sec/batch
Epoch: 5/20... Training Step: 18004... Training loss: 1.6480... 0.0969 sec/batch
Epoch: 5/20... Training Step: 18005... Training loss: 1.6952... 0.1111 sec/batch
Epoch: 5/20... Training Step: 18006... Training loss: 1.6476... 0.0835 sec/batch
Epoch: 5/20... Training Step: 18007... Training loss: 1.7295... 0.1073 sec/batch
Epoch: 5/20... Training Step: 18008... Training loss: 1.5146... 0.0977 sec/batch
Epoch: 5/20... Training Step: 18009... Training loss: 1.5279... 0.0890 sec/batch
Epoch: 5/20... Training Step: 18010... Training loss: 1.7710... 0.0927 sec/batch
Epoch: 5/20... Training Step: 18011... Training loss: 1.5806... 0.1017 sec/batch
Epoch: 5/20... Training Step: 18012... Training loss: 1.5347... 0.0927 sec/batch
Epoch: 5/20... Training Step: 18013... Training loss: 1.7071... 0.0984 sec/batch
Epoch: 5/20... Training Step: 18014... Training loss: 1.6177... 0.0976 sec/batch
Epoch: 5/20... Training Step: 18015... Training loss: 1.6201... 0.0866 sec/batch
Epoch: 5/20... Training Step: 18016... Training loss: 1.6134... 0.1174 sec/batch
Epoch: 5/20... Training Step: 18017... Training loss: 1.6055... 0.0935 sec/batch
Epoch: 5/20... Training Step: 18018... Training loss: 1.6844... 0.0924 sec/batch
Epoch: 5/20... Training Step: 18019... Training loss: 1.6638... 0.1183 sec/batch
Epoch: 5/20... Training Step: 18020... Training loss: 1.6965... 0.0960 sec/batch
Epoch: 5/20... Training Step: 18021... Training loss: 1.6234... 0.0844 sec/batch
Epoch: 5/20... Training Step: 18022... Training loss: 1.7306... 0.0945 sec/batch
Epoch: 5/20... Training Step: 18023... Training loss: 1.6691... 0.0921 sec/batch
Epoch: 5/20... Training Step: 18024... Training loss: 1.6711... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18025... Training loss: 1.6666... 0.0686 sec/batch
Epoch: 5/20... Training Step: 18026... Training loss: 1.6578... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18027... Training loss: 1.5438... 0.0749 sec/batch
Epoch: 5/20... Training Step: 18028... Training loss: 1.9221... 0.0660 sec/batch
Epoch: 5/20... Training Step: 18029... Training loss: 1.6561... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18030... Training loss: 1.6481... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18031... Training loss: 1.7697... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18032... Training loss: 1.6213... 0.0792 sec/batch
Epoch: 5/20... Training Step: 18033... Training loss: 1.5330... 0.0631 sec/batch
Epoch: 5/20... Training Step: 18034... Training loss: 1.6264... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18035... Training loss: 1.6597... 0.0648 sec/batch
Epoch: 5/20... Training Step: 18036... Training loss: 1.4379... 0.0829 sec/batch
Epoch: 5/20... Training Step: 18037... Training loss: 1.7170... 0.0700 sec/batch
Epoch: 5/20... Training Step: 18038... Training loss: 1.6087... 0.0566 sec/batch
Epoch: 5/20... Training Step: 18039... Training loss: 1.7562... 0.0671 sec/batch
Epoch: 5/20... Training Step: 18040... Training loss: 1.6703... 0.0876 sec/batch
Epoch: 5/20... Training Step: 18041... Training loss: 1.6129... 0.0775 sec/batch
Epoch: 5/20... Training Step: 18042... Training loss: 1.6132... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18043... Training loss: 1.4987... 0.0653 sec/batch
Epoch: 5/20... Training Step: 18044... Training loss: 1.6283... 0.0714 sec/batch
Epoch: 5/20... Training Step: 18045... Training loss: 1.6707... 0.0747 sec/batch
Epoch: 5/20... Training Step: 18046... Training loss: 1.5852... 0.0770 sec/batch
Epoch: 5/20... Training Step: 18047... Training loss: 1.6087... 0.0716 sec/batch
Epoch: 5/20... Training Step: 18048... Training loss: 1.5958... 0.0622 sec/batch
Epoch: 5/20... Training Step: 18049... Training loss: 1.6928... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18050... Training loss: 1.5901... 0.0853 sec/batch
Epoch: 5/20... Training Step: 18051... Training loss: 1.5271... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18052... Training loss: 1.6484... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18053... Training loss: 1.7730... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18054... Training loss: 1.5958... 0.0951 sec/batch
Epoch: 5/20... Training Step: 18055... Training loss: 1.7093... 0.0667 sec/batch
Epoch: 5/20... Training Step: 18056... Training loss: 1.6732... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18057... Training loss: 1.6293... 0.0620 sec/batch
Epoch: 5/20... Training Step: 18058... Training loss: 1.7023... 0.0702 sec/batch
Epoch: 5/20... Training Step: 18059... Training loss: 1.6536... 0.0723 sec/batch
Epoch: 5/20... Training Step: 18060... Training loss: 1.6832... 0.0658 sec/batch
Epoch: 5/20... Training Step: 18061... Training loss: 1.6260... 0.0754 sec/batch
Epoch: 5/20... Training Step: 18062... Training loss: 1.8215... 0.1063 sec/batch
Epoch: 5/20... Training Step: 18063... Training loss: 1.5523... 0.0672 sec/batch
Epoch: 5/20... Training Step: 18064... Training loss: 1.6281... 0.0811 sec/batch
Epoch: 5/20... Training Step: 18065... Training loss: 1.7262... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18066... Training loss: 1.6618... 0.0686 sec/batch
Epoch: 5/20... Training Step: 18067... Training loss: 1.7162... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18068... Training loss: 1.5347... 0.0906 sec/batch
Epoch: 5/20... Training Step: 18069... Training loss: 1.6994... 0.0685 sec/batch
Epoch: 5/20... Training Step: 18070... Training loss: 1.6693... 0.0801 sec/batch
Epoch: 5/20... Training Step: 18071... Training loss: 1.6919... 0.0853 sec/batch
Epoch: 5/20... Training Step: 18072... Training loss: 1.6190... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18073... Training loss: 1.5924... 0.0635 sec/batch
Epoch: 5/20... Training Step: 18074... Training loss: 1.5354... 0.0867 sec/batch
Epoch: 5/20... Training Step: 18075... Training loss: 1.6526... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18076... Training loss: 1.7603... 0.0643 sec/batch
Epoch: 5/20... Training Step: 18077... Training loss: 1.6072... 0.0937 sec/batch
Epoch: 5/20... Training Step: 18078... Training loss: 1.6965... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18079... Training loss: 1.6835... 0.0627 sec/batch
Epoch: 5/20... Training Step: 18080... Training loss: 1.6289... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18081... Training loss: 1.7505... 0.0642 sec/batch
Epoch: 5/20... Training Step: 18082... Training loss: 1.7722... 0.0671 sec/batch
Epoch: 5/20... Training Step: 18083... Training loss: 1.8009... 0.0715 sec/batch
Epoch: 5/20... Training Step: 18084... Training loss: 1.5329... 0.0605 sec/batch
Epoch: 5/20... Training Step: 18085... Training loss: 1.6369... 0.0625 sec/batch
Epoch: 5/20... Training Step: 18086... Training loss: 1.6544... 0.0725 sec/batch
Epoch: 5/20... Training Step: 18087... Training loss: 1.7367... 0.0723 sec/batch
Epoch: 5/20... Training Step: 18088... Training loss: 1.7050... 0.0580 sec/batch
Epoch: 5/20... Training Step: 18089... Training loss: 1.6269... 0.0587 sec/batch
Epoch: 5/20... Training Step: 18090... Training loss: 1.7081... 0.0835 sec/batch
Epoch: 5/20... Training Step: 18091... Training loss: 1.7060... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18092... Training loss: 1.5377... 0.0580 sec/batch
Epoch: 5/20... Training Step: 18093... Training loss: 1.5630... 0.0536 sec/batch
Epoch: 5/20... Training Step: 18094... Training loss: 1.7042... 0.0619 sec/batch
Epoch: 5/20... Training Step: 18095... Training loss: 1.7743... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18096... Training loss: 1.9795... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18097... Training loss: 1.6603... 0.0667 sec/batch
Epoch: 5/20... Training Step: 18098... Training loss: 1.6066... 0.0713 sec/batch
Epoch: 5/20... Training Step: 18099... Training loss: 1.6965... 0.0725 sec/batch
Epoch: 5/20... Training Step: 18100... Training loss: 1.6823... 0.0681 sec/batch
Epoch: 5/20... Training Step: 18101... Training loss: 1.5990... 0.0625 sec/batch
Epoch: 5/20... Training Step: 18102... Training loss: 1.6988... 0.0755 sec/batch
Epoch: 5/20... Training Step: 18103... Training loss: 1.5608... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18104... Training loss: 1.5163... 0.0557 sec/batch
Epoch: 5/20... Training Step: 18105... Training loss: 1.8336... 0.0679 sec/batch
Epoch: 5/20... Training Step: 18106... Training loss: 1.4874... 0.0607 sec/batch
Epoch: 5/20... Training Step: 18107... Training loss: 1.5936... 0.0593 sec/batch
Epoch: 5/20... Training Step: 18108... Training loss: 1.6849... 0.0562 sec/batch
Epoch: 5/20... Training Step: 18109... Training loss: 1.6281... 0.0595 sec/batch
Epoch: 5/20... Training Step: 18110... Training loss: 1.6521... 0.0715 sec/batch
Epoch: 5/20... Training Step: 18111... Training loss: 1.6830... 0.0572 sec/batch
Epoch: 5/20... Training Step: 18112... Training loss: 1.6105... 0.0648 sec/batch
Epoch: 5/20... Training Step: 18113... Training loss: 1.6170... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18114... Training loss: 1.5972... 0.0665 sec/batch
Epoch: 5/20... Training Step: 18115... Training loss: 1.5317... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18116... Training loss: 1.4826... 0.0642 sec/batch
Epoch: 5/20... Training Step: 18117... Training loss: 1.6235... 0.0783 sec/batch
Epoch: 5/20... Training Step: 18118... Training loss: 1.5205... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18119... Training loss: 1.6823... 0.0575 sec/batch
Epoch: 5/20... Training Step: 18120... Training loss: 1.8888... 0.0560 sec/batch
Epoch: 5/20... Training Step: 18121... Training loss: 1.8043... 0.0660 sec/batch
Epoch: 5/20... Training Step: 18122... Training loss: 1.6846... 0.0736 sec/batch
Epoch: 5/20... Training Step: 18123... Training loss: 1.7435... 0.0603 sec/batch
Epoch: 5/20... Training Step: 18124... Training loss: 1.8013... 0.0565 sec/batch
Epoch: 5/20... Training Step: 18125... Training loss: 2.0088... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18126... Training loss: 1.7939... 0.0768 sec/batch
Epoch: 5/20... Training Step: 18127... Training loss: 1.5817... 0.0591 sec/batch
Epoch: 5/20... Training Step: 18128... Training loss: 1.6791... 0.0610 sec/batch
Epoch: 5/20... Training Step: 18129... Training loss: 1.5430... 0.0752 sec/batch
Epoch: 5/20... Training Step: 18130... Training loss: 1.7154... 0.0739 sec/batch
Epoch: 5/20... Training Step: 18131... Training loss: 1.5369... 0.0651 sec/batch
Epoch: 5/20... Training Step: 18132... Training loss: 1.7404... 0.0626 sec/batch
Epoch: 5/20... Training Step: 18133... Training loss: 1.5198... 0.0923 sec/batch
Epoch: 5/20... Training Step: 18134... Training loss: 1.6599... 0.0628 sec/batch
Epoch: 5/20... Training Step: 18135... Training loss: 1.6472... 0.0544 sec/batch
Epoch: 5/20... Training Step: 18136... Training loss: 1.5846... 0.0720 sec/batch
Epoch: 5/20... Training Step: 18137... Training loss: 1.6867... 0.0605 sec/batch
Epoch: 5/20... Training Step: 18138... Training loss: 1.6142... 0.0626 sec/batch
Epoch: 5/20... Training Step: 18139... Training loss: 1.5546... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18140... Training loss: 1.8603... 0.0616 sec/batch
Epoch: 5/20... Training Step: 18141... Training loss: 1.7990... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18142... Training loss: 1.5266... 0.0585 sec/batch
Epoch: 5/20... Training Step: 18143... Training loss: 1.5680... 0.0762 sec/batch
Epoch: 5/20... Training Step: 18144... Training loss: 1.6483... 0.0675 sec/batch
Epoch: 5/20... Training Step: 18145... Training loss: 1.6273... 0.0588 sec/batch
Epoch: 5/20... Training Step: 18146... Training loss: 1.6723... 0.0592 sec/batch
Epoch: 5/20... Training Step: 18147... Training loss: 1.6070... 0.0754 sec/batch
Epoch: 5/20... Training Step: 18148... Training loss: 1.6939... 0.0705 sec/batch
Epoch: 5/20... Training Step: 18149... Training loss: 1.7089... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18150... Training loss: 1.5461... 0.0535 sec/batch
Epoch: 5/20... Training Step: 18151... Training loss: 1.5835... 0.0625 sec/batch
Epoch: 5/20... Training Step: 18152... Training loss: 1.7389... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18153... Training loss: 1.5634... 0.0546 sec/batch
Epoch: 5/20... Training Step: 18154... Training loss: 1.5519... 0.0561 sec/batch
Epoch: 5/20... Training Step: 18155... Training loss: 1.4806... 0.0633 sec/batch
Epoch: 5/20... Training Step: 18156... Training loss: 1.5743... 0.0729 sec/batch
Epoch: 5/20... Training Step: 18157... Training loss: 1.6213... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18158... Training loss: 1.6650... 0.0642 sec/batch
Epoch: 5/20... Training Step: 18159... Training loss: 1.7197... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18160... Training loss: 1.5569... 0.0649 sec/batch
Epoch: 5/20... Training Step: 18161... Training loss: 1.6242... 0.0604 sec/batch
Epoch: 5/20... Training Step: 18162... Training loss: 1.5780... 0.0535 sec/batch
Epoch: 5/20... Training Step: 18163... Training loss: 1.6479... 0.0557 sec/batch
Epoch: 5/20... Training Step: 18164... Training loss: 1.5577... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18165... Training loss: 1.5686... 0.0587 sec/batch
Epoch: 5/20... Training Step: 18166... Training loss: 1.5375... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18167... Training loss: 1.5471... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18168... Training loss: 1.5743... 0.0673 sec/batch
Epoch: 5/20... Training Step: 18169... Training loss: 1.6405... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18170... Training loss: 1.6453... 0.0524 sec/batch
Epoch: 5/20... Training Step: 18171... Training loss: 1.5239... 0.0662 sec/batch
Epoch: 5/20... Training Step: 18172... Training loss: 1.6308... 0.0651 sec/batch
Epoch: 5/20... Training Step: 18173... Training loss: 1.5506... 0.0588 sec/batch
Epoch: 5/20... Training Step: 18174... Training loss: 1.6564... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18175... Training loss: 1.6311... 0.0593 sec/batch
Epoch: 5/20... Training Step: 18176... Training loss: 1.5582... 0.0699 sec/batch
Epoch: 5/20... Training Step: 18177... Training loss: 1.6578... 0.0554 sec/batch
Epoch: 5/20... Training Step: 18178... Training loss: 1.5465... 0.0583 sec/batch
Epoch: 5/20... Training Step: 18179... Training loss: 1.8081... 0.0610 sec/batch
Epoch: 5/20... Training Step: 18180... Training loss: 1.5440... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18181... Training loss: 1.7547... 0.0591 sec/batch
Epoch: 5/20... Training Step: 18182... Training loss: 1.7276... 0.0653 sec/batch
Epoch: 5/20... Training Step: 18183... Training loss: 1.5843... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18184... Training loss: 1.5469... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18185... Training loss: 1.6030... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18186... Training loss: 1.7344... 0.0585 sec/batch
Epoch: 5/20... Training Step: 18187... Training loss: 1.6205... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18188... Training loss: 1.7354... 0.0733 sec/batch
Epoch: 5/20... Training Step: 18189... Training loss: 1.5536... 0.0566 sec/batch
Epoch: 5/20... Training Step: 18190... Training loss: 1.7132... 0.0576 sec/batch
Epoch: 5/20... Training Step: 18191... Training loss: 1.5503... 0.0679 sec/batch
Epoch: 5/20... Training Step: 18192... Training loss: 1.5675... 0.0597 sec/batch
Epoch: 5/20... Training Step: 18193... Training loss: 1.6465... 0.0522 sec/batch
Epoch: 5/20... Training Step: 18194... Training loss: 1.6919... 0.0516 sec/batch
Epoch: 5/20... Training Step: 18195... Training loss: 1.5983... 0.0540 sec/batch
Epoch: 5/20... Training Step: 18196... Training loss: 1.6219... 0.0728 sec/batch
Epoch: 5/20... Training Step: 18197... Training loss: 1.4780... 0.0675 sec/batch
Epoch: 5/20... Training Step: 18198... Training loss: 1.6470... 0.0619 sec/batch
Epoch: 5/20... Training Step: 18199... Training loss: 1.7012... 0.0749 sec/batch
Epoch: 5/20... Training Step: 18200... Training loss: 1.6580... 0.0642 sec/batch
Epoch: 5/20... Training Step: 18201... Training loss: 1.5464... 0.0740 sec/batch
Epoch: 5/20... Training Step: 18202... Training loss: 1.5145... 0.0518 sec/batch
Epoch: 5/20... Training Step: 18203... Training loss: 1.5424... 0.0593 sec/batch
Epoch: 5/20... Training Step: 18204... Training loss: 1.7184... 0.0595 sec/batch
Epoch: 5/20... Training Step: 18205... Training loss: 1.8053... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18206... Training loss: 1.7912... 0.0768 sec/batch
Epoch: 5/20... Training Step: 18207... Training loss: 1.7871... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18208... Training loss: 1.6620... 0.0612 sec/batch
Epoch: 5/20... Training Step: 18209... Training loss: 1.7236... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18210... Training loss: 1.5950... 0.0612 sec/batch
Epoch: 5/20... Training Step: 18211... Training loss: 1.5235... 0.0565 sec/batch
Epoch: 5/20... Training Step: 18212... Training loss: 1.7075... 0.0685 sec/batch
Epoch: 5/20... Training Step: 18213... Training loss: 1.5450... 0.0724 sec/batch
Epoch: 5/20... Training Step: 18214... Training loss: 1.5453... 0.0595 sec/batch
Epoch: 5/20... Training Step: 18215... Training loss: 1.6068... 0.0609 sec/batch
Epoch: 5/20... Training Step: 18216... Training loss: 1.6523... 0.0812 sec/batch
Epoch: 5/20... Training Step: 18217... Training loss: 1.8136... 0.0601 sec/batch
Epoch: 5/20... Training Step: 18218... Training loss: 1.7696... 0.0602 sec/batch
Epoch: 5/20... Training Step: 18219... Training loss: 1.7145... 0.0574 sec/batch
Epoch: 5/20... Training Step: 18220... Training loss: 1.7711... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18221... Training loss: 1.6271... 0.0724 sec/batch
Epoch: 5/20... Training Step: 18222... Training loss: 1.5759... 0.0589 sec/batch
Epoch: 5/20... Training Step: 18223... Training loss: 1.6497... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18224... Training loss: 1.5591... 0.0651 sec/batch
Epoch: 5/20... Training Step: 18225... Training loss: 1.5950... 0.0692 sec/batch
Epoch: 5/20... Training Step: 18226... Training loss: 1.4599... 0.0652 sec/batch
Epoch: 5/20... Training Step: 18227... Training loss: 1.6543... 0.0555 sec/batch
Epoch: 5/20... Training Step: 18228... Training loss: 1.5828... 0.0658 sec/batch
Epoch: 5/20... Training Step: 18229... Training loss: 1.6766... 0.0791 sec/batch
Epoch: 5/20... Training Step: 18230... Training loss: 1.7518... 0.0643 sec/batch
Epoch: 5/20... Training Step: 18231... Training loss: 1.8700... 0.0563 sec/batch
Epoch: 5/20... Training Step: 18232... Training loss: 1.7179... 0.0776 sec/batch
Epoch: 5/20... Training Step: 18233... Training loss: 1.8008... 0.0650 sec/batch
Epoch: 5/20... Training Step: 18234... Training loss: 1.6756... 0.0580 sec/batch
Epoch: 5/20... Training Step: 18235... Training loss: 1.6217... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18236... Training loss: 1.6777... 0.0887 sec/batch
Epoch: 5/20... Training Step: 18237... Training loss: 1.6504... 0.0704 sec/batch
Epoch: 5/20... Training Step: 18238... Training loss: 1.6951... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18239... Training loss: 1.5279... 0.0655 sec/batch
Epoch: 5/20... Training Step: 18240... Training loss: 1.8013... 0.0686 sec/batch
Epoch: 5/20... Training Step: 18241... Training loss: 1.6576... 0.0728 sec/batch
Epoch: 5/20... Training Step: 18242... Training loss: 1.7957... 0.0614 sec/batch
Epoch: 5/20... Training Step: 18243... Training loss: 1.6315... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18244... Training loss: 1.7423... 0.0683 sec/batch
Epoch: 5/20... Training Step: 18245... Training loss: 1.6017... 0.0662 sec/batch
Epoch: 5/20... Training Step: 18246... Training loss: 1.5896... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18247... Training loss: 1.5609... 0.0721 sec/batch
Epoch: 5/20... Training Step: 18248... Training loss: 1.5790... 0.0898 sec/batch
Epoch: 5/20... Training Step: 18249... Training loss: 1.7057... 0.0878 sec/batch
Epoch: 5/20... Training Step: 18250... Training loss: 1.7086... 0.0572 sec/batch
Epoch: 5/20... Training Step: 18251... Training loss: 1.7334... 0.0577 sec/batch
Epoch: 5/20... Training Step: 18252... Training loss: 1.6249... 0.0741 sec/batch
Epoch: 5/20... Training Step: 18253... Training loss: 1.6001... 0.0711 sec/batch
Epoch: 5/20... Training Step: 18254... Training loss: 1.6071... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18255... Training loss: 1.7397... 0.0857 sec/batch
Epoch: 5/20... Training Step: 18256... Training loss: 1.7397... 0.0638 sec/batch
Epoch: 5/20... Training Step: 18257... Training loss: 1.7187... 0.0674 sec/batch
Epoch: 5/20... Training Step: 18258... Training loss: 1.6613... 0.0786 sec/batch
Epoch: 5/20... Training Step: 18259... Training loss: 1.6342... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18260... Training loss: 1.6764... 0.0582 sec/batch
Epoch: 5/20... Training Step: 18261... Training loss: 1.5723... 0.0736 sec/batch
Epoch: 5/20... Training Step: 18262... Training loss: 1.6773... 0.0969 sec/batch
Epoch: 5/20... Training Step: 18263... Training loss: 1.6978... 0.0840 sec/batch
Epoch: 5/20... Training Step: 18264... Training loss: 1.6594... 0.0865 sec/batch
Epoch: 5/20... Training Step: 18265... Training loss: 1.6461... 0.0996 sec/batch
Epoch: 5/20... Training Step: 18266... Training loss: 1.5802... 0.1023 sec/batch
Epoch: 5/20... Training Step: 18267... Training loss: 1.4946... 0.1194 sec/batch
Epoch: 5/20... Training Step: 18268... Training loss: 1.7405... 0.0910 sec/batch
Epoch: 5/20... Training Step: 18269... Training loss: 1.6908... 0.0901 sec/batch
Epoch: 5/20... Training Step: 18270... Training loss: 1.7235... 0.1179 sec/batch
Epoch: 5/20... Training Step: 18271... Training loss: 1.7123... 0.0948 sec/batch
Epoch: 5/20... Training Step: 18272... Training loss: 1.7053... 0.0913 sec/batch
Epoch: 5/20... Training Step: 18273... Training loss: 1.8451... 0.1158 sec/batch
Epoch: 5/20... Training Step: 18274... Training loss: 1.7994... 0.1079 sec/batch
Epoch: 5/20... Training Step: 18275... Training loss: 1.6455... 0.0877 sec/batch
Epoch: 5/20... Training Step: 18276... Training loss: 1.6397... 0.1004 sec/batch
Epoch: 5/20... Training Step: 18277... Training loss: 1.6784... 0.1022 sec/batch
Epoch: 5/20... Training Step: 18278... Training loss: 1.5963... 0.0815 sec/batch
Epoch: 5/20... Training Step: 18279... Training loss: 1.7056... 0.0861 sec/batch
Epoch: 5/20... Training Step: 18280... Training loss: 1.7316... 0.1122 sec/batch
Epoch: 5/20... Training Step: 18281... Training loss: 1.6000... 0.0852 sec/batch
Epoch: 5/20... Training Step: 18282... Training loss: 1.5706... 0.1104 sec/batch
Epoch: 5/20... Training Step: 18283... Training loss: 1.5738... 0.0935 sec/batch
Epoch: 5/20... Training Step: 18284... Training loss: 1.6128... 0.0808 sec/batch
Epoch: 5/20... Training Step: 18285... Training loss: 1.5137... 0.0956 sec/batch
Epoch: 5/20... Training Step: 18286... Training loss: 1.6985... 0.1046 sec/batch
Epoch: 5/20... Training Step: 18287... Training loss: 1.6038... 0.0814 sec/batch
Epoch: 5/20... Training Step: 18288... Training loss: 1.5513... 0.1257 sec/batch
Epoch: 5/20... Training Step: 18289... Training loss: 1.5111... 0.1027 sec/batch
Epoch: 5/20... Training Step: 18290... Training loss: 1.5734... 0.0822 sec/batch
Epoch: 5/20... Training Step: 18291... Training loss: 1.6414... 0.1099 sec/batch
Epoch: 5/20... Training Step: 18292... Training loss: 1.6150... 0.0950 sec/batch
Epoch: 5/20... Training Step: 18293... Training loss: 1.6975... 0.0791 sec/batch
Epoch: 5/20... Training Step: 18294... Training loss: 1.5708... 0.1228 sec/batch
Epoch: 5/20... Training Step: 18295... Training loss: 1.6157... 0.0956 sec/batch
Epoch: 5/20... Training Step: 18296... Training loss: 1.5350... 0.0829 sec/batch
Epoch: 5/20... Training Step: 18297... Training loss: 1.5386... 0.0980 sec/batch
Epoch: 5/20... Training Step: 18298... Training loss: 1.6330... 0.0967 sec/batch
Epoch: 5/20... Training Step: 18299... Training loss: 1.5816... 0.0883 sec/batch
Epoch: 5/20... Training Step: 18300... Training loss: 1.8517... 0.0846 sec/batch
Epoch: 5/20... Training Step: 18301... Training loss: 1.7317... 0.1120 sec/batch
Epoch: 5/20... Training Step: 18302... Training loss: 1.6530... 0.0933 sec/batch
Epoch: 5/20... Training Step: 18303... Training loss: 1.6871... 0.0922 sec/batch
Epoch: 5/20... Training Step: 18304... Training loss: 1.6331... 0.0610 sec/batch
Epoch: 5/20... Training Step: 18305... Training loss: 1.5157... 0.0650 sec/batch
Epoch: 5/20... Training Step: 18306... Training loss: 1.6964... 0.0723 sec/batch
Epoch: 5/20... Training Step: 18307... Training loss: 1.6337... 0.0751 sec/batch
Epoch: 5/20... Training Step: 18308... Training loss: 1.7503... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18309... Training loss: 1.7075... 0.0858 sec/batch
Epoch: 5/20... Training Step: 18310... Training loss: 1.6400... 0.0820 sec/batch
Epoch: 5/20... Training Step: 18311... Training loss: 1.6878... 0.0796 sec/batch
Epoch: 5/20... Training Step: 18312... Training loss: 1.7800... 0.0839 sec/batch
Epoch: 5/20... Training Step: 18313... Training loss: 1.7082... 0.0643 sec/batch
Epoch: 5/20... Training Step: 18314... Training loss: 1.6097... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18315... Training loss: 1.5627... 0.0867 sec/batch
Epoch: 5/20... Training Step: 18316... Training loss: 1.6972... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18317... Training loss: 1.5961... 0.0669 sec/batch
Epoch: 5/20... Training Step: 18318... Training loss: 1.7033... 0.0848 sec/batch
Epoch: 5/20... Training Step: 18319... Training loss: 1.7009... 0.0590 sec/batch
Epoch: 5/20... Training Step: 18320... Training loss: 1.6229... 0.0704 sec/batch
Epoch: 5/20... Training Step: 18321... Training loss: 1.6514... 0.0919 sec/batch
Epoch: 5/20... Training Step: 18322... Training loss: 1.6561... 0.0690 sec/batch
Epoch: 5/20... Training Step: 18323... Training loss: 1.5973... 0.0699 sec/batch
Epoch: 5/20... Training Step: 18324... Training loss: 1.7473... 0.0855 sec/batch
Epoch: 5/20... Training Step: 18325... Training loss: 1.6376... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18326... Training loss: 1.6136... 0.0643 sec/batch
Epoch: 5/20... Training Step: 18327... Training loss: 1.7491... 0.0674 sec/batch
Epoch: 5/20... Training Step: 18328... Training loss: 1.5033... 0.0715 sec/batch
Epoch: 5/20... Training Step: 18329... Training loss: 1.6428... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18330... Training loss: 1.7699... 0.0936 sec/batch
Epoch: 5/20... Training Step: 18331... Training loss: 1.7662... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18332... Training loss: 1.6668... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18333... Training loss: 1.6916... 0.0745 sec/batch
Epoch: 5/20... Training Step: 18334... Training loss: 1.5885... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18335... Training loss: 1.5884... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18336... Training loss: 1.6994... 0.0874 sec/batch
Epoch: 5/20... Training Step: 18337... Training loss: 1.7976... 0.0718 sec/batch
Epoch: 5/20... Training Step: 18338... Training loss: 1.7579... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18339... Training loss: 1.7686... 0.0710 sec/batch
Epoch: 5/20... Training Step: 18340... Training loss: 1.7094... 0.0779 sec/batch
Epoch: 5/20... Training Step: 18341... Training loss: 1.5180... 0.0705 sec/batch
Epoch: 5/20... Training Step: 18342... Training loss: 1.6934... 0.0797 sec/batch
Epoch: 5/20... Training Step: 18343... Training loss: 1.5137... 0.0687 sec/batch
Epoch: 5/20... Training Step: 18344... Training loss: 1.4910... 0.0623 sec/batch
Epoch: 5/20... Training Step: 18345... Training loss: 1.7090... 0.0856 sec/batch
Epoch: 5/20... Training Step: 18346... Training loss: 1.6205... 0.0745 sec/batch
Epoch: 5/20... Training Step: 18347... Training loss: 1.7102... 0.0711 sec/batch
Epoch: 5/20... Training Step: 18348... Training loss: 1.6578... 0.0933 sec/batch
Epoch: 5/20... Training Step: 18349... Training loss: 1.6138... 0.0663 sec/batch
Epoch: 5/20... Training Step: 18350... Training loss: 1.6765... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18351... Training loss: 1.6048... 0.0814 sec/batch
Epoch: 5/20... Training Step: 18352... Training loss: 1.5677... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18353... Training loss: 1.5421... 0.0739 sec/batch
Epoch: 5/20... Training Step: 18354... Training loss: 1.5658... 0.0769 sec/batch
Epoch: 5/20... Training Step: 18355... Training loss: 1.4933... 0.0675 sec/batch
Epoch: 5/20... Training Step: 18356... Training loss: 1.5650... 0.0667 sec/batch
Epoch: 5/20... Training Step: 18357... Training loss: 1.5381... 0.0733 sec/batch
Epoch: 5/20... Training Step: 18358... Training loss: 1.6101... 0.0638 sec/batch
Epoch: 5/20... Training Step: 18359... Training loss: 1.5943... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18360... Training loss: 1.7007... 0.0697 sec/batch
Epoch: 5/20... Training Step: 18361... Training loss: 1.5167... 0.0670 sec/batch
Epoch: 5/20... Training Step: 18362... Training loss: 1.5708... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18363... Training loss: 1.8339... 0.0646 sec/batch
Epoch: 5/20... Training Step: 18364... Training loss: 1.6099... 0.0708 sec/batch
Epoch: 5/20... Training Step: 18365... Training loss: 1.6445... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18366... Training loss: 1.5982... 0.0555 sec/batch
Epoch: 5/20... Training Step: 18367... Training loss: 1.5612... 0.0562 sec/batch
Epoch: 5/20... Training Step: 18368... Training loss: 1.5451... 0.0645 sec/batch
Epoch: 5/20... Training Step: 18369... Training loss: 1.6921... 0.0729 sec/batch
Epoch: 5/20... Training Step: 18370... Training loss: 1.6880... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18371... Training loss: 1.7511... 0.0535 sec/batch
Epoch: 5/20... Training Step: 18372... Training loss: 1.6494... 0.0705 sec/batch
Epoch: 5/20... Training Step: 18373... Training loss: 1.6104... 0.0730 sec/batch
Epoch: 5/20... Training Step: 18374... Training loss: 1.6298... 0.0571 sec/batch
Epoch: 5/20... Training Step: 18375... Training loss: 1.5926... 0.0589 sec/batch
Epoch: 5/20... Training Step: 18376... Training loss: 1.5845... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18377... Training loss: 1.5616... 0.0728 sec/batch
Epoch: 5/20... Training Step: 18378... Training loss: 1.6179... 0.0563 sec/batch
Epoch: 5/20... Training Step: 18379... Training loss: 1.6989... 0.0650 sec/batch
Epoch: 5/20... Training Step: 18380... Training loss: 1.5566... 0.0827 sec/batch
Epoch: 5/20... Training Step: 18381... Training loss: 1.6031... 0.0719 sec/batch
Epoch: 5/20... Training Step: 18382... Training loss: 1.5878... 0.0685 sec/batch
Epoch: 5/20... Training Step: 18383... Training loss: 1.6097... 0.0603 sec/batch
Epoch: 5/20... Training Step: 18384... Training loss: 1.5675... 0.0728 sec/batch
Epoch: 5/20... Training Step: 18385... Training loss: 1.7409... 0.0730 sec/batch
Epoch: 5/20... Training Step: 18386... Training loss: 1.6879... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18387... Training loss: 1.5468... 0.0912 sec/batch
Epoch: 5/20... Training Step: 18388... Training loss: 1.6068... 0.0616 sec/batch
Epoch: 5/20... Training Step: 18389... Training loss: 1.7400... 0.0590 sec/batch
Epoch: 5/20... Training Step: 18390... Training loss: 1.6857... 0.0742 sec/batch
Epoch: 5/20... Training Step: 18391... Training loss: 1.6585... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18392... Training loss: 1.5168... 0.0608 sec/batch
Epoch: 5/20... Training Step: 18393... Training loss: 1.5088... 0.0755 sec/batch
Epoch: 5/20... Training Step: 18394... Training loss: 1.6747... 0.0665 sec/batch
Epoch: 5/20... Training Step: 18395... Training loss: 1.5167... 0.0627 sec/batch
Epoch: 5/20... Training Step: 18396... Training loss: 1.7200... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18397... Training loss: 1.6585... 0.0840 sec/batch
Epoch: 5/20... Training Step: 18398... Training loss: 1.7319... 0.0617 sec/batch
Epoch: 5/20... Training Step: 18399... Training loss: 1.6824... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18400... Training loss: 1.7669... 0.0650 sec/batch
Epoch: 5/20... Training Step: 18401... Training loss: 1.6152... 0.0585 sec/batch
Epoch: 5/20... Training Step: 18402... Training loss: 1.6208... 0.0613 sec/batch
Epoch: 5/20... Training Step: 18403... Training loss: 1.6568... 0.0635 sec/batch
Epoch: 5/20... Training Step: 18404... Training loss: 1.7192... 0.0582 sec/batch
Epoch: 5/20... Training Step: 18405... Training loss: 1.7353... 0.0617 sec/batch
Epoch: 5/20... Training Step: 18406... Training loss: 1.4928... 0.0779 sec/batch
Epoch: 5/20... Training Step: 18407... Training loss: 1.8127... 0.0607 sec/batch
Epoch: 5/20... Training Step: 18408... Training loss: 1.6502... 0.0668 sec/batch
Epoch: 5/20... Training Step: 18409... Training loss: 1.7047... 0.0724 sec/batch
Epoch: 5/20... Training Step: 18410... Training loss: 1.5381... 0.0651 sec/batch
Epoch: 5/20... Training Step: 18411... Training loss: 1.5663... 0.0689 sec/batch
Epoch: 5/20... Training Step: 18412... Training loss: 1.5982... 0.0734 sec/batch
Epoch: 5/20... Training Step: 18413... Training loss: 1.7559... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18414... Training loss: 1.5825... 0.0712 sec/batch
Epoch: 5/20... Training Step: 18415... Training loss: 1.6400... 0.0670 sec/batch
Epoch: 5/20... Training Step: 18416... Training loss: 1.6671... 0.0622 sec/batch
Epoch: 5/20... Training Step: 18417... Training loss: 1.6697... 0.0513 sec/batch
Epoch: 5/20... Training Step: 18418... Training loss: 1.6465... 0.0736 sec/batch
Epoch: 5/20... Training Step: 18419... Training loss: 1.7374... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18420... Training loss: 1.7062... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18421... Training loss: 1.6497... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18422... Training loss: 1.6747... 0.0682 sec/batch
Epoch: 5/20... Training Step: 18423... Training loss: 1.5990... 0.0674 sec/batch
Epoch: 5/20... Training Step: 18424... Training loss: 1.6767... 0.0578 sec/batch
Epoch: 5/20... Training Step: 18425... Training loss: 1.6048... 0.0614 sec/batch
Epoch: 5/20... Training Step: 18426... Training loss: 1.6271... 0.0761 sec/batch
Epoch: 5/20... Training Step: 18427... Training loss: 1.6958... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18428... Training loss: 1.6371... 0.0581 sec/batch
Epoch: 5/20... Training Step: 18429... Training loss: 1.6774... 0.0546 sec/batch
Epoch: 5/20... Training Step: 18430... Training loss: 1.6680... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18431... Training loss: 1.6104... 0.0722 sec/batch
Epoch: 5/20... Training Step: 18432... Training loss: 1.7889... 0.0586 sec/batch
Epoch: 5/20... Training Step: 18433... Training loss: 1.5449... 0.0531 sec/batch
Epoch: 5/20... Training Step: 18434... Training loss: 1.5128... 0.0645 sec/batch
Epoch: 5/20... Training Step: 18435... Training loss: 1.6741... 0.0616 sec/batch
Epoch: 5/20... Training Step: 18436... Training loss: 1.6087... 0.0567 sec/batch
Epoch: 5/20... Training Step: 18437... Training loss: 1.6509... 0.0522 sec/batch
Epoch: 5/20... Training Step: 18438... Training loss: 1.6231... 0.0518 sec/batch
Epoch: 5/20... Training Step: 18439... Training loss: 1.6928... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18440... Training loss: 1.5810... 0.0603 sec/batch
Epoch: 5/20... Training Step: 18441... Training loss: 1.4903... 0.0571 sec/batch
Epoch: 5/20... Training Step: 18442... Training loss: 1.5906... 0.0826 sec/batch
Epoch: 5/20... Training Step: 18443... Training loss: 1.5964... 0.0550 sec/batch
Epoch: 5/20... Training Step: 18444... Training loss: 1.5267... 0.0633 sec/batch
Epoch: 5/20... Training Step: 18445... Training loss: 1.4583... 0.0494 sec/batch
Epoch: 5/20... Training Step: 18446... Training loss: 1.6728... 0.0689 sec/batch
Epoch: 5/20... Training Step: 18447... Training loss: 1.5148... 0.0729 sec/batch
Epoch: 5/20... Training Step: 18448... Training loss: 1.7135... 0.0577 sec/batch
Epoch: 5/20... Training Step: 18449... Training loss: 1.5903... 0.0553 sec/batch
Epoch: 5/20... Training Step: 18450... Training loss: 1.6192... 0.0545 sec/batch
Epoch: 5/20... Training Step: 18451... Training loss: 1.6147... 0.0779 sec/batch
Epoch: 5/20... Training Step: 18452... Training loss: 1.6033... 0.0622 sec/batch
Epoch: 5/20... Training Step: 18453... Training loss: 1.5944... 0.0597 sec/batch
Epoch: 5/20... Training Step: 18454... Training loss: 1.6739... 0.0701 sec/batch
Epoch: 5/20... Training Step: 18455... Training loss: 1.5314... 0.0774 sec/batch
Epoch: 5/20... Training Step: 18456... Training loss: 1.6143... 0.0605 sec/batch
Epoch: 5/20... Training Step: 18457... Training loss: 1.7454... 0.0966 sec/batch
Epoch: 5/20... Training Step: 18458... Training loss: 1.6316... 0.0645 sec/batch
Epoch: 5/20... Training Step: 18459... Training loss: 1.6696... 0.0580 sec/batch
Epoch: 5/20... Training Step: 18460... Training loss: 1.6896... 0.0710 sec/batch
Epoch: 5/20... Training Step: 18461... Training loss: 1.6013... 0.0544 sec/batch
Epoch: 5/20... Training Step: 18462... Training loss: 1.7824... 0.0555 sec/batch
Epoch: 5/20... Training Step: 18463... Training loss: 1.5078... 0.0784 sec/batch
Epoch: 5/20... Training Step: 18464... Training loss: 1.5603... 0.0565 sec/batch
Epoch: 5/20... Training Step: 18465... Training loss: 1.5920... 0.0584 sec/batch
Epoch: 5/20... Training Step: 18466... Training loss: 1.8063... 0.0578 sec/batch
Epoch: 5/20... Training Step: 18467... Training loss: 1.7643... 0.0567 sec/batch
Epoch: 5/20... Training Step: 18468... Training loss: 1.8241... 0.0796 sec/batch
Epoch: 5/20... Training Step: 18469... Training loss: 1.6598... 0.0507 sec/batch
Epoch: 5/20... Training Step: 18470... Training loss: 1.7576... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18471... Training loss: 1.6714... 0.0643 sec/batch
Epoch: 5/20... Training Step: 18472... Training loss: 1.6697... 0.0668 sec/batch
Epoch: 5/20... Training Step: 18473... Training loss: 1.6875... 0.0588 sec/batch
Epoch: 5/20... Training Step: 18474... Training loss: 1.6135... 0.0564 sec/batch
Epoch: 5/20... Training Step: 18475... Training loss: 1.5864... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18476... Training loss: 1.6801... 0.0666 sec/batch
Epoch: 5/20... Training Step: 18477... Training loss: 1.4941... 0.0548 sec/batch
Epoch: 5/20... Training Step: 18478... Training loss: 1.5671... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18479... Training loss: 1.5544... 0.0653 sec/batch
Epoch: 5/20... Training Step: 18480... Training loss: 1.6193... 0.0673 sec/batch
Epoch: 5/20... Training Step: 18481... Training loss: 1.5985... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18482... Training loss: 1.5830... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18483... Training loss: 1.7927... 0.0814 sec/batch
Epoch: 5/20... Training Step: 18484... Training loss: 1.6601... 0.0615 sec/batch
Epoch: 5/20... Training Step: 18485... Training loss: 1.5689... 0.0615 sec/batch
Epoch: 5/20... Training Step: 18486... Training loss: 1.5724... 0.0683 sec/batch
Epoch: 5/20... Training Step: 18487... Training loss: 1.6426... 0.0771 sec/batch
Epoch: 5/20... Training Step: 18488... Training loss: 1.7081... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18489... Training loss: 1.5109... 0.0607 sec/batch
Epoch: 5/20... Training Step: 18490... Training loss: 1.7640... 0.0587 sec/batch
Epoch: 5/20... Training Step: 18491... Training loss: 1.6649... 0.0751 sec/batch
Epoch: 5/20... Training Step: 18492... Training loss: 1.5610... 0.0684 sec/batch
Epoch: 5/20... Training Step: 18493... Training loss: 1.6548... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18494... Training loss: 1.6040... 0.0660 sec/batch
Epoch: 5/20... Training Step: 18495... Training loss: 1.4508... 0.0822 sec/batch
Epoch: 5/20... Training Step: 18496... Training loss: 1.4898... 0.0710 sec/batch
Epoch: 5/20... Training Step: 18497... Training loss: 1.6570... 0.0614 sec/batch
Epoch: 5/20... Training Step: 18498... Training loss: 1.6485... 0.0559 sec/batch
Epoch: 5/20... Training Step: 18499... Training loss: 1.4795... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18500... Training loss: 1.5084... 0.0765 sec/batch
Epoch: 5/20... Training Step: 18501... Training loss: 1.7352... 0.0663 sec/batch
Epoch: 5/20... Training Step: 18502... Training loss: 1.4807... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18503... Training loss: 1.7092... 0.0678 sec/batch
Epoch: 5/20... Training Step: 18504... Training loss: 1.7893... 0.0613 sec/batch
Epoch: 5/20... Training Step: 18505... Training loss: 1.6844... 0.0575 sec/batch
Epoch: 5/20... Training Step: 18506... Training loss: 1.6524... 0.0732 sec/batch
Epoch: 5/20... Training Step: 18507... Training loss: 1.7186... 0.0744 sec/batch
Epoch: 5/20... Training Step: 18508... Training loss: 1.6275... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18509... Training loss: 1.5645... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18510... Training loss: 1.8656... 0.0673 sec/batch
Epoch: 5/20... Training Step: 18511... Training loss: 1.6718... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18512... Training loss: 1.5462... 0.0664 sec/batch
Epoch: 5/20... Training Step: 18513... Training loss: 1.6830... 0.0801 sec/batch
Epoch: 5/20... Training Step: 18514... Training loss: 1.7036... 0.0631 sec/batch
Epoch: 5/20... Training Step: 18515... Training loss: 1.4941... 0.0617 sec/batch
Epoch: 5/20... Training Step: 18516... Training loss: 1.7516... 0.0792 sec/batch
Epoch: 5/20... Training Step: 18517... Training loss: 1.7687... 0.0683 sec/batch
Epoch: 5/20... Training Step: 18518... Training loss: 1.5538... 0.0610 sec/batch
Epoch: 5/20... Training Step: 18519... Training loss: 1.7484... 0.0682 sec/batch
Epoch: 5/20... Training Step: 18520... Training loss: 1.5101... 0.0759 sec/batch
Epoch: 5/20... Training Step: 18521... Training loss: 1.6934... 0.0628 sec/batch
Epoch: 5/20... Training Step: 18522... Training loss: 1.6811... 0.0840 sec/batch
Epoch: 5/20... Training Step: 18523... Training loss: 1.5960... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18524... Training loss: 1.6202... 0.0648 sec/batch
Epoch: 5/20... Training Step: 18525... Training loss: 1.6608... 0.0726 sec/batch
Epoch: 5/20... Training Step: 18526... Training loss: 1.6781... 0.0635 sec/batch
Epoch: 5/20... Training Step: 18527... Training loss: 1.6177... 0.0616 sec/batch
Epoch: 5/20... Training Step: 18528... Training loss: 1.7139... 0.0649 sec/batch
Epoch: 5/20... Training Step: 18529... Training loss: 1.5886... 0.0645 sec/batch
Epoch: 5/20... Training Step: 18530... Training loss: 1.8055... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18531... Training loss: 1.7753... 0.0702 sec/batch
Epoch: 5/20... Training Step: 18532... Training loss: 1.7470... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18533... Training loss: 1.6911... 0.0728 sec/batch
Epoch: 5/20... Training Step: 18534... Training loss: 1.7032... 0.0633 sec/batch
Epoch: 5/20... Training Step: 18535... Training loss: 1.6123... 0.0577 sec/batch
Epoch: 5/20... Training Step: 18536... Training loss: 1.6326... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18537... Training loss: 1.7623... 0.0658 sec/batch
Epoch: 5/20... Training Step: 18538... Training loss: 1.6048... 0.0604 sec/batch
Epoch: 5/20... Training Step: 18539... Training loss: 1.5769... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18540... Training loss: 1.8342... 0.0701 sec/batch
Epoch: 5/20... Training Step: 18541... Training loss: 1.4856... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18542... Training loss: 1.6683... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18543... Training loss: 1.5795... 0.0631 sec/batch
Epoch: 5/20... Training Step: 18544... Training loss: 1.6328... 0.0694 sec/batch
Epoch: 5/20... Training Step: 18545... Training loss: 1.8299... 0.0669 sec/batch
Epoch: 5/20... Training Step: 18546... Training loss: 1.6543... 0.0527 sec/batch
Epoch: 5/20... Training Step: 18547... Training loss: 1.6108... 0.0573 sec/batch
Epoch: 5/20... Training Step: 18548... Training loss: 1.5577... 0.0675 sec/batch
Epoch: 5/20... Training Step: 18549... Training loss: 1.7302... 0.0694 sec/batch
Epoch: 5/20... Training Step: 18550... Training loss: 1.6314... 0.0579 sec/batch
Epoch: 5/20... Training Step: 18551... Training loss: 1.9512... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18552... Training loss: 1.6088... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18553... Training loss: 1.7595... 0.0738 sec/batch
Epoch: 5/20... Training Step: 18554... Training loss: 1.7352... 0.0564 sec/batch
Epoch: 5/20... Training Step: 18555... Training loss: 1.5328... 0.0638 sec/batch
Epoch: 5/20... Training Step: 18556... Training loss: 1.6486... 0.0840 sec/batch
Epoch: 5/20... Training Step: 18557... Training loss: 1.6268... 0.0551 sec/batch
Epoch: 5/20... Training Step: 18558... Training loss: 1.6261... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18559... Training loss: 1.7209... 0.0810 sec/batch
Epoch: 5/20... Training Step: 18560... Training loss: 1.5791... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18561... Training loss: 1.6303... 0.0551 sec/batch
Epoch: 5/20... Training Step: 18562... Training loss: 1.6093... 0.0515 sec/batch
Epoch: 5/20... Training Step: 18563... Training loss: 1.7237... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18564... Training loss: 1.5849... 0.0746 sec/batch
Epoch: 5/20... Training Step: 18565... Training loss: 1.5948... 0.0585 sec/batch
Epoch: 5/20... Training Step: 18566... Training loss: 1.5218... 0.0563 sec/batch
Epoch: 5/20... Training Step: 18567... Training loss: 1.5914... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18568... Training loss: 1.6303... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18569... Training loss: 1.5884... 0.0574 sec/batch
Epoch: 5/20... Training Step: 18570... Training loss: 1.7082... 0.0582 sec/batch
Epoch: 5/20... Training Step: 18571... Training loss: 1.7185... 0.0717 sec/batch
Epoch: 5/20... Training Step: 18572... Training loss: 1.6210... 0.0665 sec/batch
Epoch: 5/20... Training Step: 18573... Training loss: 1.5019... 0.0560 sec/batch
Epoch: 5/20... Training Step: 18574... Training loss: 1.7106... 0.0578 sec/batch
Epoch: 5/20... Training Step: 18575... Training loss: 1.8669... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18576... Training loss: 1.6743... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18577... Training loss: 1.6577... 0.0540 sec/batch
Epoch: 5/20... Training Step: 18578... Training loss: 1.6151... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18579... Training loss: 1.8632... 0.0631 sec/batch
Epoch: 5/20... Training Step: 18580... Training loss: 1.7391... 0.0688 sec/batch
Epoch: 5/20... Training Step: 18581... Training loss: 1.7445... 0.0574 sec/batch
Epoch: 5/20... Training Step: 18582... Training loss: 1.6900... 0.0544 sec/batch
Epoch: 5/20... Training Step: 18583... Training loss: 1.7686... 0.0704 sec/batch
Epoch: 5/20... Training Step: 18584... Training loss: 1.8012... 0.0674 sec/batch
Epoch: 5/20... Training Step: 18585... Training loss: 1.8246... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18586... Training loss: 1.7336... 0.0680 sec/batch
Epoch: 5/20... Training Step: 18587... Training loss: 1.5711... 0.0720 sec/batch
Epoch: 5/20... Training Step: 18588... Training loss: 1.6007... 0.0553 sec/batch
Epoch: 5/20... Training Step: 18589... Training loss: 1.6129... 0.0553 sec/batch
Epoch: 5/20... Training Step: 18590... Training loss: 1.6101... 0.0516 sec/batch
Epoch: 5/20... Training Step: 18591... Training loss: 1.4977... 0.0703 sec/batch
Epoch: 5/20... Training Step: 18592... Training loss: 1.7044... 0.0727 sec/batch
Epoch: 5/20... Training Step: 18593... Training loss: 1.6912... 0.0557 sec/batch
Epoch: 5/20... Training Step: 18594... Training loss: 1.6680... 0.0568 sec/batch
Epoch: 5/20... Training Step: 18595... Training loss: 1.6088... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18596... Training loss: 1.6457... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18597... Training loss: 1.5885... 0.0573 sec/batch
Epoch: 5/20... Training Step: 18598... Training loss: 1.5836... 0.0564 sec/batch
Epoch: 5/20... Training Step: 18599... Training loss: 1.6282... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18600... Training loss: 1.5094... 0.0690 sec/batch
Epoch: 5/20... Training Step: 18601... Training loss: 1.6383... 0.0569 sec/batch
Epoch: 5/20... Training Step: 18602... Training loss: 1.5164... 0.0648 sec/batch
Epoch: 5/20... Training Step: 18603... Training loss: 1.6240... 0.0551 sec/batch
Epoch: 5/20... Training Step: 18604... Training loss: 1.6801... 0.0542 sec/batch
Epoch: 5/20... Training Step: 18605... Training loss: 1.6315... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18606... Training loss: 1.8006... 0.0677 sec/batch
Epoch: 5/20... Training Step: 18607... Training loss: 1.7529... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18608... Training loss: 1.7395... 0.0614 sec/batch
Epoch: 5/20... Training Step: 18609... Training loss: 1.6812... 0.0732 sec/batch
Epoch: 5/20... Training Step: 18610... Training loss: 1.6182... 0.0655 sec/batch
Epoch: 5/20... Training Step: 18611... Training loss: 1.6964... 0.0615 sec/batch
Epoch: 5/20... Training Step: 18612... Training loss: 1.6047... 0.0613 sec/batch
Epoch: 5/20... Training Step: 18613... Training loss: 1.6813... 0.0865 sec/batch
Epoch: 5/20... Training Step: 18614... Training loss: 1.6419... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18615... Training loss: 1.5853... 0.0513 sec/batch
Epoch: 5/20... Training Step: 18616... Training loss: 1.7460... 0.0823 sec/batch
Epoch: 5/20... Training Step: 18617... Training loss: 1.6289... 0.0601 sec/batch
Epoch: 5/20... Training Step: 18618... Training loss: 1.6795... 0.0571 sec/batch
Epoch: 5/20... Training Step: 18619... Training loss: 1.8708... 0.0664 sec/batch
Epoch: 5/20... Training Step: 18620... Training loss: 1.5255... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18621... Training loss: 1.6620... 0.0671 sec/batch
Epoch: 5/20... Training Step: 18622... Training loss: 1.7244... 0.0607 sec/batch
Epoch: 5/20... Training Step: 18623... Training loss: 1.8344... 0.0784 sec/batch
Epoch: 5/20... Training Step: 18624... Training loss: 1.6840... 0.0789 sec/batch
Epoch: 5/20... Training Step: 18625... Training loss: 1.7179... 0.0591 sec/batch
Epoch: 5/20... Training Step: 18626... Training loss: 1.7982... 0.0567 sec/batch
Epoch: 5/20... Training Step: 18627... Training loss: 1.6193... 0.0679 sec/batch
Epoch: 5/20... Training Step: 18628... Training loss: 1.7222... 0.0607 sec/batch
Epoch: 5/20... Training Step: 18629... Training loss: 1.7763... 0.0544 sec/batch
Epoch: 5/20... Training Step: 18630... Training loss: 1.5888... 0.0564 sec/batch
Epoch: 5/20... Training Step: 18631... Training loss: 1.5449... 0.0694 sec/batch
Epoch: 5/20... Training Step: 18632... Training loss: 1.7482... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18633... Training loss: 1.5920... 0.0527 sec/batch
Epoch: 5/20... Training Step: 18634... Training loss: 1.6686... 0.0626 sec/batch
Epoch: 5/20... Training Step: 18635... Training loss: 1.7554... 0.0725 sec/batch
Epoch: 5/20... Training Step: 18636... Training loss: 1.6589... 0.0636 sec/batch
Epoch: 5/20... Training Step: 18637... Training loss: 1.7197... 0.0654 sec/batch
Epoch: 5/20... Training Step: 18638... Training loss: 1.6971... 0.0556 sec/batch
Epoch: 5/20... Training Step: 18639... Training loss: 1.6539... 0.0743 sec/batch
Epoch: 5/20... Training Step: 18640... Training loss: 1.6271... 0.0621 sec/batch
Epoch: 5/20... Training Step: 18641... Training loss: 1.5631... 0.0532 sec/batch
Epoch: 5/20... Training Step: 18642... Training loss: 1.6525... 0.0552 sec/batch
Epoch: 5/20... Training Step: 18643... Training loss: 1.6882... 0.0598 sec/batch
Epoch: 5/20... Training Step: 18644... Training loss: 1.5063... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18645... Training loss: 1.8021... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18646... Training loss: 1.6046... 0.0632 sec/batch
Epoch: 5/20... Training Step: 18647... Training loss: 1.5207... 0.0830 sec/batch
Epoch: 5/20... Training Step: 18648... Training loss: 1.7000... 0.0655 sec/batch
Epoch: 5/20... Training Step: 18649... Training loss: 1.8153... 0.0596 sec/batch
Epoch: 5/20... Training Step: 18650... Training loss: 1.7307... 0.0619 sec/batch
Epoch: 5/20... Training Step: 18651... Training loss: 1.6753... 0.0554 sec/batch
Epoch: 5/20... Training Step: 18652... Training loss: 1.7695... 0.0749 sec/batch
Epoch: 5/20... Training Step: 18653... Training loss: 1.6078... 0.0558 sec/batch
Epoch: 5/20... Training Step: 18654... Training loss: 1.6708... 0.0641 sec/batch
Epoch: 5/20... Training Step: 18655... Training loss: 1.6835... 0.0754 sec/batch
Epoch: 5/20... Training Step: 18656... Training loss: 1.6226... 0.0612 sec/batch
Epoch: 5/20... Training Step: 18657... Training loss: 1.6783... 0.0581 sec/batch
Epoch: 5/20... Training Step: 18658... Training loss: 1.4995... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18659... Training loss: 1.5659... 0.0619 sec/batch
Epoch: 5/20... Training Step: 18660... Training loss: 1.6515... 0.0682 sec/batch
Epoch: 5/20... Training Step: 18661... Training loss: 1.5679... 0.0540 sec/batch
Epoch: 5/20... Training Step: 18662... Training loss: 1.8034... 0.0600 sec/batch
Epoch: 5/20... Training Step: 18663... Training loss: 1.5466... 0.0784 sec/batch
Epoch: 5/20... Training Step: 18664... Training loss: 1.5611... 0.0721 sec/batch
Epoch: 5/20... Training Step: 18665... Training loss: 1.5625... 0.0538 sec/batch
Epoch: 5/20... Training Step: 18666... Training loss: 1.7925... 0.0638 sec/batch
Epoch: 5/20... Training Step: 18667... Training loss: 1.5619... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18668... Training loss: 1.5564... 0.0794 sec/batch
Epoch: 5/20... Training Step: 18669... Training loss: 1.5855... 0.0613 sec/batch
Epoch: 5/20... Training Step: 18670... Training loss: 1.5434... 0.0671 sec/batch
Epoch: 5/20... Training Step: 18671... Training loss: 1.6469... 0.0770 sec/batch
Epoch: 5/20... Training Step: 18672... Training loss: 1.5681... 0.0557 sec/batch
Epoch: 5/20... Training Step: 18673... Training loss: 1.5134... 0.0661 sec/batch
Epoch: 5/20... Training Step: 18674... Training loss: 1.4945... 0.0603 sec/batch
Epoch: 5/20... Training Step: 18675... Training loss: 1.5291... 0.0732 sec/batch
Epoch: 5/20... Training Step: 18676... Training loss: 1.7635... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18677... Training loss: 1.4999... 0.0726 sec/batch
Epoch: 5/20... Training Step: 18678... Training loss: 1.6857... 0.0843 sec/batch
Epoch: 5/20... Training Step: 18679... Training loss: 1.6305... 0.0602 sec/batch
Epoch: 5/20... Training Step: 18680... Training loss: 1.4727... 0.0635 sec/batch
Epoch: 5/20... Training Step: 18681... Training loss: 1.6529... 0.0987 sec/batch
Epoch: 5/20... Training Step: 18682... Training loss: 1.6862... 0.0579 sec/batch
Epoch: 5/20... Training Step: 18683... Training loss: 1.7333... 0.0658 sec/batch
Epoch: 5/20... Training Step: 18684... Training loss: 1.6379... 0.0868 sec/batch
Epoch: 5/20... Training Step: 18685... Training loss: 1.6089... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18686... Training loss: 1.7264... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18687... Training loss: 1.7164... 0.0689 sec/batch
Epoch: 5/20... Training Step: 18688... Training loss: 1.6703... 0.0684 sec/batch
Epoch: 5/20... Training Step: 18689... Training loss: 1.6232... 0.0699 sec/batch
Epoch: 5/20... Training Step: 18690... Training loss: 1.7518... 0.0845 sec/batch
Epoch: 5/20... Training Step: 18691... Training loss: 1.6975... 0.0563 sec/batch
Epoch: 5/20... Training Step: 18692... Training loss: 1.5977... 0.0599 sec/batch
Epoch: 5/20... Training Step: 18693... Training loss: 1.6261... 0.0903 sec/batch
Epoch: 5/20... Training Step: 18694... Training loss: 1.5966... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18695... Training loss: 1.7495... 0.0616 sec/batch
Epoch: 5/20... Training Step: 18696... Training loss: 1.7503... 0.0917 sec/batch
Epoch: 5/20... Training Step: 18697... Training loss: 1.6275... 0.0590 sec/batch
Epoch: 5/20... Training Step: 18698... Training loss: 1.7616... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18699... Training loss: 1.5520... 0.0752 sec/batch
Epoch: 5/20... Training Step: 18700... Training loss: 1.7514... 0.0688 sec/batch
Epoch: 5/20... Training Step: 18701... Training loss: 1.7779... 0.0700 sec/batch
Epoch: 5/20... Training Step: 18702... Training loss: 1.6185... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18703... Training loss: 1.7619... 0.0703 sec/batch
Epoch: 5/20... Training Step: 18704... Training loss: 1.4459... 0.0669 sec/batch
Epoch: 5/20... Training Step: 18705... Training loss: 1.8411... 0.0763 sec/batch
Epoch: 5/20... Training Step: 18706... Training loss: 1.6946... 0.0739 sec/batch
Epoch: 5/20... Training Step: 18707... Training loss: 1.6627... 0.0780 sec/batch
Epoch: 5/20... Training Step: 18708... Training loss: 1.7783... 0.0832 sec/batch
Epoch: 5/20... Training Step: 18709... Training loss: 1.6587... 0.0835 sec/batch
Epoch: 5/20... Training Step: 18710... Training loss: 1.7484... 0.0579 sec/batch
Epoch: 5/20... Training Step: 18711... Training loss: 1.7463... 0.0577 sec/batch
Epoch: 5/20... Training Step: 18712... Training loss: 1.7681... 0.0743 sec/batch
Epoch: 5/20... Training Step: 18713... Training loss: 1.7311... 0.0696 sec/batch
Epoch: 5/20... Training Step: 18714... Training loss: 1.8268... 0.0652 sec/batch
Epoch: 5/20... Training Step: 18715... Training loss: 1.6696... 0.0627 sec/batch
Epoch: 5/20... Training Step: 18716... Training loss: 1.7879... 0.0617 sec/batch
Epoch: 5/20... Training Step: 18717... Training loss: 1.8285... 0.0831 sec/batch
Epoch: 5/20... Training Step: 18718... Training loss: 1.7334... 0.0669 sec/batch
Epoch: 5/20... Training Step: 18719... Training loss: 1.7756... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18720... Training loss: 1.6424... 0.0567 sec/batch
Epoch: 5/20... Training Step: 18721... Training loss: 1.6963... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18722... Training loss: 1.7564... 0.0680 sec/batch
Epoch: 5/20... Training Step: 18723... Training loss: 1.4749... 0.0751 sec/batch
Epoch: 5/20... Training Step: 18724... Training loss: 1.5021... 0.0719 sec/batch
Epoch: 5/20... Training Step: 18725... Training loss: 1.6853... 0.0572 sec/batch
Epoch: 5/20... Training Step: 18726... Training loss: 1.6725... 0.0594 sec/batch
Epoch: 5/20... Training Step: 18727... Training loss: 1.6665... 0.0875 sec/batch
Epoch: 5/20... Training Step: 18728... Training loss: 1.6686... 0.0663 sec/batch
Epoch: 5/20... Training Step: 18729... Training loss: 1.5988... 0.0551 sec/batch
Epoch: 5/20... Training Step: 18730... Training loss: 1.8081... 0.0667 sec/batch
Epoch: 5/20... Training Step: 18731... Training loss: 1.6669... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18732... Training loss: 1.7951... 0.0693 sec/batch
Epoch: 5/20... Training Step: 18733... Training loss: 1.8108... 0.0604 sec/batch
Epoch: 5/20... Training Step: 18734... Training loss: 1.6385... 0.0541 sec/batch
Epoch: 5/20... Training Step: 18735... Training loss: 1.6094... 0.0563 sec/batch
Epoch: 5/20... Training Step: 18736... Training loss: 1.7740... 0.0803 sec/batch
Epoch: 5/20... Training Step: 18737... Training loss: 1.6993... 0.0548 sec/batch
Epoch: 5/20... Training Step: 18738... Training loss: 1.7371... 0.0649 sec/batch
Epoch: 5/20... Training Step: 18739... Training loss: 1.4960... 0.0701 sec/batch
Epoch: 5/20... Training Step: 18740... Training loss: 1.5929... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18741... Training loss: 1.7050... 0.0651 sec/batch
Epoch: 5/20... Training Step: 18742... Training loss: 1.5205... 0.0714 sec/batch
Epoch: 5/20... Training Step: 18743... Training loss: 1.6507... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18744... Training loss: 1.6745... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18745... Training loss: 1.5853... 0.0852 sec/batch
Epoch: 5/20... Training Step: 18746... Training loss: 1.5974... 0.0593 sec/batch
Epoch: 5/20... Training Step: 18747... Training loss: 1.5786... 0.0518 sec/batch
Epoch: 5/20... Training Step: 18748... Training loss: 1.6432... 0.0639 sec/batch
Epoch: 5/20... Training Step: 18749... Training loss: 1.7305... 0.0725 sec/batch
Epoch: 5/20... Training Step: 18750... Training loss: 1.5933... 0.0768 sec/batch
Epoch: 5/20... Training Step: 18751... Training loss: 1.5040... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18752... Training loss: 1.5294... 0.0542 sec/batch
Epoch: 5/20... Training Step: 18753... Training loss: 1.6602... 0.0549 sec/batch
Epoch: 5/20... Training Step: 18754... Training loss: 1.6335... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18755... Training loss: 1.6570... 0.0626 sec/batch
Epoch: 5/20... Training Step: 18756... Training loss: 1.8421... 0.0541 sec/batch
Epoch: 5/20... Training Step: 18757... Training loss: 1.6969... 0.0655 sec/batch
Epoch: 5/20... Training Step: 18758... Training loss: 1.5928... 0.0715 sec/batch
Epoch: 5/20... Training Step: 18759... Training loss: 1.8662... 0.0608 sec/batch
Epoch: 5/20... Training Step: 18760... Training loss: 1.5809... 0.0581 sec/batch
Epoch: 5/20... Training Step: 18761... Training loss: 1.7812... 0.0719 sec/batch
Epoch: 5/20... Training Step: 18762... Training loss: 1.7491... 0.0717 sec/batch
Epoch: 5/20... Training Step: 18763... Training loss: 1.7482... 0.0619 sec/batch
Epoch: 5/20... Training Step: 18764... Training loss: 1.6109... 0.0581 sec/batch
Epoch: 5/20... Training Step: 18765... Training loss: 1.6168... 0.0719 sec/batch
Epoch: 5/20... Training Step: 18766... Training loss: 1.5616... 0.0700 sec/batch
Epoch: 5/20... Training Step: 18767... Training loss: 1.5922... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18768... Training loss: 1.6445... 0.0541 sec/batch
Epoch: 5/20... Training Step: 18769... Training loss: 1.6914... 0.0620 sec/batch
Epoch: 5/20... Training Step: 18770... Training loss: 1.5747... 0.0713 sec/batch
Epoch: 5/20... Training Step: 18771... Training loss: 1.7623... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18772... Training loss: 1.5197... 0.0668 sec/batch
Epoch: 5/20... Training Step: 18773... Training loss: 1.7448... 0.0872 sec/batch
Epoch: 5/20... Training Step: 18774... Training loss: 1.6431... 0.0722 sec/batch
Epoch: 5/20... Training Step: 18775... Training loss: 1.5472... 0.0610 sec/batch
Epoch: 5/20... Training Step: 18776... Training loss: 1.5863... 0.0617 sec/batch
Epoch: 5/20... Training Step: 18777... Training loss: 1.5380... 0.0748 sec/batch
Epoch: 5/20... Training Step: 18778... Training loss: 1.7174... 0.0656 sec/batch
Epoch: 5/20... Training Step: 18779... Training loss: 1.6349... 0.0580 sec/batch
Epoch: 5/20... Training Step: 18780... Training loss: 1.5263... 0.0556 sec/batch
Epoch: 5/20... Training Step: 18781... Training loss: 1.4469... 0.0847 sec/batch
Epoch: 5/20... Training Step: 18782... Training loss: 1.6630... 0.0785 sec/batch
Epoch: 5/20... Training Step: 18783... Training loss: 1.5333... 0.0517 sec/batch
Epoch: 5/20... Training Step: 18784... Training loss: 1.6410... 0.0596 sec/batch
Epoch: 5/20... Training Step: 18785... Training loss: 1.6860... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18786... Training loss: 1.6551... 0.0730 sec/batch
Epoch: 5/20... Training Step: 18787... Training loss: 1.6705... 0.0614 sec/batch
Epoch: 5/20... Training Step: 18788... Training loss: 1.7171... 0.0680 sec/batch
Epoch: 5/20... Training Step: 18789... Training loss: 1.5559... 0.0858 sec/batch
Epoch: 5/20... Training Step: 18790... Training loss: 1.7044... 0.0689 sec/batch
Epoch: 5/20... Training Step: 18791... Training loss: 1.5306... 0.0851 sec/batch
Epoch: 5/20... Training Step: 18792... Training loss: 1.7580... 0.0748 sec/batch
Epoch: 5/20... Training Step: 18793... Training loss: 1.7138... 0.0695 sec/batch
Epoch: 5/20... Training Step: 18794... Training loss: 1.6748... 0.0555 sec/batch
Epoch: 5/20... Training Step: 18795... Training loss: 1.8025... 0.0715 sec/batch
Epoch: 5/20... Training Step: 18796... Training loss: 1.6719... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18797... Training loss: 1.7062... 0.0637 sec/batch
Epoch: 5/20... Training Step: 18798... Training loss: 1.6638... 0.0856 sec/batch
Epoch: 5/20... Training Step: 18799... Training loss: 1.7075... 0.0700 sec/batch
Epoch: 5/20... Training Step: 18800... Training loss: 1.7963... 0.0575 sec/batch
Epoch: 5/20... Training Step: 18801... Training loss: 1.7702... 0.0836 sec/batch
Epoch: 5/20... Training Step: 18802... Training loss: 1.6115... 0.0660 sec/batch
Epoch: 5/20... Training Step: 18803... Training loss: 1.6058... 0.0653 sec/batch
Epoch: 5/20... Training Step: 18804... Training loss: 1.8130... 0.0645 sec/batch
Epoch: 5/20... Training Step: 18805... Training loss: 1.6360... 0.0713 sec/batch
Epoch: 5/20... Training Step: 18806... Training loss: 1.5891... 0.0665 sec/batch
Epoch: 5/20... Training Step: 18807... Training loss: 1.6242... 0.0704 sec/batch
Epoch: 5/20... Training Step: 18808... Training loss: 1.6445... 0.0605 sec/batch
Epoch: 5/20... Training Step: 18809... Training loss: 1.5361... 0.0847 sec/batch
Epoch: 5/20... Training Step: 18810... Training loss: 1.5958... 0.0649 sec/batch
Epoch: 5/20... Training Step: 18811... Training loss: 1.5963... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18812... Training loss: 1.5191... 0.0605 sec/batch
Epoch: 5/20... Training Step: 18813... Training loss: 1.5821... 0.0688 sec/batch
Epoch: 5/20... Training Step: 18814... Training loss: 1.6513... 0.0722 sec/batch
Epoch: 5/20... Training Step: 18815... Training loss: 1.6305... 0.0670 sec/batch
Epoch: 5/20... Training Step: 18816... Training loss: 1.6277... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18817... Training loss: 1.6988... 0.0781 sec/batch
Epoch: 5/20... Training Step: 18818... Training loss: 1.6803... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18819... Training loss: 1.5453... 0.0659 sec/batch
Epoch: 5/20... Training Step: 18820... Training loss: 1.6530... 0.0761 sec/batch
Epoch: 5/20... Training Step: 18821... Training loss: 1.7774... 0.0596 sec/batch
Epoch: 5/20... Training Step: 18822... Training loss: 1.7911... 0.0595 sec/batch
Epoch: 5/20... Training Step: 18823... Training loss: 1.6514... 0.0611 sec/batch
Epoch: 5/20... Training Step: 18824... Training loss: 1.5781... 0.0698 sec/batch
Epoch: 5/20... Training Step: 18825... Training loss: 1.5516... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18826... Training loss: 1.6513... 0.0663 sec/batch
Epoch: 5/20... Training Step: 18827... Training loss: 1.5916... 0.0809 sec/batch
Epoch: 5/20... Training Step: 18828... Training loss: 1.5648... 0.0683 sec/batch
Epoch: 5/20... Training Step: 18829... Training loss: 1.5623... 0.0555 sec/batch
Epoch: 5/20... Training Step: 18830... Training loss: 1.4213... 0.0818 sec/batch
Epoch: 5/20... Training Step: 18831... Training loss: 1.5282... 0.0730 sec/batch
Epoch: 5/20... Training Step: 18832... Training loss: 1.5284... 0.0918 sec/batch
Epoch: 5/20... Training Step: 18833... Training loss: 1.5753... 0.0765 sec/batch
Epoch: 5/20... Training Step: 18834... Training loss: 1.5603... 0.0654 sec/batch
Epoch: 5/20... Training Step: 18835... Training loss: 1.5003... 0.0707 sec/batch
Epoch: 5/20... Training Step: 18836... Training loss: 1.5858... 0.1073 sec/batch
Epoch: 5/20... Training Step: 18837... Training loss: 1.5608... 0.0782 sec/batch
Epoch: 5/20... Training Step: 18838... Training loss: 1.5890... 0.1299 sec/batch
Epoch: 5/20... Training Step: 18839... Training loss: 1.5941... 0.0743 sec/batch
Epoch: 5/20... Training Step: 18840... Training loss: 1.7018... 0.0627 sec/batch
Epoch: 5/20... Training Step: 18841... Training loss: 1.6622... 0.0604 sec/batch
Epoch: 5/20... Training Step: 18842... Training loss: 1.6593... 0.0867 sec/batch
Epoch: 5/20... Training Step: 18843... Training loss: 1.5849... 0.0699 sec/batch
Epoch: 5/20... Training Step: 18844... Training loss: 1.4861... 0.0618 sec/batch
Epoch: 5/20... Training Step: 18845... Training loss: 1.5777... 0.0640 sec/batch
Epoch: 5/20... Training Step: 18846... Training loss: 1.6292... 0.0703 sec/batch
Epoch: 5/20... Training Step: 18847... Training loss: 1.5803... 0.0720 sec/batch
Epoch: 5/20... Training Step: 18848... Training loss: 1.6718... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18849... Training loss: 1.5888... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18850... Training loss: 1.5872... 0.0777 sec/batch
Epoch: 5/20... Training Step: 18851... Training loss: 1.7794... 0.0750 sec/batch
Epoch: 5/20... Training Step: 18852... Training loss: 1.5429... 0.0624 sec/batch
Epoch: 5/20... Training Step: 18853... Training loss: 1.5235... 0.0881 sec/batch
Epoch: 5/20... Training Step: 18854... Training loss: 1.6789... 0.0654 sec/batch
Epoch: 5/20... Training Step: 18855... Training loss: 1.6596... 0.0750 sec/batch
Epoch: 5/20... Training Step: 18856... Training loss: 1.6319... 0.0980 sec/batch
Epoch: 5/20... Training Step: 18857... Training loss: 1.5321... 0.1065 sec/batch
Epoch: 5/20... Training Step: 18858... Training loss: 1.6659... 0.1439 sec/batch
Epoch: 5/20... Training Step: 18859... Training loss: 1.6989... 0.1157 sec/batch
Epoch: 5/20... Training Step: 18860... Training loss: 1.8273... 0.1156 sec/batch
Epoch: 5/20... Training Step: 18861... Training loss: 1.7840... 0.0693 sec/batch
Epoch: 5/20... Training Step: 18862... Training loss: 1.7540... 0.0978 sec/batch
Epoch: 5/20... Training Step: 18863... Training loss: 1.5109... 0.0837 sec/batch
Epoch: 5/20... Training Step: 18864... Training loss: 1.4690... 0.0655 sec/batch
Epoch: 5/20... Training Step: 18865... Training loss: 1.6846... 0.0709 sec/batch
Epoch: 5/20... Training Step: 18866... Training loss: 1.6829... 0.0840 sec/batch
Epoch: 5/20... Training Step: 18867... Training loss: 1.6387... 0.0691 sec/batch
Epoch: 5/20... Training Step: 18868... Training loss: 1.7526... 0.0721 sec/batch
Epoch: 5/20... Training Step: 18869... Training loss: 1.6805... 0.0720 sec/batch
Epoch: 5/20... Training Step: 18870... Training loss: 1.6685... 0.0718 sec/batch
Epoch: 5/20... Training Step: 18871... Training loss: 1.6860... 0.0722 sec/batch
Epoch: 5/20... Training Step: 18872... Training loss: 1.6728... 0.1090 sec/batch
Epoch: 5/20... Training Step: 18873... Training loss: 1.4659... 0.0683 sec/batch
Epoch: 5/20... Training Step: 18874... Training loss: 1.6996... 0.0806 sec/batch
Epoch: 5/20... Training Step: 18875... Training loss: 1.6805... 0.0752 sec/batch
Epoch: 5/20... Training Step: 18876... Training loss: 1.7443... 0.0669 sec/batch
Epoch: 5/20... Training Step: 18877... Training loss: 1.8189... 0.0682 sec/batch
Epoch: 5/20... Training Step: 18878... Training loss: 1.7232... 0.0757 sec/batch
Epoch: 5/20... Training Step: 18879... Training loss: 1.5218... 0.0684 sec/batch
Epoch: 5/20... Training Step: 18880... Training loss: 1.6084... 0.0699 sec/batch
Epoch: 5/20... Training Step: 18881... Training loss: 1.5951... 0.0687 sec/batch
Epoch: 5/20... Training Step: 18882... Training loss: 1.4941... 0.0736 sec/batch
Epoch: 5/20... Training Step: 18883... Training loss: 1.6992... 0.0602 sec/batch
Epoch: 5/20... Training Step: 18884... Training loss: 1.5923... 0.0696 sec/batch
Epoch: 5/20... Training Step: 18885... Training loss: 1.7253... 0.0658 sec/batch
Epoch: 5/20... Training Step: 18886... Training loss: 1.6547... 0.0596 sec/batch
Epoch: 5/20... Training Step: 18887... Training loss: 1.5917... 0.0531 sec/batch
Epoch: 5/20... Training Step: 18888... Training loss: 1.7706... 0.0488 sec/batch
Epoch: 5/20... Training Step: 18889... Training loss: 1.6013... 0.0484 sec/batch
Epoch: 5/20... Training Step: 18890... Training loss: 1.7545... 0.0749 sec/batch
Epoch: 5/20... Training Step: 18891... Training loss: 1.5418... 0.0626 sec/batch
Epoch: 5/20... Training Step: 18892... Training loss: 1.7779... 0.0598 sec/batch
Epoch: 5/20... Training Step: 18893... Training loss: 1.5562... 0.0629 sec/batch
Epoch: 5/20... Training Step: 18894... Training loss: 1.7011... 0.0510 sec/batch
Epoch: 5/20... Training Step: 18895... Training loss: 1.7678... 0.0662 sec/batch
Epoch: 5/20... Training Step: 18896... Training loss: 1.6126... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18897... Training loss: 1.6649... 0.0571 sec/batch
Epoch: 5/20... Training Step: 18898... Training loss: 1.7180... 0.0877 sec/batch
Epoch: 5/20... Training Step: 18899... Training loss: 1.5396... 0.0504 sec/batch
Epoch: 5/20... Training Step: 18900... Training loss: 1.8167... 0.0542 sec/batch
Epoch: 5/20... Training Step: 18901... Training loss: 1.6589... 0.0796 sec/batch
Epoch: 5/20... Training Step: 18902... Training loss: 1.6812... 0.0716 sec/batch
Epoch: 5/20... Training Step: 18903... Training loss: 1.6209... 0.0581 sec/batch
Epoch: 5/20... Training Step: 18904... Training loss: 1.5434... 0.0630 sec/batch
Epoch: 5/20... Training Step: 18905... Training loss: 1.6372... 0.0465 sec/batch
Epoch: 5/20... Training Step: 18906... Training loss: 1.6202... 0.0464 sec/batch
Epoch: 5/20... Training Step: 18907... Training loss: 1.6438... 0.0575 sec/batch
Epoch: 5/20... Training Step: 18908... Training loss: 1.7223... 0.0587 sec/batch
Epoch: 5/20... Training Step: 18909... Training loss: 1.8700... 0.0646 sec/batch
Epoch: 5/20... Training Step: 18910... Training loss: 1.6388... 0.0620 sec/batch
Epoch: 5/20... Training Step: 18911... Training loss: 1.7112... 0.0551 sec/batch
Epoch: 5/20... Training Step: 18912... Training loss: 1.7361... 0.0586 sec/batch
Epoch: 5/20... Training Step: 18913... Training loss: 1.6561... 0.0925 sec/batch
Epoch: 5/20... Training Step: 18914... Training loss: 1.7182... 0.0768 sec/batch
Epoch: 5/20... Training Step: 18915... Training loss: 1.4290... 0.0657 sec/batch
Epoch: 5/20... Training Step: 18916... Training loss: 1.7760... 0.0506 sec/batch
Epoch: 5/20... Training Step: 18917... Training loss: 1.7172... 0.0689 sec/batch
Epoch: 5/20... Training Step: 18918... Training loss: 1.7826... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18919... Training loss: 1.5727... 0.0644 sec/batch
Epoch: 5/20... Training Step: 18920... Training loss: 1.7138... 0.0606 sec/batch
Epoch: 5/20... Training Step: 18921... Training loss: 1.5722... 0.0486 sec/batch
Epoch: 5/20... Training Step: 18922... Training loss: 1.6706... 0.0497 sec/batch
Epoch: 5/20... Training Step: 18923... Training loss: 1.7528... 0.0493 sec/batch
Epoch: 5/20... Training Step: 18924... Training loss: 1.6495... 0.0498 sec/batch
Epoch: 5/20... Training Step: 18925... Training loss: 1.7108... 0.0503 sec/batch
Epoch: 5/20... Training Step: 18926... Training loss: 1.7230... 0.0593 sec/batch
Epoch: 5/20... Training Step: 18927... Training loss: 1.5409... 0.0538 sec/batch
Epoch: 5/20... Training Step: 18928... Training loss: 1.8376... 0.0702 sec/batch
Epoch: 5/20... Training Step: 18929... Training loss: 1.7355... 0.0642 sec/batch
Epoch: 5/20... Training Step: 18930... Training loss: 1.7151... 0.0513 sec/batch
Epoch: 5/20... Training Step: 18931... Training loss: 1.7308... 0.0460 sec/batch
Epoch: 5/20... Training Step: 18932... Training loss: 1.7084... 0.0460 sec/batch
Epoch: 5/20... Training Step: 18933... Training loss: 1.8638... 0.0455 sec/batch
Epoch: 5/20... Training Step: 18934... Training loss: 1.6099... 0.0494 sec/batch
Epoch: 5/20... Training Step: 18935... Training loss: 1.6044... 0.0492 sec/batch
Epoch: 5/20... Training Step: 18936... Training loss: 1.6064... 0.0820 sec/batch
Epoch: 5/20... Training Step: 18937... Training loss: 1.7892... 0.1008 sec/batch
Epoch: 5/20... Training Step: 18938... Training loss: 1.6700... 0.0535 sec/batch
Epoch: 5/20... Training Step: 18939... Training loss: 1.7911... 0.0572 sec/batch
Epoch: 5/20... Training Step: 18940... Training loss: 1.5899... 0.0489 sec/batch
Epoch: 5/20... Training Step: 18941... Training loss: 1.8277... 0.0830 sec/batch
Epoch: 5/20... Training Step: 18942... Training loss: 1.6295... 0.0539 sec/batch
Epoch: 5/20... Training Step: 18943... Training loss: 1.7491... 0.0851 sec/batch
Epoch: 5/20... Training Step: 18944... Training loss: 1.5975... 0.0532 sec/batch
Epoch: 5/20... Training Step: 18945... Training loss: 1.6025... 0.0518 sec/batch
Epoch: 5/20... Training Step: 18946... Training loss: 1.6097... 0.0505 sec/batch
Epoch: 5/20... Training Step: 18947... Training loss: 1.7329... 0.0634 sec/batch
Epoch: 5/20... Training Step: 18948... Training loss: 1.6344... 0.0625 sec/batch
Epoch: 5/20... Training Step: 18949... Training loss: 1.6595... 0.0547 sec/batch
Epoch: 5/20... Training Step: 18950... Training loss: 1.6521... 0.0516 sec/batch
Epoch: 5/20... Training Step: 18951... Training loss: 1.5862... 0.0476 sec/batch
Epoch: 5/20... Training Step: 18952... Training loss: 1.6396... 0.0464 sec/batch
Epoch: 5/20... Training Step: 18953... Training loss: 1.6777... 0.0457 sec/batch
Epoch: 5/20... Training Step: 18954... Training loss: 1.4654... 0.0462 sec/batch
Epoch: 5/20... Training Step: 18955... Training loss: 1.7211... 0.0546 sec/batch
Epoch: 5/20... Training Step: 18956... Training loss: 1.6448... 0.0413 sec/batch
Epoch: 5/20... Training Step: 18957... Training loss: 1.6457... 0.0428 sec/batch
Epoch: 5/20... Training Step: 18958... Training loss: 1.6269... 0.0421 sec/batch
Epoch: 5/20... Training Step: 18959... Training loss: 1.4883... 0.0490 sec/batch
Epoch: 5/20... Training Step: 18960... Training loss: 1.6452... 0.0512 sec/batch
Epoch: 5/20... Training Step: 18961... Training loss: 1.6697... 0.0432 sec/batch
Epoch: 5/20... Training Step: 18962... Training loss: 1.5103... 0.0418 sec/batch
Epoch: 5/20... Training Step: 18963... Training loss: 1.7003... 0.0447 sec/batch
Epoch: 5/20... Training Step: 18964... Training loss: 1.5935... 0.0519 sec/batch
Epoch: 5/20... Training Step: 18965... Training loss: 1.6300... 0.0470 sec/batch
Epoch: 5/20... Training Step: 18966... Training loss: 1.6115... 0.0420 sec/batch
Epoch: 5/20... Training Step: 18967... Training loss: 1.6082... 0.0425 sec/batch
Epoch: 5/20... Training Step: 18968... Training loss: 1.5503... 0.0441 sec/batch
Epoch: 5/20... Training Step: 18969... Training loss: 1.6218... 0.0443 sec/batch
Epoch: 5/20... Training Step: 18970... Training loss: 1.5833... 0.0516 sec/batch
Epoch: 5/20... Training Step: 18971... Training loss: 1.4865... 0.0415 sec/batch
Epoch: 5/20... Training Step: 18972... Training loss: 1.5421... 0.0410 sec/batch
Epoch: 5/20... Training Step: 18973... Training loss: 1.7073... 0.0423 sec/batch
Epoch: 5/20... Training Step: 18974... Training loss: 1.8201... 0.0440 sec/batch
Epoch: 5/20... Training Step: 18975... Training loss: 1.7212... 0.0463 sec/batch
Epoch: 5/20... Training Step: 18976... Training loss: 1.6197... 0.0414 sec/batch
Epoch: 5/20... Training Step: 18977... Training loss: 1.6924... 0.0417 sec/batch
Epoch: 5/20... Training Step: 18978... Training loss: 1.6905... 0.0418 sec/batch
Epoch: 5/20... Training Step: 18979... Training loss: 1.5386... 0.0474 sec/batch
Epoch: 5/20... Training Step: 18980... Training loss: 1.5430... 0.0458 sec/batch
Epoch: 5/20... Training Step: 18981... Training loss: 1.7115... 0.0477 sec/batch
Epoch: 5/20... Training Step: 18982... Training loss: 1.5953... 0.0507 sec/batch
Epoch: 5/20... Training Step: 18983... Training loss: 1.4778... 0.0444 sec/batch
Epoch: 5/20... Training Step: 18984... Training loss: 1.6045... 0.0489 sec/batch
Epoch: 5/20... Training Step: 18985... Training loss: 1.5173... 0.0509 sec/batch
Epoch: 5/20... Training Step: 18986... Training loss: 1.5754... 0.0458 sec/batch
Epoch: 5/20... Training Step: 18987... Training loss: 1.7305... 0.0442 sec/batch
Epoch: 5/20... Training Step: 18988... Training loss: 1.5390... 0.0460 sec/batch
Epoch: 5/20... Training Step: 18989... Training loss: 1.5066... 0.0436 sec/batch
Epoch: 5/20... Training Step: 18990... Training loss: 1.6169... 0.0486 sec/batch
Epoch: 5/20... Training Step: 18991... Training loss: 1.5971... 0.0461 sec/batch
Epoch: 5/20... Training Step: 18992... Training loss: 1.6347... 0.0414 sec/batch
Epoch: 5/20... Training Step: 18993... Training loss: 1.6569... 0.0463 sec/batch
Epoch: 5/20... Training Step: 18994... Training loss: 1.5915... 0.0434 sec/batch
Epoch: 5/20... Training Step: 18995... Training loss: 1.6344... 0.0464 sec/batch
Epoch: 5/20... Training Step: 18996... Training loss: 1.5703... 0.0463 sec/batch
Epoch: 5/20... Training Step: 18997... Training loss: 1.5339... 0.0424 sec/batch
Epoch: 5/20... Training Step: 18998... Training loss: 1.6271... 0.0438 sec/batch
Epoch: 5/20... Training Step: 18999... Training loss: 1.6882... 0.0528 sec/batch
Epoch: 5/20... Training Step: 19000... Training loss: 1.6032... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19001... Training loss: 1.6151... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19002... Training loss: 1.5892... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19003... Training loss: 1.7183... 0.0471 sec/batch
Epoch: 5/20... Training Step: 19004... Training loss: 1.6216... 0.0462 sec/batch
Epoch: 5/20... Training Step: 19005... Training loss: 1.6151... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19006... Training loss: 1.6497... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19007... Training loss: 1.6724... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19008... Training loss: 1.6059... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19009... Training loss: 1.5013... 0.0484 sec/batch
Epoch: 5/20... Training Step: 19010... Training loss: 1.5857... 0.0493 sec/batch
Epoch: 5/20... Training Step: 19011... Training loss: 1.6312... 0.0453 sec/batch
Epoch: 5/20... Training Step: 19012... Training loss: 1.6083... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19013... Training loss: 1.7811... 0.0494 sec/batch
Epoch: 5/20... Training Step: 19014... Training loss: 1.6382... 0.0536 sec/batch
Epoch: 5/20... Training Step: 19015... Training loss: 1.7323... 0.0454 sec/batch
Epoch: 5/20... Training Step: 19016... Training loss: 1.7376... 0.0461 sec/batch
Epoch: 5/20... Training Step: 19017... Training loss: 1.5728... 0.0455 sec/batch
Epoch: 5/20... Training Step: 19018... Training loss: 1.6697... 0.0515 sec/batch
Epoch: 5/20... Training Step: 19019... Training loss: 1.6194... 0.0489 sec/batch
Epoch: 5/20... Training Step: 19020... Training loss: 1.7714... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19021... Training loss: 1.7487... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19022... Training loss: 1.5147... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19023... Training loss: 1.6856... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19024... Training loss: 1.5327... 0.0568 sec/batch
Epoch: 5/20... Training Step: 19025... Training loss: 1.7006... 0.0454 sec/batch
Epoch: 5/20... Training Step: 19026... Training loss: 1.6475... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19027... Training loss: 1.6152... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19028... Training loss: 1.6728... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19029... Training loss: 1.5996... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19030... Training loss: 1.5369... 0.0491 sec/batch
Epoch: 5/20... Training Step: 19031... Training loss: 1.6351... 0.0602 sec/batch
Epoch: 5/20... Training Step: 19032... Training loss: 1.7084... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19033... Training loss: 1.6853... 0.0489 sec/batch
Epoch: 5/20... Training Step: 19034... Training loss: 1.5897... 0.0513 sec/batch
Epoch: 5/20... Training Step: 19035... Training loss: 1.6415... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19036... Training loss: 1.7268... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19037... Training loss: 1.6193... 0.0524 sec/batch
Epoch: 5/20... Training Step: 19038... Training loss: 1.6561... 0.0464 sec/batch
Epoch: 5/20... Training Step: 19039... Training loss: 1.5855... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19040... Training loss: 1.6127... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19041... Training loss: 1.6988... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19042... Training loss: 1.6156... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19043... Training loss: 1.7837... 0.0536 sec/batch
Epoch: 5/20... Training Step: 19044... Training loss: 1.5954... 0.0459 sec/batch
Epoch: 5/20... Training Step: 19045... Training loss: 1.5307... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19046... Training loss: 1.5456... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19047... Training loss: 1.5665... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19048... Training loss: 1.5604... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19049... Training loss: 1.6321... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19050... Training loss: 1.6616... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19051... Training loss: 1.6811... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19052... Training loss: 1.6307... 0.0467 sec/batch
Epoch: 5/20... Training Step: 19053... Training loss: 1.7022... 0.0507 sec/batch
Epoch: 5/20... Training Step: 19054... Training loss: 1.4760... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19055... Training loss: 1.5894... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19056... Training loss: 1.6924... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19057... Training loss: 1.7523... 0.0513 sec/batch
Epoch: 5/20... Training Step: 19058... Training loss: 1.5257... 0.0476 sec/batch
Epoch: 5/20... Training Step: 19059... Training loss: 1.6828... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19060... Training loss: 1.5360... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19061... Training loss: 1.6606... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19062... Training loss: 1.5692... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19063... Training loss: 1.6110... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19064... Training loss: 1.6545... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19065... Training loss: 1.7603... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19066... Training loss: 1.6808... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19067... Training loss: 1.5034... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19068... Training loss: 1.7543... 0.0540 sec/batch
Epoch: 5/20... Training Step: 19069... Training loss: 1.6399... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19070... Training loss: 1.6161... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19071... Training loss: 1.5305... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19072... Training loss: 1.5228... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19073... Training loss: 1.6872... 0.0497 sec/batch
Epoch: 5/20... Training Step: 19074... Training loss: 1.5858... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19075... Training loss: 1.5546... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19076... Training loss: 1.6077... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19077... Training loss: 1.6528... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19078... Training loss: 1.7120... 0.0467 sec/batch
Epoch: 5/20... Training Step: 19079... Training loss: 1.6188... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19080... Training loss: 1.8071... 0.0639 sec/batch
Epoch: 5/20... Training Step: 19081... Training loss: 1.7478... 0.0707 sec/batch
Epoch: 5/20... Training Step: 19082... Training loss: 1.6093... 0.0519 sec/batch
Epoch: 5/20... Training Step: 19083... Training loss: 1.5348... 0.0554 sec/batch
Epoch: 5/20... Training Step: 19084... Training loss: 1.7764... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19085... Training loss: 1.6354... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19086... Training loss: 1.6739... 0.0489 sec/batch
Epoch: 5/20... Training Step: 19087... Training loss: 1.6568... 0.0480 sec/batch
Epoch: 5/20... Training Step: 19088... Training loss: 1.7221... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19089... Training loss: 1.5832... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19090... Training loss: 1.5342... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19091... Training loss: 1.7299... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19092... Training loss: 1.7481... 0.0448 sec/batch
Epoch: 5/20... Training Step: 19093... Training loss: 1.7388... 0.0543 sec/batch
Epoch: 5/20... Training Step: 19094... Training loss: 1.6514... 0.0600 sec/batch
Epoch: 5/20... Training Step: 19095... Training loss: 1.6833... 0.0569 sec/batch
Epoch: 5/20... Training Step: 19096... Training loss: 1.6904... 0.0477 sec/batch
Epoch: 5/20... Training Step: 19097... Training loss: 1.5311... 0.0522 sec/batch
Epoch: 5/20... Training Step: 19098... Training loss: 1.5046... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19099... Training loss: 1.5202... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19100... Training loss: 1.6730... 0.0479 sec/batch
Epoch: 5/20... Training Step: 19101... Training loss: 1.5571... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19102... Training loss: 1.7075... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19103... Training loss: 1.6098... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19104... Training loss: 1.6397... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19105... Training loss: 1.5832... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19106... Training loss: 1.4508... 0.0538 sec/batch
Epoch: 5/20... Training Step: 19107... Training loss: 1.6823... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19108... Training loss: 1.6160... 0.0575 sec/batch
Epoch: 5/20... Training Step: 19109... Training loss: 1.5808... 0.0759 sec/batch
Epoch: 5/20... Training Step: 19110... Training loss: 1.7243... 0.0520 sec/batch
Epoch: 5/20... Training Step: 19111... Training loss: 1.7229... 0.0710 sec/batch
Epoch: 5/20... Training Step: 19112... Training loss: 1.5572... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19113... Training loss: 1.5822... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19114... Training loss: 1.5037... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19115... Training loss: 1.6703... 0.0465 sec/batch
Epoch: 5/20... Training Step: 19116... Training loss: 1.6545... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19117... Training loss: 1.5745... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19118... Training loss: 1.5556... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19119... Training loss: 1.6029... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19120... Training loss: 1.6210... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19121... Training loss: 1.6556... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19122... Training loss: 1.5400... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19123... Training loss: 1.6314... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19124... Training loss: 1.6951... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19125... Training loss: 1.7138... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19126... Training loss: 1.7183... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19127... Training loss: 1.6576... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19128... Training loss: 1.6942... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19129... Training loss: 1.7920... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19130... Training loss: 1.6975... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19131... Training loss: 1.7936... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19132... Training loss: 1.7767... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19133... Training loss: 1.8765... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19134... Training loss: 1.6513... 0.0498 sec/batch
Epoch: 5/20... Training Step: 19135... Training loss: 1.6689... 0.0451 sec/batch
Epoch: 5/20... Training Step: 19136... Training loss: 1.6359... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19137... Training loss: 1.5324... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19138... Training loss: 1.6393... 0.0391 sec/batch
Epoch: 5/20... Training Step: 19139... Training loss: 1.7968... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19140... Training loss: 1.5822... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19141... Training loss: 1.6549... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19142... Training loss: 1.6096... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19143... Training loss: 1.5943... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19144... Training loss: 1.5497... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19145... Training loss: 1.5878... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19146... Training loss: 1.6939... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19147... Training loss: 1.5474... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19148... Training loss: 1.6210... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19149... Training loss: 1.5410... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19150... Training loss: 1.7052... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19151... Training loss: 1.5988... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19152... Training loss: 1.5760... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19153... Training loss: 1.6897... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19154... Training loss: 1.5907... 0.0477 sec/batch
Epoch: 5/20... Training Step: 19155... Training loss: 1.5553... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19156... Training loss: 1.6266... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19157... Training loss: 1.6264... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19158... Training loss: 1.6108... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19159... Training loss: 1.6297... 0.0468 sec/batch
Epoch: 5/20... Training Step: 19160... Training loss: 1.5156... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19161... Training loss: 1.6393... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19162... Training loss: 1.5333... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19163... Training loss: 1.6499... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19164... Training loss: 1.5968... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19165... Training loss: 1.5616... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19166... Training loss: 1.6227... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19167... Training loss: 1.5715... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19168... Training loss: 1.6954... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19169... Training loss: 1.6972... 0.0491 sec/batch
Epoch: 5/20... Training Step: 19170... Training loss: 1.7258... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19171... Training loss: 1.7581... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19172... Training loss: 1.7368... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19173... Training loss: 1.6011... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19174... Training loss: 1.7325... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19175... Training loss: 1.6461... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19176... Training loss: 1.7203... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19177... Training loss: 1.7226... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19178... Training loss: 1.7194... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19179... Training loss: 1.6390... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19180... Training loss: 1.7534... 0.0487 sec/batch
Epoch: 5/20... Training Step: 19181... Training loss: 1.7331... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19182... Training loss: 1.7292... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19183... Training loss: 1.6710... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19184... Training loss: 1.7267... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19185... Training loss: 1.7438... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19186... Training loss: 1.6287... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19187... Training loss: 1.6178... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19188... Training loss: 1.6057... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19189... Training loss: 1.6080... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19190... Training loss: 1.7314... 0.0452 sec/batch
Epoch: 5/20... Training Step: 19191... Training loss: 1.6689... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19192... Training loss: 1.6012... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19193... Training loss: 1.6578... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19194... Training loss: 1.7673... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19195... Training loss: 1.6735... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19196... Training loss: 1.7555... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19197... Training loss: 1.6915... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19198... Training loss: 1.6836... 0.0380 sec/batch
Epoch: 5/20... Training Step: 19199... Training loss: 1.7022... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19200... Training loss: 1.6840... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19201... Training loss: 1.6256... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19202... Training loss: 1.5100... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19203... Training loss: 1.6455... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19204... Training loss: 1.5524... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19205... Training loss: 1.7124... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19206... Training loss: 1.7897... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19207... Training loss: 1.7259... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19208... Training loss: 1.6310... 0.0453 sec/batch
Epoch: 5/20... Training Step: 19209... Training loss: 1.8358... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19210... Training loss: 1.8021... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19211... Training loss: 1.5501... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19212... Training loss: 1.7961... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19213... Training loss: 1.6278... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19214... Training loss: 1.6775... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19215... Training loss: 1.8039... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19216... Training loss: 1.8383... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19217... Training loss: 1.7087... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19218... Training loss: 1.7402... 0.0451 sec/batch
Epoch: 5/20... Training Step: 19219... Training loss: 1.5169... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19220... Training loss: 1.6179... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19221... Training loss: 1.6870... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19222... Training loss: 1.5880... 0.0448 sec/batch
Epoch: 5/20... Training Step: 19223... Training loss: 1.6313... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19224... Training loss: 1.6413... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19225... Training loss: 1.6285... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19226... Training loss: 1.8241... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19227... Training loss: 1.8110... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19228... Training loss: 1.7560... 0.0502 sec/batch
Epoch: 5/20... Training Step: 19229... Training loss: 1.7653... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19230... Training loss: 1.5653... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19231... Training loss: 1.8138... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19232... Training loss: 1.5464... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19233... Training loss: 1.5616... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19234... Training loss: 1.5548... 0.0451 sec/batch
Epoch: 5/20... Training Step: 19235... Training loss: 1.5600... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19236... Training loss: 1.5771... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19237... Training loss: 1.6634... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19238... Training loss: 1.7430... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19239... Training loss: 1.6718... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19240... Training loss: 1.6164... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19241... Training loss: 1.6534... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19242... Training loss: 1.5378... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19243... Training loss: 1.5938... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19244... Training loss: 1.5599... 0.0464 sec/batch
Epoch: 5/20... Training Step: 19245... Training loss: 1.6273... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19246... Training loss: 1.5702... 0.0392 sec/batch
Epoch: 5/20... Training Step: 19247... Training loss: 1.5961... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19248... Training loss: 1.7204... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19249... Training loss: 1.7210... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19250... Training loss: 1.5899... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19251... Training loss: 1.7373... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19252... Training loss: 1.6669... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19253... Training loss: 1.6817... 0.0468 sec/batch
Epoch: 5/20... Training Step: 19254... Training loss: 1.6581... 0.0491 sec/batch
Epoch: 5/20... Training Step: 19255... Training loss: 1.6559... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19256... Training loss: 1.6494... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19257... Training loss: 1.7755... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19258... Training loss: 1.5377... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19259... Training loss: 1.4763... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19260... Training loss: 1.5481... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19261... Training loss: 1.5454... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19262... Training loss: 1.5923... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19263... Training loss: 1.5534... 0.0461 sec/batch
Epoch: 5/20... Training Step: 19264... Training loss: 1.6437... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19265... Training loss: 1.6052... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19266... Training loss: 1.5123... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19267... Training loss: 1.7472... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19268... Training loss: 1.6682... 0.0477 sec/batch
Epoch: 5/20... Training Step: 19269... Training loss: 1.5965... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19270... Training loss: 1.6818... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19271... Training loss: 1.6257... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19272... Training loss: 1.6037... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19273... Training loss: 1.5967... 0.0482 sec/batch
Epoch: 5/20... Training Step: 19274... Training loss: 1.6680... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19275... Training loss: 1.5238... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19276... Training loss: 1.7556... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19277... Training loss: 1.6863... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19278... Training loss: 1.6299... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19279... Training loss: 1.7780... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19280... Training loss: 1.6940... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19281... Training loss: 1.6509... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19282... Training loss: 1.7698... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19283... Training loss: 1.5557... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19284... Training loss: 1.5762... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19285... Training loss: 1.5793... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19286... Training loss: 1.5860... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19287... Training loss: 1.4582... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19288... Training loss: 1.5733... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19289... Training loss: 1.6145... 0.0457 sec/batch
Epoch: 5/20... Training Step: 19290... Training loss: 1.7083... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19291... Training loss: 1.6370... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19292... Training loss: 1.4720... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19293... Training loss: 1.5327... 0.0489 sec/batch
Epoch: 5/20... Training Step: 19294... Training loss: 1.5905... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19295... Training loss: 1.6129... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19296... Training loss: 1.6141... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19297... Training loss: 1.7537... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19298... Training loss: 1.7323... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19299... Training loss: 1.7169... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19300... Training loss: 1.8483... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19301... Training loss: 1.7054... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19302... Training loss: 1.6791... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19303... Training loss: 1.7224... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19304... Training loss: 1.6061... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19305... Training loss: 1.7236... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19306... Training loss: 1.5913... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19307... Training loss: 1.6069... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19308... Training loss: 1.8134... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19309... Training loss: 1.6309... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19310... Training loss: 1.5393... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19311... Training loss: 1.6832... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19312... Training loss: 1.6771... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19313... Training loss: 1.6982... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19314... Training loss: 1.7209... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19315... Training loss: 1.6694... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19316... Training loss: 1.6281... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19317... Training loss: 1.6324... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19318... Training loss: 1.6341... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19319... Training loss: 1.7609... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19320... Training loss: 1.6923... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19321... Training loss: 1.6030... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19322... Training loss: 1.5396... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19323... Training loss: 1.6493... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19324... Training loss: 1.6953... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19325... Training loss: 1.6330... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19326... Training loss: 1.6032... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19327... Training loss: 1.6466... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19328... Training loss: 1.5686... 0.0501 sec/batch
Epoch: 5/20... Training Step: 19329... Training loss: 1.6152... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19330... Training loss: 1.6069... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19331... Training loss: 1.7154... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19332... Training loss: 1.7300... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19333... Training loss: 1.6287... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19334... Training loss: 1.5161... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19335... Training loss: 1.6124... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19336... Training loss: 1.7647... 0.0387 sec/batch
Epoch: 5/20... Training Step: 19337... Training loss: 1.5829... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19338... Training loss: 1.5611... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19339... Training loss: 1.5766... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19340... Training loss: 1.6490... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19341... Training loss: 1.5969... 0.0387 sec/batch
Epoch: 5/20... Training Step: 19342... Training loss: 1.5514... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19343... Training loss: 1.8259... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19344... Training loss: 1.5942... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19345... Training loss: 1.5753... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19346... Training loss: 1.4906... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19347... Training loss: 1.4899... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19348... Training loss: 1.5699... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19349... Training loss: 1.7440... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19350... Training loss: 1.6592... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19351... Training loss: 1.5912... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19352... Training loss: 1.6296... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19353... Training loss: 1.7159... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19354... Training loss: 1.6500... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19355... Training loss: 1.7213... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19356... Training loss: 1.4826... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19357... Training loss: 1.7949... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19358... Training loss: 1.7142... 0.0480 sec/batch
Epoch: 5/20... Training Step: 19359... Training loss: 1.6314... 0.0470 sec/batch
Epoch: 5/20... Training Step: 19360... Training loss: 1.4756... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19361... Training loss: 1.6135... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19362... Training loss: 1.6518... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19363... Training loss: 1.7641... 0.0469 sec/batch
Epoch: 5/20... Training Step: 19364... Training loss: 1.7734... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19365... Training loss: 1.6570... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19366... Training loss: 1.8322... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19367... Training loss: 1.6742... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19368... Training loss: 1.7861... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19369... Training loss: 1.6912... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19370... Training loss: 1.8519... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19371... Training loss: 1.8193... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19372... Training loss: 1.6348... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19373... Training loss: 1.5922... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19374... Training loss: 1.6889... 0.0457 sec/batch
Epoch: 5/20... Training Step: 19375... Training loss: 1.7016... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19376... Training loss: 1.6474... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19377... Training loss: 1.6651... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19378... Training loss: 1.6563... 0.0464 sec/batch
Epoch: 5/20... Training Step: 19379... Training loss: 1.6829... 0.0527 sec/batch
Epoch: 5/20... Training Step: 19380... Training loss: 1.6944... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19381... Training loss: 1.5494... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19382... Training loss: 1.6062... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19383... Training loss: 1.5322... 0.0488 sec/batch
Epoch: 5/20... Training Step: 19384... Training loss: 1.6729... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19385... Training loss: 1.6246... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19386... Training loss: 1.6703... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19387... Training loss: 1.6147... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19388... Training loss: 1.6100... 0.0459 sec/batch
Epoch: 5/20... Training Step: 19389... Training loss: 1.5996... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19390... Training loss: 1.5994... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19391... Training loss: 1.6955... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19392... Training loss: 1.6270... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19393... Training loss: 1.6324... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19394... Training loss: 1.5223... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19395... Training loss: 1.4897... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19396... Training loss: 1.5329... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19397... Training loss: 1.5880... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19398... Training loss: 1.6721... 0.0455 sec/batch
Epoch: 5/20... Training Step: 19399... Training loss: 1.6549... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19400... Training loss: 1.6277... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19401... Training loss: 1.7153... 0.0517 sec/batch
Epoch: 5/20... Training Step: 19402... Training loss: 1.7519... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19403... Training loss: 1.6982... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19404... Training loss: 1.6018... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19405... Training loss: 1.3798... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19406... Training loss: 1.7436... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19407... Training loss: 1.6206... 0.0545 sec/batch
Epoch: 5/20... Training Step: 19408... Training loss: 1.6034... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19409... Training loss: 1.5898... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19410... Training loss: 1.5158... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19411... Training loss: 1.5583... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19412... Training loss: 1.6211... 0.0501 sec/batch
Epoch: 5/20... Training Step: 19413... Training loss: 1.6953... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19414... Training loss: 1.6604... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19415... Training loss: 1.6654... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19416... Training loss: 1.5576... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19417... Training loss: 1.5797... 0.0533 sec/batch
Epoch: 5/20... Training Step: 19418... Training loss: 1.6557... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19419... Training loss: 1.7086... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19420... Training loss: 1.6514... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19421... Training loss: 1.7629... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19422... Training loss: 1.5974... 0.0483 sec/batch
Epoch: 5/20... Training Step: 19423... Training loss: 1.4752... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19424... Training loss: 1.6049... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19425... Training loss: 1.6308... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19426... Training loss: 1.5649... 0.0481 sec/batch
Epoch: 5/20... Training Step: 19427... Training loss: 1.6756... 0.0480 sec/batch
Epoch: 5/20... Training Step: 19428... Training loss: 1.5268... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19429... Training loss: 1.6077... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19430... Training loss: 1.7213... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19431... Training loss: 1.5710... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19432... Training loss: 1.6285... 0.0484 sec/batch
Epoch: 5/20... Training Step: 19433... Training loss: 1.7471... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19434... Training loss: 1.7176... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19435... Training loss: 1.5140... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19436... Training loss: 1.7156... 0.0576 sec/batch
Epoch: 5/20... Training Step: 19437... Training loss: 1.5407... 0.0493 sec/batch
Epoch: 5/20... Training Step: 19438... Training loss: 1.4778... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19439... Training loss: 1.7006... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19440... Training loss: 1.6181... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19441... Training loss: 1.5630... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19442... Training loss: 1.5465... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19443... Training loss: 1.7079... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19444... Training loss: 1.5774... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19445... Training loss: 1.6329... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19446... Training loss: 1.6467... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19447... Training loss: 1.6202... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19448... Training loss: 1.5858... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19449... Training loss: 1.6615... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19450... Training loss: 1.4160... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19451... Training loss: 1.6489... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19452... Training loss: 1.7373... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19453... Training loss: 1.6270... 0.0464 sec/batch
Epoch: 5/20... Training Step: 19454... Training loss: 1.5565... 0.0460 sec/batch
Epoch: 5/20... Training Step: 19455... Training loss: 1.5634... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19456... Training loss: 1.5149... 0.0481 sec/batch
Epoch: 5/20... Training Step: 19457... Training loss: 1.6114... 0.0490 sec/batch
Epoch: 5/20... Training Step: 19458... Training loss: 1.4846... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19459... Training loss: 1.5457... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19460... Training loss: 1.5148... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19461... Training loss: 1.5802... 0.0488 sec/batch
Epoch: 5/20... Training Step: 19462... Training loss: 1.5956... 0.0471 sec/batch
Epoch: 5/20... Training Step: 19463... Training loss: 1.5566... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19464... Training loss: 1.6577... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19465... Training loss: 1.6125... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19466... Training loss: 1.7758... 0.0467 sec/batch
Epoch: 5/20... Training Step: 19467... Training loss: 1.7622... 0.0493 sec/batch
Epoch: 5/20... Training Step: 19468... Training loss: 1.5478... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19469... Training loss: 1.5428... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19470... Training loss: 1.6451... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19471... Training loss: 1.6031... 0.0477 sec/batch
Epoch: 5/20... Training Step: 19472... Training loss: 1.4993... 0.0462 sec/batch
Epoch: 5/20... Training Step: 19473... Training loss: 1.5915... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19474... Training loss: 2.1760... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19475... Training loss: 1.8173... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19476... Training loss: 1.6426... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19477... Training loss: 1.8157... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19478... Training loss: 1.7319... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19479... Training loss: 1.7284... 0.0467 sec/batch
Epoch: 5/20... Training Step: 19480... Training loss: 1.6313... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19481... Training loss: 1.6883... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19482... Training loss: 1.6324... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19483... Training loss: 1.7303... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19484... Training loss: 1.7262... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19485... Training loss: 1.6625... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19486... Training loss: 1.5965... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19487... Training loss: 1.6400... 0.0467 sec/batch
Epoch: 5/20... Training Step: 19488... Training loss: 1.6252... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19489... Training loss: 1.7531... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19490... Training loss: 1.6828... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19491... Training loss: 1.6679... 0.0479 sec/batch
Epoch: 5/20... Training Step: 19492... Training loss: 1.6509... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19493... Training loss: 1.5964... 0.0453 sec/batch
Epoch: 5/20... Training Step: 19494... Training loss: 1.6756... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19495... Training loss: 1.6556... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19496... Training loss: 1.6244... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19497... Training loss: 1.6691... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19498... Training loss: 1.7461... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19499... Training loss: 1.6348... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19500... Training loss: 1.6570... 0.0391 sec/batch
Epoch: 5/20... Training Step: 19501... Training loss: 1.5412... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19502... Training loss: 1.7353... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19503... Training loss: 1.7129... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19504... Training loss: 1.8262... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19505... Training loss: 1.6321... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19506... Training loss: 1.6051... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19507... Training loss: 1.8333... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19508... Training loss: 1.6541... 0.0471 sec/batch
Epoch: 5/20... Training Step: 19509... Training loss: 1.6869... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19510... Training loss: 1.5508... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19511... Training loss: 1.7219... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19512... Training loss: 1.8071... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19513... Training loss: 1.7484... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19514... Training loss: 1.6807... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19515... Training loss: 1.6706... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19516... Training loss: 1.7805... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19517... Training loss: 1.5570... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19518... Training loss: 1.7572... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19519... Training loss: 1.6899... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19520... Training loss: 1.6096... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19521... Training loss: 1.7396... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19522... Training loss: 1.6707... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19523... Training loss: 1.8171... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19524... Training loss: 1.7622... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19525... Training loss: 1.7239... 0.0388 sec/batch
Epoch: 5/20... Training Step: 19526... Training loss: 1.5028... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19527... Training loss: 1.6496... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19528... Training loss: 1.7499... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19529... Training loss: 1.6031... 0.0387 sec/batch
Epoch: 5/20... Training Step: 19530... Training loss: 1.7796... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19531... Training loss: 1.7248... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19532... Training loss: 1.6274... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19533... Training loss: 1.6498... 0.0448 sec/batch
Epoch: 5/20... Training Step: 19534... Training loss: 1.5924... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19535... Training loss: 1.8274... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19536... Training loss: 1.8690... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19537... Training loss: 1.6926... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19538... Training loss: 1.7597... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19539... Training loss: 1.6461... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19540... Training loss: 1.7381... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19541... Training loss: 1.7129... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19542... Training loss: 1.6410... 0.0485 sec/batch
Epoch: 5/20... Training Step: 19543... Training loss: 1.7283... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19544... Training loss: 1.8510... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19545... Training loss: 1.6982... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19546... Training loss: 1.8196... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19547... Training loss: 1.5340... 0.0485 sec/batch
Epoch: 5/20... Training Step: 19548... Training loss: 1.8085... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19549... Training loss: 1.7259... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19550... Training loss: 1.6987... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19551... Training loss: 1.5934... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19552... Training loss: 1.7227... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19553... Training loss: 1.7198... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19554... Training loss: 1.6937... 0.0388 sec/batch
Epoch: 5/20... Training Step: 19555... Training loss: 1.6262... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19556... Training loss: 1.8401... 0.0388 sec/batch
Epoch: 5/20... Training Step: 19557... Training loss: 1.7618... 0.0465 sec/batch
Epoch: 5/20... Training Step: 19558... Training loss: 1.7210... 0.0454 sec/batch
Epoch: 5/20... Training Step: 19559... Training loss: 1.7822... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19560... Training loss: 1.8410... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19561... Training loss: 1.6725... 0.0385 sec/batch
Epoch: 5/20... Training Step: 19562... Training loss: 1.7390... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19563... Training loss: 1.5627... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19564... Training loss: 1.6020... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19565... Training loss: 1.6827... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19566... Training loss: 1.7471... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19567... Training loss: 1.6869... 0.0483 sec/batch
Epoch: 5/20... Training Step: 19568... Training loss: 1.6428... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19569... Training loss: 1.6612... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19570... Training loss: 1.6771... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19571... Training loss: 1.7676... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19572... Training loss: 1.7097... 0.0473 sec/batch
Epoch: 5/20... Training Step: 19573... Training loss: 1.7894... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19574... Training loss: 1.6778... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19575... Training loss: 1.6947... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19576... Training loss: 1.8089... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19577... Training loss: 1.6351... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19578... Training loss: 1.7411... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19579... Training loss: 1.6706... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19580... Training loss: 1.5731... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19581... Training loss: 1.8492... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19582... Training loss: 1.7419... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19583... Training loss: 1.5753... 0.0453 sec/batch
Epoch: 5/20... Training Step: 19584... Training loss: 1.7996... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19585... Training loss: 1.5917... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19586... Training loss: 1.6093... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19587... Training loss: 1.6198... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19588... Training loss: 1.7340... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19589... Training loss: 1.6812... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19590... Training loss: 1.6835... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19591... Training loss: 1.6711... 0.0502 sec/batch
Epoch: 5/20... Training Step: 19592... Training loss: 1.8085... 0.0452 sec/batch
Epoch: 5/20... Training Step: 19593... Training loss: 1.6732... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19594... Training loss: 1.8071... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19595... Training loss: 1.9288... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19596... Training loss: 1.6222... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19597... Training loss: 1.7357... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19598... Training loss: 1.6444... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19599... Training loss: 1.6941... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19600... Training loss: 1.5901... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19601... Training loss: 1.5887... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19602... Training loss: 1.5358... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19603... Training loss: 1.8544... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19604... Training loss: 1.7562... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19605... Training loss: 1.6945... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19606... Training loss: 1.7837... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19607... Training loss: 1.8540... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19608... Training loss: 1.8130... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19609... Training loss: 1.8415... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19610... Training loss: 1.8088... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19611... Training loss: 1.6947... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19612... Training loss: 1.7989... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19613... Training loss: 1.5835... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19614... Training loss: 1.6016... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19615... Training loss: 1.7704... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19616... Training loss: 1.7074... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19617... Training loss: 1.6796... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19618... Training loss: 1.8645... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19619... Training loss: 1.8822... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19620... Training loss: 1.6722... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19621... Training loss: 1.7407... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19622... Training loss: 1.7988... 0.0510 sec/batch
Epoch: 5/20... Training Step: 19623... Training loss: 1.7715... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19624... Training loss: 1.8251... 0.0409 sec/batch
Epoch: 5/20... Training Step: 19625... Training loss: 1.5588... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19626... Training loss: 1.8197... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19627... Training loss: 1.7439... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19628... Training loss: 1.8824... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19629... Training loss: 1.6180... 0.0392 sec/batch
Epoch: 5/20... Training Step: 19630... Training loss: 1.6896... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19631... Training loss: 1.5601... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19632... Training loss: 1.8903... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19633... Training loss: 1.6544... 0.0456 sec/batch
Epoch: 5/20... Training Step: 19634... Training loss: 1.7065... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19635... Training loss: 1.7057... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19636... Training loss: 1.7544... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19637... Training loss: 1.7101... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19638... Training loss: 1.8816... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19639... Training loss: 1.7705... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19640... Training loss: 1.5781... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19641... Training loss: 1.6138... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19642... Training loss: 1.6283... 0.0449 sec/batch
Epoch: 5/20... Training Step: 19643... Training loss: 1.6352... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19644... Training loss: 1.7248... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19645... Training loss: 1.6274... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19646... Training loss: 1.6813... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19647... Training loss: 1.6488... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19648... Training loss: 1.7444... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19649... Training loss: 1.8063... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19650... Training loss: 1.5735... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19651... Training loss: 1.6206... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19652... Training loss: 1.5648... 0.0442 sec/batch
Epoch: 5/20... Training Step: 19653... Training loss: 1.7386... 0.0479 sec/batch
Epoch: 5/20... Training Step: 19654... Training loss: 1.6358... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19655... Training loss: 1.7593... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19656... Training loss: 1.6493... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19657... Training loss: 1.6738... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19658... Training loss: 1.7312... 0.0522 sec/batch
Epoch: 5/20... Training Step: 19659... Training loss: 1.6610... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19660... Training loss: 1.8757... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19661... Training loss: 1.6579... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19662... Training loss: 1.5579... 0.0474 sec/batch
Epoch: 5/20... Training Step: 19663... Training loss: 1.6279... 0.0472 sec/batch
Epoch: 5/20... Training Step: 19664... Training loss: 1.6850... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19665... Training loss: 1.5136... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19666... Training loss: 1.5147... 0.0459 sec/batch
Epoch: 5/20... Training Step: 19667... Training loss: 1.7293... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19668... Training loss: 1.5763... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19669... Training loss: 1.5313... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19670... Training loss: 1.5993... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19671... Training loss: 1.5740... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19672... Training loss: 1.7370... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19673... Training loss: 1.7301... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19674... Training loss: 1.8402... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19675... Training loss: 1.8152... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19676... Training loss: 1.8092... 0.0458 sec/batch
Epoch: 5/20... Training Step: 19677... Training loss: 1.8595... 0.0491 sec/batch
Epoch: 5/20... Training Step: 19678... Training loss: 1.8987... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19679... Training loss: 1.8038... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19680... Training loss: 1.6151... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19681... Training loss: 1.8099... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19682... Training loss: 1.6101... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19683... Training loss: 1.5725... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19684... Training loss: 1.6334... 0.0562 sec/batch
Epoch: 5/20... Training Step: 19685... Training loss: 1.6962... 0.0433 sec/batch
Epoch: 5/20... Training Step: 19686... Training loss: 1.5314... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19687... Training loss: 1.7656... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19688... Training loss: 1.9858... 0.0404 sec/batch
Epoch: 5/20... Training Step: 19689... Training loss: 1.6334... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19690... Training loss: 1.7635... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19691... Training loss: 1.4944... 0.0390 sec/batch
Epoch: 5/20... Training Step: 19692... Training loss: 1.7069... 0.0524 sec/batch
Epoch: 5/20... Training Step: 19693... Training loss: 1.4665... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19694... Training loss: 1.5693... 0.0392 sec/batch
Epoch: 5/20... Training Step: 19695... Training loss: 1.7898... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19696... Training loss: 2.0519... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19697... Training loss: 2.0423... 0.0446 sec/batch
Epoch: 5/20... Training Step: 19698... Training loss: 1.8669... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19699... Training loss: 1.9530... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19700... Training loss: 1.9458... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19701... Training loss: 1.8482... 0.0392 sec/batch
Epoch: 5/20... Training Step: 19702... Training loss: 1.8409... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19703... Training loss: 1.9355... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19704... Training loss: 1.7531... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19705... Training loss: 1.7768... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19706... Training loss: 1.6857... 0.0391 sec/batch
Epoch: 5/20... Training Step: 19707... Training loss: 1.5909... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19708... Training loss: 1.7437... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19709... Training loss: 1.6059... 0.0387 sec/batch
Epoch: 5/20... Training Step: 19710... Training loss: 1.6365... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19711... Training loss: 1.6528... 0.0391 sec/batch
Epoch: 5/20... Training Step: 19712... Training loss: 1.5375... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19713... Training loss: 1.6072... 0.0442 sec/batch
Epoch: 5/20... Training Step: 19714... Training loss: 1.5342... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19715... Training loss: 1.7678... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19716... Training loss: 1.5638... 0.0441 sec/batch
Epoch: 5/20... Training Step: 19717... Training loss: 1.7295... 0.0492 sec/batch
Epoch: 5/20... Training Step: 19718... Training loss: 1.6701... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19719... Training loss: 1.7754... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19720... Training loss: 1.8099... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19721... Training loss: 1.7216... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19722... Training loss: 1.8477... 0.0482 sec/batch
Epoch: 5/20... Training Step: 19723... Training loss: 1.9852... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19724... Training loss: 1.9617... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19725... Training loss: 1.8633... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19726... Training loss: 1.6942... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19727... Training loss: 1.6575... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19728... Training loss: 1.7506... 0.0432 sec/batch
Epoch: 5/20... Training Step: 19729... Training loss: 1.5643... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19730... Training loss: 1.6020... 0.0395 sec/batch
Epoch: 5/20... Training Step: 19731... Training loss: 1.8086... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19732... Training loss: 1.7930... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19733... Training loss: 1.8620... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19734... Training loss: 1.6051... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19735... Training loss: 1.8432... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19736... Training loss: 1.6332... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19737... Training loss: 1.7910... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19738... Training loss: 1.5942... 0.0431 sec/batch
Epoch: 5/20... Training Step: 19739... Training loss: 1.8389... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19740... Training loss: 1.7620... 0.0396 sec/batch
Epoch: 5/20... Training Step: 19741... Training loss: 1.7198... 0.0391 sec/batch
Epoch: 5/20... Training Step: 19742... Training loss: 1.5985... 0.0430 sec/batch
Epoch: 5/20... Training Step: 19743... Training loss: 1.6505... 0.0481 sec/batch
Epoch: 5/20... Training Step: 19744... Training loss: 1.7157... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19745... Training loss: 1.7418... 0.0425 sec/batch
Epoch: 5/20... Training Step: 19746... Training loss: 1.7132... 0.0419 sec/batch
Epoch: 5/20... Training Step: 19747... Training loss: 1.7457... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19748... Training loss: 1.6094... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19749... Training loss: 1.6835... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19750... Training loss: 1.7747... 0.0399 sec/batch
Epoch: 5/20... Training Step: 19751... Training loss: 1.6854... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19752... Training loss: 1.6821... 0.0455 sec/batch
Epoch: 5/20... Training Step: 19753... Training loss: 1.6744... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19754... Training loss: 1.6952... 0.0387 sec/batch
Epoch: 5/20... Training Step: 19755... Training loss: 1.5757... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19756... Training loss: 1.5914... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19757... Training loss: 1.7055... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19758... Training loss: 1.7052... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19759... Training loss: 1.6437... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19760... Training loss: 1.7072... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19761... Training loss: 1.7185... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19762... Training loss: 1.7771... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19763... Training loss: 1.8013... 0.0473 sec/batch
Epoch: 5/20... Training Step: 19764... Training loss: 1.6617... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19765... Training loss: 1.7149... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19766... Training loss: 1.6171... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19767... Training loss: 1.7087... 0.0518 sec/batch
Epoch: 5/20... Training Step: 19768... Training loss: 1.7843... 0.0448 sec/batch
Epoch: 5/20... Training Step: 19769... Training loss: 1.6674... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19770... Training loss: 1.7800... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19771... Training loss: 1.8125... 0.0464 sec/batch
Epoch: 5/20... Training Step: 19772... Training loss: 1.6056... 0.0497 sec/batch
Epoch: 5/20... Training Step: 19773... Training loss: 1.6830... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19774... Training loss: 1.6981... 0.0406 sec/batch
Epoch: 5/20... Training Step: 19775... Training loss: 1.7175... 0.0402 sec/batch
Epoch: 5/20... Training Step: 19776... Training loss: 1.9909... 0.0411 sec/batch
Epoch: 5/20... Training Step: 19777... Training loss: 1.8813... 0.0471 sec/batch
Epoch: 5/20... Training Step: 19778... Training loss: 1.7575... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19779... Training loss: 1.7682... 0.0443 sec/batch
Epoch: 5/20... Training Step: 19780... Training loss: 1.7884... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19781... Training loss: 1.6953... 0.0444 sec/batch
Epoch: 5/20... Training Step: 19782... Training loss: 2.0208... 0.0474 sec/batch
Epoch: 5/20... Training Step: 19783... Training loss: 1.8072... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19784... Training loss: 1.9012... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19785... Training loss: 1.9831... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19786... Training loss: 2.0095... 0.0421 sec/batch
Epoch: 5/20... Training Step: 19787... Training loss: 1.7806... 0.0488 sec/batch
Epoch: 5/20... Training Step: 19788... Training loss: 1.7892... 0.0420 sec/batch
Epoch: 5/20... Training Step: 19789... Training loss: 1.8521... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19790... Training loss: 1.9296... 0.0412 sec/batch
Epoch: 5/20... Training Step: 19791... Training loss: 1.9494... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19792... Training loss: 1.7580... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19793... Training loss: 1.6553... 0.0422 sec/batch
Epoch: 5/20... Training Step: 19794... Training loss: 1.7322... 0.0415 sec/batch
Epoch: 5/20... Training Step: 19795... Training loss: 1.5891... 0.0394 sec/batch
Epoch: 5/20... Training Step: 19796... Training loss: 1.6234... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19797... Training loss: 1.7238... 0.0517 sec/batch
Epoch: 5/20... Training Step: 19798... Training loss: 1.6981... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19799... Training loss: 1.7266... 0.0403 sec/batch
Epoch: 5/20... Training Step: 19800... Training loss: 1.9766... 0.0418 sec/batch
Epoch: 5/20... Training Step: 19801... Training loss: 1.6150... 0.0438 sec/batch
Epoch: 5/20... Training Step: 19802... Training loss: 1.8195... 0.0478 sec/batch
Epoch: 5/20... Training Step: 19803... Training loss: 1.6361... 0.0410 sec/batch
Epoch: 5/20... Training Step: 19804... Training loss: 1.8017... 0.0423 sec/batch
Epoch: 5/20... Training Step: 19805... Training loss: 1.7971... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19806... Training loss: 1.7374... 0.0442 sec/batch
Epoch: 5/20... Training Step: 19807... Training loss: 1.6921... 0.0427 sec/batch
Epoch: 5/20... Training Step: 19808... Training loss: 1.6355... 0.0407 sec/batch
Epoch: 5/20... Training Step: 19809... Training loss: 1.7187... 0.0405 sec/batch
Epoch: 5/20... Training Step: 19810... Training loss: 1.7758... 0.0413 sec/batch
Epoch: 5/20... Training Step: 19811... Training loss: 1.6646... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19812... Training loss: 1.8995... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19813... Training loss: 1.8617... 0.0450 sec/batch
Epoch: 5/20... Training Step: 19814... Training loss: 1.6663... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19815... Training loss: 1.6677... 0.0434 sec/batch
Epoch: 5/20... Training Step: 19816... Training loss: 1.6288... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19817... Training loss: 1.8045... 0.0497 sec/batch
Epoch: 5/20... Training Step: 19818... Training loss: 1.6491... 0.0429 sec/batch
Epoch: 5/20... Training Step: 19819... Training loss: 1.8808... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19820... Training loss: 1.7020... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19821... Training loss: 1.7498... 0.0463 sec/batch
Epoch: 5/20... Training Step: 19822... Training loss: 1.7337... 0.0439 sec/batch
Epoch: 5/20... Training Step: 19823... Training loss: 1.6244... 0.0424 sec/batch
Epoch: 5/20... Training Step: 19824... Training loss: 1.6558... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19825... Training loss: 1.6644... 0.0437 sec/batch
Epoch: 5/20... Training Step: 19826... Training loss: 1.9592... 0.0435 sec/batch
Epoch: 5/20... Training Step: 19827... Training loss: 1.7841... 0.0493 sec/batch
Epoch: 5/20... Training Step: 19828... Training loss: 1.7122... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19829... Training loss: 1.6626... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19830... Training loss: 1.6302... 0.0414 sec/batch
Epoch: 5/20... Training Step: 19831... Training loss: 1.7007... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19832... Training loss: 1.5742... 0.0436 sec/batch
Epoch: 5/20... Training Step: 19833... Training loss: 1.6121... 0.0408 sec/batch
Epoch: 5/20... Training Step: 19834... Training loss: 1.7100... 0.0392 sec/batch
Epoch: 5/20... Training Step: 19835... Training loss: 1.5464... 0.0416 sec/batch
Epoch: 5/20... Training Step: 19836... Training loss: 1.7524... 0.0401 sec/batch
Epoch: 5/20... Training Step: 19837... Training loss: 1.7521... 0.0445 sec/batch
Epoch: 5/20... Training Step: 19838... Training loss: 1.7636... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19839... Training loss: 1.5278... 0.0398 sec/batch
Epoch: 5/20... Training Step: 19840... Training loss: 1.7363... 0.0417 sec/batch
Epoch: 5/20... Training Step: 19841... Training loss: 1.5743... 0.0440 sec/batch
Epoch: 5/20... Training Step: 19842... Training loss: 1.7550... 0.0486 sec/batch
Epoch: 5/20... Training Step: 19843... Training loss: 1.7672... 0.0426 sec/batch
Epoch: 5/20... Training Step: 19844... Training loss: 1.6224... 0.0397 sec/batch
Epoch: 5/20... Training Step: 19845... Training loss: 1.7051... 0.0400 sec/batch
Epoch: 5/20... Training Step: 19846... Training loss: 1.6352... 0.0389 sec/batch
Epoch: 5/20... Training Step: 19847... Training loss: 1.6080... 0.0447 sec/batch
Epoch: 5/20... Training Step: 19848... Training loss: 1.6299... 0.0428 sec/batch
Epoch: 5/20... Training Step: 19849... Training loss: 1.7441... 0.0393 sec/batch
Epoch: 5/20... Training Step: 19850... Training loss: 1.8014... 0.0419 sec/batch
Epoch: 6/20... Training Step: 19851... Training loss: 1.6079... 0.0404 sec/batch
Epoch: 6/20... Training Step: 19852... Training loss: 1.6372... 0.0434 sec/batch
Epoch: 6/20... Training Step: 19853... Training loss: 1.6229... 0.0434 sec/batch
Epoch: 6/20... Training Step: 19854... Training loss: 1.5907... 0.0393 sec/batch
Epoch: 6/20... Training Step: 19855... Training loss: 1.6903... 0.0398 sec/batch
Epoch: 6/20... Training Step: 19856... Training loss: 1.6041... 0.0394 sec/batch
Epoch: 6/20... Training Step: 19857... Training loss: 1.5757... 0.0427 sec/batch
Epoch: 6/20... Training Step: 19858... Training loss: 1.6439... 0.0433 sec/batch
Epoch: 6/20... Training Step: 19859... Training loss: 1.7721... 0.0405 sec/batch
Epoch: 6/20... Training Step: 19860... Training loss: 1.6750... 0.0404 sec/batch
Epoch: 6/20... Training Step: 19861... Training loss: 1.5949... 0.0409 sec/batch
Epoch: 6/20... Training Step: 19862... Training loss: 1.5431... 0.0413 sec/batch
Epoch: 6/20... Training Step: 19863... Training loss: 1.6259... 0.0419 sec/batch
Epoch: 6/20... Training Step: 19864... Training loss: 1.5944... 0.0423 sec/batch
Epoch: 6/20... Training Step: 19865... Training loss: 1.4475... 0.0394 sec/batch
Epoch: 6/20... Training Step: 19866... Training loss: 1.5715... 0.0400 sec/batch
Epoch: 6/20... Training Step: 19867... Training loss: 1.6391... 0.0427 sec/batch
Epoch: 6/20... Training Step: 19868... Training loss: 1.5800... 0.0493 sec/batch
Epoch: 6/20... Training Step: 19869... Training loss: 1.6987... 0.0455 sec/batch
Epoch: 6/20... Training Step: 19870... Training loss: 1.7219... 0.0456 sec/batch
Epoch: 6/20... Training Step: 19871... Training loss: 1.6860... 0.0480 sec/batch
Epoch: 6/20... Training Step: 19872... Training loss: 1.5613... 0.0467 sec/batch
Epoch: 6/20... Training Step: 19873... Training loss: 1.5998... 0.0419 sec/batch
Epoch: 6/20... Training Step: 19874... Training loss: 1.5823... 0.0462 sec/batch
Epoch: 6/20... Training Step: 19875... Training loss: 1.5946... 0.0420 sec/batch
Epoch: 6/20... Training Step: 19876... Training loss: 1.7505... 0.0480 sec/batch
Epoch: 6/20... Training Step: 19877... Training loss: 1.8330... 0.0499 sec/batch
Epoch: 6/20... Training Step: 19878... Training loss: 1.8692... 0.0476 sec/batch
Epoch: 6/20... Training Step: 19879... Training loss: 1.8224... 0.0409 sec/batch
Epoch: 6/20... Training Step: 19880... Training loss: 1.6318... 0.0432 sec/batch
Epoch: 6/20... Training Step: 19881... Training loss: 1.7733... 0.0492 sec/batch
Epoch: 6/20... Training Step: 19882... Training loss: 1.7872... 0.0464 sec/batch
Epoch: 6/20... Training Step: 19883... Training loss: 1.6644... 0.0459 sec/batch
Epoch: 6/20... Training Step: 19884... Training loss: 1.6497... 0.0412 sec/batch
Epoch: 6/20... Training Step: 19885... Training loss: 1.6809... 0.0411 sec/batch
Epoch: 6/20... Training Step: 19886... Training loss: 1.5347... 0.0481 sec/batch
Epoch: 6/20... Training Step: 19887... Training loss: 1.6764... 0.0447 sec/batch
Epoch: 6/20... Training Step: 19888... Training loss: 1.7135... 0.0436 sec/batch
Epoch: 6/20... Training Step: 19889... Training loss: 1.6949... 0.0427 sec/batch
Epoch: 6/20... Training Step: 19890... Training loss: 1.7355... 0.0456 sec/batch
Epoch: 6/20... Training Step: 19891... Training loss: 1.7111... 0.0448 sec/batch
Epoch: 6/20... Training Step: 19892... Training loss: 1.5947... 0.0442 sec/batch
Epoch: 6/20... Training Step: 19893... Training loss: 1.6226... 0.0464 sec/batch
Epoch: 6/20... Training Step: 19894... Training loss: 1.6097... 0.0437 sec/batch
Epoch: 6/20... Training Step: 19895... Training loss: 1.6249... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19896... Training loss: 1.7478... 0.0410 sec/batch
Epoch: 6/20... Training Step: 19897... Training loss: 1.6077... 0.0432 sec/batch
Epoch: 6/20... Training Step: 19898... Training loss: 1.6045... 0.0468 sec/batch
Epoch: 6/20... Training Step: 19899... Training loss: 1.7017... 0.0422 sec/batch
Epoch: 6/20... Training Step: 19900... Training loss: 1.8015... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19901... Training loss: 1.7205... 0.0455 sec/batch
Epoch: 6/20... Training Step: 19902... Training loss: 1.6353... 0.0551 sec/batch
Epoch: 6/20... Training Step: 19903... Training loss: 1.7661... 0.0428 sec/batch
Epoch: 6/20... Training Step: 19904... Training loss: 1.7628... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19905... Training loss: 1.6994... 0.0446 sec/batch
Epoch: 6/20... Training Step: 19906... Training loss: 1.6645... 0.0541 sec/batch
Epoch: 6/20... Training Step: 19907... Training loss: 1.6638... 0.0461 sec/batch
Epoch: 6/20... Training Step: 19908... Training loss: 1.6371... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19909... Training loss: 1.6499... 0.0417 sec/batch
Epoch: 6/20... Training Step: 19910... Training loss: 1.6241... 0.0418 sec/batch
Epoch: 6/20... Training Step: 19911... Training loss: 1.8533... 0.0421 sec/batch
Epoch: 6/20... Training Step: 19912... Training loss: 1.8875... 0.0445 sec/batch
Epoch: 6/20... Training Step: 19913... Training loss: 1.6748... 0.0496 sec/batch
Epoch: 6/20... Training Step: 19914... Training loss: 1.4871... 0.0428 sec/batch
Epoch: 6/20... Training Step: 19915... Training loss: 1.7793... 0.0434 sec/batch
Epoch: 6/20... Training Step: 19916... Training loss: 1.7782... 0.0617 sec/batch
Epoch: 6/20... Training Step: 19917... Training loss: 1.6947... 0.0492 sec/batch
Epoch: 6/20... Training Step: 19918... Training loss: 1.7955... 0.0466 sec/batch
Epoch: 6/20... Training Step: 19919... Training loss: 1.8826... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19920... Training loss: 1.6419... 0.0406 sec/batch
Epoch: 6/20... Training Step: 19921... Training loss: 1.7234... 0.0639 sec/batch
Epoch: 6/20... Training Step: 19922... Training loss: 1.6362... 0.0492 sec/batch
Epoch: 6/20... Training Step: 19923... Training loss: 1.5972... 0.0421 sec/batch
Epoch: 6/20... Training Step: 19924... Training loss: 1.6977... 0.0443 sec/batch
Epoch: 6/20... Training Step: 19925... Training loss: 1.5190... 0.0443 sec/batch
Epoch: 6/20... Training Step: 19926... Training loss: 1.6138... 0.0509 sec/batch
Epoch: 6/20... Training Step: 19927... Training loss: 1.5531... 0.0507 sec/batch
Epoch: 6/20... Training Step: 19928... Training loss: 1.5857... 0.0417 sec/batch
Epoch: 6/20... Training Step: 19929... Training loss: 1.5516... 0.0441 sec/batch
Epoch: 6/20... Training Step: 19930... Training loss: 1.6183... 0.0458 sec/batch
Epoch: 6/20... Training Step: 19931... Training loss: 1.5505... 0.0449 sec/batch
Epoch: 6/20... Training Step: 19932... Training loss: 1.6303... 0.0507 sec/batch
Epoch: 6/20... Training Step: 19933... Training loss: 1.5845... 0.0465 sec/batch
Epoch: 6/20... Training Step: 19934... Training loss: 1.6382... 0.0417 sec/batch
Epoch: 6/20... Training Step: 19935... Training loss: 1.4825... 0.0468 sec/batch
Epoch: 6/20... Training Step: 19936... Training loss: 1.5718... 0.0431 sec/batch
Epoch: 6/20... Training Step: 19937... Training loss: 1.7097... 0.0459 sec/batch
Epoch: 6/20... Training Step: 19938... Training loss: 1.5943... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19939... Training loss: 1.5778... 0.0426 sec/batch
Epoch: 6/20... Training Step: 19940... Training loss: 1.7194... 0.0415 sec/batch
Epoch: 6/20... Training Step: 19941... Training loss: 1.6152... 0.0441 sec/batch
Epoch: 6/20... Training Step: 19942... Training loss: 1.6584... 0.0540 sec/batch
Epoch: 6/20... Training Step: 19943... Training loss: 1.5920... 0.0443 sec/batch
Epoch: 6/20... Training Step: 19944... Training loss: 1.6725... 0.0422 sec/batch
Epoch: 6/20... Training Step: 19945... Training loss: 1.5519... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19946... Training loss: 1.6426... 0.0498 sec/batch
Epoch: 6/20... Training Step: 19947... Training loss: 1.7505... 0.0450 sec/batch
Epoch: 6/20... Training Step: 19948... Training loss: 1.7195... 0.0427 sec/batch
Epoch: 6/20... Training Step: 19949... Training loss: 1.6236... 0.0424 sec/batch
Epoch: 6/20... Training Step: 19950... Training loss: 1.6304... 0.0449 sec/batch
Epoch: 6/20... Training Step: 19951... Training loss: 1.7121... 0.0437 sec/batch
Epoch: 6/20... Training Step: 19952... Training loss: 1.5524... 0.0489 sec/batch
Epoch: 6/20... Training Step: 19953... Training loss: 1.7032... 0.0431 sec/batch
Epoch: 6/20... Training Step: 19954... Training loss: 1.5588... 0.0411 sec/batch
Epoch: 6/20... Training Step: 19955... Training loss: 1.6382... 0.0421 sec/batch
Epoch: 6/20... Training Step: 19956... Training loss: 1.6155... 0.0403 sec/batch
Epoch: 6/20... Training Step: 19957... Training loss: 1.5983... 0.0519 sec/batch
Epoch: 6/20... Training Step: 19958... Training loss: 1.5488... 0.0422 sec/batch
Epoch: 6/20... Training Step: 19959... Training loss: 1.5602... 0.0420 sec/batch
Epoch: 6/20... Training Step: 19960... Training loss: 1.6797... 0.0404 sec/batch
Epoch: 6/20... Training Step: 19961... Training loss: 1.5290... 0.0414 sec/batch
Epoch: 6/20... Training Step: 19962... Training loss: 1.6134... 0.0491 sec/batch
Epoch: 6/20... Training Step: 19963... Training loss: 1.5768... 0.0415 sec/batch
Epoch: 6/20... Training Step: 19964... Training loss: 1.6206... 0.0417 sec/batch
Epoch: 6/20... Training Step: 19965... Training loss: 1.5612... 0.0418 sec/batch
Epoch: 6/20... Training Step: 19966... Training loss: 1.6766... 0.0425 sec/batch
Epoch: 6/20... Training Step: 19967... Training loss: 1.5846... 0.0502 sec/batch
Epoch: 6/20... Training Step: 19968... Training loss: 1.5303... 0.0435 sec/batch
Epoch: 6/20... Training Step: 19969... Training loss: 1.5521... 0.0435 sec/batch
Epoch: 6/20... Training Step: 19970... Training loss: 1.5880... 0.0410 sec/batch
Epoch: 6/20... Training Step: 19971... Training loss: 1.5510... 0.0502 sec/batch
Epoch: 6/20... Training Step: 19972... Training loss: 1.5652... 0.0410 sec/batch
Epoch: 6/20... Training Step: 19973... Training loss: 1.6388... 0.0434 sec/batch
Epoch: 6/20... Training Step: 19974... Training loss: 1.5570... 0.0437 sec/batch
Epoch: 6/20... Training Step: 19975... Training loss: 1.6369... 0.0422 sec/batch
Epoch: 6/20... Training Step: 19976... Training loss: 1.6551... 0.0395 sec/batch
Epoch: 6/20... Training Step: 19977... Training loss: 1.6162... 0.0412 sec/batch
Epoch: 6/20... Training Step: 19978... Training loss: 1.5582... 0.0468 sec/batch
Epoch: 6/20... Training Step: 19979... Training loss: 1.7264... 0.0424 sec/batch
Epoch: 6/20... Training Step: 19980... Training loss: 1.6076... 0.0412 sec/batch
Epoch: 6/20... Training Step: 19981... Training loss: 1.6558... 0.0422 sec/batch
Epoch: 6/20... Training Step: 19982... Training loss: 1.6202... 0.0445 sec/batch
Epoch: 6/20... Training Step: 19983... Training loss: 1.6694... 0.0431 sec/batch
Epoch: 6/20... Training Step: 19984... Training loss: 1.5711... 0.0417 sec/batch
Epoch: 6/20... Training Step: 19985... Training loss: 1.7082... 0.0414 sec/batch
Epoch: 6/20... Training Step: 19986... Training loss: 1.5386... 0.0413 sec/batch
Epoch: 6/20... Training Step: 19987... Training loss: 1.7103... 0.0495 sec/batch
Epoch: 6/20... Training Step: 19988... Training loss: 1.5056... 0.0431 sec/batch
Epoch: 6/20... Training Step: 19989... Training loss: 1.6179... 0.0409 sec/batch
Epoch: 6/20... Training Step: 19990... Training loss: 1.5179... 0.0403 sec/batch
Epoch: 6/20... Training Step: 19991... Training loss: 1.6466... 0.0426 sec/batch
Epoch: 6/20... Training Step: 19992... Training loss: 1.6029... 0.0462 sec/batch
Epoch: 6/20... Training Step: 19993... Training loss: 1.6612... 0.0415 sec/batch
Epoch: 6/20... Training Step: 19994... Training loss: 1.7087... 0.0406 sec/batch
Epoch: 6/20... Training Step: 19995... Training loss: 1.5616... 0.0396 sec/batch
Epoch: 6/20... Training Step: 19996... Training loss: 1.7022... 0.0402 sec/batch
Epoch: 6/20... Training Step: 19997... Training loss: 1.7139... 0.0484 sec/batch
Epoch: 6/20... Training Step: 19998... Training loss: 1.5598... 0.0436 sec/batch
Epoch: 6/20... Training Step: 19999... Training loss: 1.5071... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20000... Training loss: 1.4771... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20001... Training loss: 1.7146... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20002... Training loss: 1.6105... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20003... Training loss: 1.5071... 0.0393 sec/batch
Epoch: 6/20... Training Step: 20004... Training loss: 1.5736... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20005... Training loss: 1.7859... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20006... Training loss: 1.6427... 0.0389 sec/batch
Epoch: 6/20... Training Step: 20007... Training loss: 1.4172... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20008... Training loss: 1.7863... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20009... Training loss: 1.9041... 0.0422 sec/batch
Epoch: 6/20... Training Step: 20010... Training loss: 1.8039... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20011... Training loss: 1.7054... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20012... Training loss: 1.6289... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20013... Training loss: 1.6421... 0.0423 sec/batch
Epoch: 6/20... Training Step: 20014... Training loss: 1.6943... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20015... Training loss: 1.8121... 0.0401 sec/batch
Epoch: 6/20... Training Step: 20016... Training loss: 1.6245... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20017... Training loss: 1.6397... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20018... Training loss: 1.8073... 0.0478 sec/batch
Epoch: 6/20... Training Step: 20019... Training loss: 1.6220... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20020... Training loss: 1.6872... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20021... Training loss: 1.6270... 0.0412 sec/batch
Epoch: 6/20... Training Step: 20022... Training loss: 1.7437... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20023... Training loss: 1.6759... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20024... Training loss: 1.6645... 0.0406 sec/batch
Epoch: 6/20... Training Step: 20025... Training loss: 1.6234... 0.0398 sec/batch
Epoch: 6/20... Training Step: 20026... Training loss: 1.6075... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20027... Training loss: 1.6141... 0.0452 sec/batch
Epoch: 6/20... Training Step: 20028... Training loss: 1.5691... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20029... Training loss: 1.5988... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20030... Training loss: 1.5585... 0.0408 sec/batch
Epoch: 6/20... Training Step: 20031... Training loss: 1.6456... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20032... Training loss: 1.6882... 0.0445 sec/batch
Epoch: 6/20... Training Step: 20033... Training loss: 1.6198... 0.0438 sec/batch
Epoch: 6/20... Training Step: 20034... Training loss: 1.5195... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20035... Training loss: 1.5656... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20036... Training loss: 1.6267... 0.0414 sec/batch
Epoch: 6/20... Training Step: 20037... Training loss: 1.4394... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20038... Training loss: 1.6162... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20039... Training loss: 1.6141... 0.0393 sec/batch
Epoch: 6/20... Training Step: 20040... Training loss: 1.7382... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20041... Training loss: 1.6346... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20042... Training loss: 1.5491... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20043... Training loss: 1.6116... 0.0444 sec/batch
Epoch: 6/20... Training Step: 20044... Training loss: 1.7176... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20045... Training loss: 1.6605... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20046... Training loss: 1.5756... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20047... Training loss: 1.4569... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20048... Training loss: 1.7190... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20049... Training loss: 1.5854... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20050... Training loss: 1.6748... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20051... Training loss: 1.5940... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20052... Training loss: 1.6955... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20053... Training loss: 1.6429... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20054... Training loss: 1.7897... 0.0438 sec/batch
Epoch: 6/20... Training Step: 20055... Training loss: 1.6623... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20056... Training loss: 1.6310... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20057... Training loss: 1.5868... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20058... Training loss: 1.6405... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20059... Training loss: 1.7608... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20060... Training loss: 1.6251... 0.0414 sec/batch
Epoch: 6/20... Training Step: 20061... Training loss: 1.6128... 0.0401 sec/batch
Epoch: 6/20... Training Step: 20062... Training loss: 1.6832... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20063... Training loss: 1.6696... 0.0428 sec/batch
Epoch: 6/20... Training Step: 20064... Training loss: 1.7240... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20065... Training loss: 1.6320... 0.0384 sec/batch
Epoch: 6/20... Training Step: 20066... Training loss: 1.6698... 0.0401 sec/batch
Epoch: 6/20... Training Step: 20067... Training loss: 1.8041... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20068... Training loss: 1.7324... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20069... Training loss: 1.7032... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20070... Training loss: 1.6276... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20071... Training loss: 1.6540... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20072... Training loss: 1.6793... 0.0455 sec/batch
Epoch: 6/20... Training Step: 20073... Training loss: 1.7892... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20074... Training loss: 1.7544... 0.0406 sec/batch
Epoch: 6/20... Training Step: 20075... Training loss: 1.6434... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20076... Training loss: 1.7219... 0.0408 sec/batch
Epoch: 6/20... Training Step: 20077... Training loss: 1.8059... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20078... Training loss: 1.8384... 0.0478 sec/batch
Epoch: 6/20... Training Step: 20079... Training loss: 1.6035... 0.0391 sec/batch
Epoch: 6/20... Training Step: 20080... Training loss: 1.5563... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20081... Training loss: 1.6530... 0.0391 sec/batch
Epoch: 6/20... Training Step: 20082... Training loss: 1.6252... 0.0466 sec/batch
Epoch: 6/20... Training Step: 20083... Training loss: 1.5573... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20084... Training loss: 1.6844... 0.0401 sec/batch
Epoch: 6/20... Training Step: 20085... Training loss: 1.6784... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20086... Training loss: 1.7735... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20087... Training loss: 1.7222... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20088... Training loss: 1.7602... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20089... Training loss: 1.6920... 0.0392 sec/batch
Epoch: 6/20... Training Step: 20090... Training loss: 1.6453... 0.0398 sec/batch
Epoch: 6/20... Training Step: 20091... Training loss: 1.6665... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20092... Training loss: 1.8605... 0.0475 sec/batch
Epoch: 6/20... Training Step: 20093... Training loss: 1.6786... 0.0451 sec/batch
Epoch: 6/20... Training Step: 20094... Training loss: 1.7574... 0.0396 sec/batch
Epoch: 6/20... Training Step: 20095... Training loss: 1.7570... 0.0398 sec/batch
Epoch: 6/20... Training Step: 20096... Training loss: 1.7766... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20097... Training loss: 1.6920... 0.0438 sec/batch
Epoch: 6/20... Training Step: 20098... Training loss: 1.7006... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20099... Training loss: 1.5248... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20100... Training loss: 1.6701... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20101... Training loss: 1.6751... 0.0396 sec/batch
Epoch: 6/20... Training Step: 20102... Training loss: 1.7507... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20103... Training loss: 1.7101... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20104... Training loss: 1.6599... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20105... Training loss: 1.8178... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20106... Training loss: 1.5972... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20107... Training loss: 1.6641... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20108... Training loss: 1.6152... 0.0428 sec/batch
Epoch: 6/20... Training Step: 20109... Training loss: 1.5968... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20110... Training loss: 1.6209... 0.0393 sec/batch
Epoch: 6/20... Training Step: 20111... Training loss: 1.6540... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20112... Training loss: 1.5731... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20113... Training loss: 1.7330... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20114... Training loss: 1.6461... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20115... Training loss: 1.5278... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20116... Training loss: 1.6142... 0.0396 sec/batch
Epoch: 6/20... Training Step: 20117... Training loss: 1.5657... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20118... Training loss: 1.6648... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20119... Training loss: 1.6964... 0.0391 sec/batch
Epoch: 6/20... Training Step: 20120... Training loss: 1.7294... 0.0408 sec/batch
Epoch: 6/20... Training Step: 20121... Training loss: 1.6517... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20122... Training loss: 1.5897... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20123... Training loss: 1.8117... 0.0452 sec/batch
Epoch: 6/20... Training Step: 20124... Training loss: 1.7443... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20125... Training loss: 1.7546... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20126... Training loss: 1.6397... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20127... Training loss: 1.5781... 0.0511 sec/batch
Epoch: 6/20... Training Step: 20128... Training loss: 1.7445... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20129... Training loss: 1.6655... 0.0401 sec/batch
Epoch: 6/20... Training Step: 20130... Training loss: 1.6864... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20131... Training loss: 1.5673... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20132... Training loss: 1.6295... 0.0444 sec/batch
Epoch: 6/20... Training Step: 20133... Training loss: 1.5855... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20134... Training loss: 1.7129... 0.0392 sec/batch
Epoch: 6/20... Training Step: 20135... Training loss: 1.5423... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20136... Training loss: 1.5836... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20137... Training loss: 1.5446... 0.0451 sec/batch
Epoch: 6/20... Training Step: 20138... Training loss: 1.6761... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20139... Training loss: 1.5641... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20140... Training loss: 1.5974... 0.0412 sec/batch
Epoch: 6/20... Training Step: 20141... Training loss: 1.6940... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20142... Training loss: 1.5950... 0.0497 sec/batch
Epoch: 6/20... Training Step: 20143... Training loss: 1.5431... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20144... Training loss: 1.5331... 0.0402 sec/batch
Epoch: 6/20... Training Step: 20145... Training loss: 1.7016... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20146... Training loss: 1.7162... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20147... Training loss: 1.4983... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20148... Training loss: 1.6785... 0.0444 sec/batch
Epoch: 6/20... Training Step: 20149... Training loss: 1.5079... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20150... Training loss: 1.5452... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20151... Training loss: 1.6953... 0.0423 sec/batch
Epoch: 6/20... Training Step: 20152... Training loss: 1.5434... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20153... Training loss: 1.6428... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20154... Training loss: 1.6245... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20155... Training loss: 1.7396... 0.0399 sec/batch
Epoch: 6/20... Training Step: 20156... Training loss: 1.5269... 0.0451 sec/batch
Epoch: 6/20... Training Step: 20157... Training loss: 1.7483... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20158... Training loss: 1.6541... 0.0451 sec/batch
Epoch: 6/20... Training Step: 20159... Training loss: 1.6974... 0.0409 sec/batch
Epoch: 6/20... Training Step: 20160... Training loss: 1.7943... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20161... Training loss: 1.6362... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20162... Training loss: 1.6697... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20163... Training loss: 1.6201... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20164... Training loss: 1.6428... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20165... Training loss: 1.7876... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20166... Training loss: 1.7779... 0.0415 sec/batch
Epoch: 6/20... Training Step: 20167... Training loss: 1.6088... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20168... Training loss: 1.5580... 0.0481 sec/batch
Epoch: 6/20... Training Step: 20169... Training loss: 1.7556... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20170... Training loss: 1.7393... 0.0391 sec/batch
Epoch: 6/20... Training Step: 20171... Training loss: 1.6555... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20172... Training loss: 1.4645... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20173... Training loss: 1.7529... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20174... Training loss: 1.6492... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20175... Training loss: 1.5882... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20176... Training loss: 1.7694... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20177... Training loss: 1.7459... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20178... Training loss: 1.6242... 0.0444 sec/batch
Epoch: 6/20... Training Step: 20179... Training loss: 1.7656... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20180... Training loss: 1.6485... 0.0398 sec/batch
Epoch: 6/20... Training Step: 20181... Training loss: 1.7170... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20182... Training loss: 1.7435... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20183... Training loss: 1.6706... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20184... Training loss: 1.7615... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20185... Training loss: 1.7817... 0.0396 sec/batch
Epoch: 6/20... Training Step: 20186... Training loss: 1.7450... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20187... Training loss: 1.5548... 0.0411 sec/batch
Epoch: 6/20... Training Step: 20188... Training loss: 1.5462... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20189... Training loss: 1.6768... 0.0406 sec/batch
Epoch: 6/20... Training Step: 20190... Training loss: 1.6617... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20191... Training loss: 1.8957... 0.0402 sec/batch
Epoch: 6/20... Training Step: 20192... Training loss: 1.9604... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20193... Training loss: 1.8751... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20194... Training loss: 1.6723... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20195... Training loss: 1.6796... 0.0412 sec/batch
Epoch: 6/20... Training Step: 20196... Training loss: 1.7646... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20197... Training loss: 1.8041... 0.0454 sec/batch
Epoch: 6/20... Training Step: 20198... Training loss: 1.7363... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20199... Training loss: 1.7021... 0.0397 sec/batch
Epoch: 6/20... Training Step: 20200... Training loss: 1.7080... 0.0414 sec/batch
Epoch: 6/20... Training Step: 20201... Training loss: 1.6140... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20202... Training loss: 1.5524... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20203... Training loss: 1.6361... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20204... Training loss: 1.7582... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20205... Training loss: 1.6486... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20206... Training loss: 1.6126... 0.0402 sec/batch
Epoch: 6/20... Training Step: 20207... Training loss: 1.7317... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20208... Training loss: 1.7134... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20209... Training loss: 1.5226... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20210... Training loss: 1.7471... 0.0421 sec/batch
Epoch: 6/20... Training Step: 20211... Training loss: 1.7123... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20212... Training loss: 1.6562... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20213... Training loss: 1.7659... 0.0454 sec/batch
Epoch: 6/20... Training Step: 20214... Training loss: 1.5907... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20215... Training loss: 1.7785... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20216... Training loss: 1.6703... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20217... Training loss: 1.6929... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20218... Training loss: 1.7130... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20219... Training loss: 1.7268... 0.0412 sec/batch
Epoch: 6/20... Training Step: 20220... Training loss: 1.7289... 0.0395 sec/batch
Epoch: 6/20... Training Step: 20221... Training loss: 1.6308... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20222... Training loss: 1.5893... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20223... Training loss: 1.7309... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20224... Training loss: 1.5508... 0.0389 sec/batch
Epoch: 6/20... Training Step: 20225... Training loss: 1.7380... 0.0404 sec/batch
Epoch: 6/20... Training Step: 20226... Training loss: 1.6323... 0.0398 sec/batch
Epoch: 6/20... Training Step: 20227... Training loss: 1.7646... 0.0439 sec/batch
Epoch: 6/20... Training Step: 20228... Training loss: 1.6578... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20229... Training loss: 1.6669... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20230... Training loss: 1.6901... 0.0399 sec/batch
Epoch: 6/20... Training Step: 20231... Training loss: 1.6736... 0.0409 sec/batch
Epoch: 6/20... Training Step: 20232... Training loss: 1.7201... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20233... Training loss: 1.8283... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20234... Training loss: 1.8778... 0.0414 sec/batch
Epoch: 6/20... Training Step: 20235... Training loss: 1.6696... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20236... Training loss: 1.6782... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20237... Training loss: 1.6595... 0.0542 sec/batch
Epoch: 6/20... Training Step: 20238... Training loss: 1.7148... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20239... Training loss: 1.5713... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20240... Training loss: 1.6336... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20241... Training loss: 1.6508... 0.0506 sec/batch
Epoch: 6/20... Training Step: 20242... Training loss: 1.8146... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20243... Training loss: 1.6502... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20244... Training loss: 1.5947... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20245... Training loss: 1.7055... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20246... Training loss: 1.6710... 0.0486 sec/batch
Epoch: 6/20... Training Step: 20247... Training loss: 1.6386... 0.0530 sec/batch
Epoch: 6/20... Training Step: 20248... Training loss: 1.6843... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20249... Training loss: 1.6209... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20250... Training loss: 1.6526... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20251... Training loss: 1.5292... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20252... Training loss: 1.6291... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20253... Training loss: 1.6321... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20254... Training loss: 1.5936... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20255... Training loss: 1.7109... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20256... Training loss: 1.6467... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20257... Training loss: 1.6614... 0.0538 sec/batch
Epoch: 6/20... Training Step: 20258... Training loss: 1.7999... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20259... Training loss: 1.5945... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20260... Training loss: 1.6032... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20261... Training loss: 1.6878... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20262... Training loss: 1.6374... 0.0568 sec/batch
Epoch: 6/20... Training Step: 20263... Training loss: 1.8893... 0.0540 sec/batch
Epoch: 6/20... Training Step: 20264... Training loss: 1.6891... 0.0519 sec/batch
Epoch: 6/20... Training Step: 20265... Training loss: 1.6965... 0.0524 sec/batch
Epoch: 6/20... Training Step: 20266... Training loss: 1.7637... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20267... Training loss: 1.8646... 0.0525 sec/batch
Epoch: 6/20... Training Step: 20268... Training loss: 1.5629... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20269... Training loss: 1.6531... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20270... Training loss: 1.6483... 0.0452 sec/batch
Epoch: 6/20... Training Step: 20271... Training loss: 1.6209... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20272... Training loss: 1.6685... 0.0449 sec/batch
Epoch: 6/20... Training Step: 20273... Training loss: 1.6132... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20274... Training loss: 1.6103... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20275... Training loss: 1.6686... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20276... Training loss: 1.7050... 0.0531 sec/batch
Epoch: 6/20... Training Step: 20277... Training loss: 1.7284... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20278... Training loss: 1.6362... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20279... Training loss: 1.6625... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20280... Training loss: 1.6199... 0.0758 sec/batch
Epoch: 6/20... Training Step: 20281... Training loss: 1.5692... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20282... Training loss: 1.6943... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20283... Training loss: 1.5845... 0.0654 sec/batch
Epoch: 6/20... Training Step: 20284... Training loss: 1.6273... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20285... Training loss: 1.5975... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20286... Training loss: 1.6163... 0.0530 sec/batch
Epoch: 6/20... Training Step: 20287... Training loss: 1.7106... 0.0447 sec/batch
Epoch: 6/20... Training Step: 20288... Training loss: 1.6039... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20289... Training loss: 1.5710... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20290... Training loss: 1.6660... 0.0518 sec/batch
Epoch: 6/20... Training Step: 20291... Training loss: 1.8004... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20292... Training loss: 1.6173... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20293... Training loss: 1.7168... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20294... Training loss: 1.6287... 0.0517 sec/batch
Epoch: 6/20... Training Step: 20295... Training loss: 1.6810... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20296... Training loss: 1.7370... 0.0447 sec/batch
Epoch: 6/20... Training Step: 20297... Training loss: 1.8044... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20298... Training loss: 1.5874... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20299... Training loss: 1.5480... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20300... Training loss: 1.7228... 0.0576 sec/batch
Epoch: 6/20... Training Step: 20301... Training loss: 1.6484... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20302... Training loss: 1.6232... 0.0441 sec/batch
Epoch: 6/20... Training Step: 20303... Training loss: 1.5395... 0.0534 sec/batch
Epoch: 6/20... Training Step: 20304... Training loss: 1.6509... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20305... Training loss: 1.5921... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20306... Training loss: 1.5427... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20307... Training loss: 1.6119... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20308... Training loss: 1.7277... 0.0438 sec/batch
Epoch: 6/20... Training Step: 20309... Training loss: 1.7596... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20310... Training loss: 1.7651... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20311... Training loss: 1.6411... 0.0446 sec/batch
Epoch: 6/20... Training Step: 20312... Training loss: 1.7330... 0.0487 sec/batch
Epoch: 6/20... Training Step: 20313... Training loss: 1.6609... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20314... Training loss: 1.6133... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20315... Training loss: 1.6305... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20316... Training loss: 1.6969... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20317... Training loss: 1.6604... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20318... Training loss: 1.6403... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20319... Training loss: 1.5928... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20320... Training loss: 1.6891... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20321... Training loss: 1.5616... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20322... Training loss: 1.6275... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20323... Training loss: 1.6588... 0.0408 sec/batch
Epoch: 6/20... Training Step: 20324... Training loss: 1.7580... 0.0520 sec/batch
Epoch: 6/20... Training Step: 20325... Training loss: 1.6219... 0.0591 sec/batch
Epoch: 6/20... Training Step: 20326... Training loss: 1.5250... 0.0513 sec/batch
Epoch: 6/20... Training Step: 20327... Training loss: 1.6221... 0.0586 sec/batch
Epoch: 6/20... Training Step: 20328... Training loss: 1.5309... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20329... Training loss: 1.5970... 0.0578 sec/batch
Epoch: 6/20... Training Step: 20330... Training loss: 1.6994... 0.0610 sec/batch
Epoch: 6/20... Training Step: 20331... Training loss: 1.5912... 0.0528 sec/batch
Epoch: 6/20... Training Step: 20332... Training loss: 1.5560... 0.0511 sec/batch
Epoch: 6/20... Training Step: 20333... Training loss: 1.6692... 0.0534 sec/batch
Epoch: 6/20... Training Step: 20334... Training loss: 1.6001... 0.0539 sec/batch
Epoch: 6/20... Training Step: 20335... Training loss: 1.7673... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20336... Training loss: 1.7425... 0.0570 sec/batch
Epoch: 6/20... Training Step: 20337... Training loss: 1.7940... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20338... Training loss: 1.6236... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20339... Training loss: 1.7133... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20340... Training loss: 1.7012... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20341... Training loss: 1.6289... 0.0553 sec/batch
Epoch: 6/20... Training Step: 20342... Training loss: 1.6582... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20343... Training loss: 1.7856... 0.0444 sec/batch
Epoch: 6/20... Training Step: 20344... Training loss: 1.7213... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20345... Training loss: 1.5884... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20346... Training loss: 1.6246... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20347... Training loss: 1.5148... 0.0406 sec/batch
Epoch: 6/20... Training Step: 20348... Training loss: 1.6550... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20349... Training loss: 1.4836... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20350... Training loss: 1.7554... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20351... Training loss: 1.5850... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20352... Training loss: 1.7229... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20353... Training loss: 1.6679... 0.0445 sec/batch
Epoch: 6/20... Training Step: 20354... Training loss: 1.6348... 0.0422 sec/batch
Epoch: 6/20... Training Step: 20355... Training loss: 1.6205... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20356... Training loss: 1.6736... 0.0554 sec/batch
Epoch: 6/20... Training Step: 20357... Training loss: 1.7097... 0.0585 sec/batch
Epoch: 6/20... Training Step: 20358... Training loss: 1.8197... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20359... Training loss: 1.8031... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20360... Training loss: 1.6485... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20361... Training loss: 1.6507... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20362... Training loss: 1.8722... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20363... Training loss: 1.6258... 0.0520 sec/batch
Epoch: 6/20... Training Step: 20364... Training loss: 1.5411... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20365... Training loss: 1.7345... 0.0508 sec/batch
Epoch: 6/20... Training Step: 20366... Training loss: 1.7889... 0.0428 sec/batch
Epoch: 6/20... Training Step: 20367... Training loss: 1.7188... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20368... Training loss: 1.6073... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20369... Training loss: 1.6565... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20370... Training loss: 1.5801... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20371... Training loss: 1.6305... 0.0402 sec/batch
Epoch: 6/20... Training Step: 20372... Training loss: 1.6998... 0.0400 sec/batch
Epoch: 6/20... Training Step: 20373... Training loss: 1.6783... 0.0413 sec/batch
Epoch: 6/20... Training Step: 20374... Training loss: 1.6319... 0.0447 sec/batch
Epoch: 6/20... Training Step: 20375... Training loss: 1.6323... 0.0850 sec/batch
Epoch: 6/20... Training Step: 20376... Training loss: 1.5945... 0.0648 sec/batch
Epoch: 6/20... Training Step: 20377... Training loss: 1.6120... 0.0636 sec/batch
Epoch: 6/20... Training Step: 20378... Training loss: 1.6871... 0.0523 sec/batch
Epoch: 6/20... Training Step: 20379... Training loss: 1.7223... 0.0635 sec/batch
Epoch: 6/20... Training Step: 20380... Training loss: 1.6591... 0.0486 sec/batch
Epoch: 6/20... Training Step: 20381... Training loss: 1.7332... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20382... Training loss: 1.6576... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20383... Training loss: 1.5811... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20384... Training loss: 1.6176... 0.0487 sec/batch
Epoch: 6/20... Training Step: 20385... Training loss: 1.6928... 0.0643 sec/batch
Epoch: 6/20... Training Step: 20386... Training loss: 1.6306... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20387... Training loss: 1.4409... 0.0563 sec/batch
Epoch: 6/20... Training Step: 20388... Training loss: 1.5426... 0.0553 sec/batch
Epoch: 6/20... Training Step: 20389... Training loss: 1.4913... 0.0550 sec/batch
Epoch: 6/20... Training Step: 20390... Training loss: 1.5289... 0.0672 sec/batch
Epoch: 6/20... Training Step: 20391... Training loss: 1.5433... 0.0449 sec/batch
Epoch: 6/20... Training Step: 20392... Training loss: 1.6258... 0.0643 sec/batch
Epoch: 6/20... Training Step: 20393... Training loss: 1.6595... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20394... Training loss: 1.5557... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20395... Training loss: 1.6888... 0.0488 sec/batch
Epoch: 6/20... Training Step: 20396... Training loss: 1.5822... 0.0517 sec/batch
Epoch: 6/20... Training Step: 20397... Training loss: 1.6562... 0.0533 sec/batch
Epoch: 6/20... Training Step: 20398... Training loss: 1.5345... 0.0715 sec/batch
Epoch: 6/20... Training Step: 20399... Training loss: 1.5775... 0.0516 sec/batch
Epoch: 6/20... Training Step: 20400... Training loss: 1.4924... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20401... Training loss: 1.5671... 0.0525 sec/batch
Epoch: 6/20... Training Step: 20402... Training loss: 1.6461... 0.0530 sec/batch
Epoch: 6/20... Training Step: 20403... Training loss: 1.5830... 0.0422 sec/batch
Epoch: 6/20... Training Step: 20404... Training loss: 1.6364... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20405... Training loss: 1.7187... 0.0441 sec/batch
Epoch: 6/20... Training Step: 20406... Training loss: 1.6456... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20407... Training loss: 1.6066... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20408... Training loss: 1.5050... 0.0483 sec/batch
Epoch: 6/20... Training Step: 20409... Training loss: 1.6556... 0.0452 sec/batch
Epoch: 6/20... Training Step: 20410... Training loss: 1.6359... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20411... Training loss: 1.5946... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20412... Training loss: 1.7081... 0.0542 sec/batch
Epoch: 6/20... Training Step: 20413... Training loss: 1.5581... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20414... Training loss: 1.4879... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20415... Training loss: 1.5395... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20416... Training loss: 1.6226... 0.0819 sec/batch
Epoch: 6/20... Training Step: 20417... Training loss: 1.5381... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20418... Training loss: 1.7317... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20419... Training loss: 1.6403... 0.0455 sec/batch
Epoch: 6/20... Training Step: 20420... Training loss: 1.7076... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20421... Training loss: 1.6630... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20422... Training loss: 1.6437... 0.0478 sec/batch
Epoch: 6/20... Training Step: 20423... Training loss: 1.4887... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20424... Training loss: 1.5411... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20425... Training loss: 1.5413... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20426... Training loss: 1.5718... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20427... Training loss: 1.6622... 0.0530 sec/batch
Epoch: 6/20... Training Step: 20428... Training loss: 1.7416... 0.0475 sec/batch
Epoch: 6/20... Training Step: 20429... Training loss: 1.6550... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20430... Training loss: 1.8395... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20431... Training loss: 1.7876... 0.0495 sec/batch
Epoch: 6/20... Training Step: 20432... Training loss: 1.5702... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20433... Training loss: 1.7394... 0.0487 sec/batch
Epoch: 6/20... Training Step: 20434... Training loss: 1.7384... 0.0516 sec/batch
Epoch: 6/20... Training Step: 20435... Training loss: 1.4800... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20436... Training loss: 1.7593... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20437... Training loss: 1.5239... 0.0532 sec/batch
Epoch: 6/20... Training Step: 20438... Training loss: 1.6164... 0.0439 sec/batch
Epoch: 6/20... Training Step: 20439... Training loss: 1.9574... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20440... Training loss: 1.7492... 0.0437 sec/batch
Epoch: 6/20... Training Step: 20441... Training loss: 1.8111... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20442... Training loss: 1.5955... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20443... Training loss: 1.6239... 0.0407 sec/batch
Epoch: 6/20... Training Step: 20444... Training loss: 1.8219... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20445... Training loss: 1.8073... 0.0631 sec/batch
Epoch: 6/20... Training Step: 20446... Training loss: 1.7796... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20447... Training loss: 1.4986... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20448... Training loss: 1.7198... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20449... Training loss: 1.6673... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20450... Training loss: 1.6009... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20451... Training loss: 1.8179... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20452... Training loss: 1.6354... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20453... Training loss: 1.6148... 0.0481 sec/batch
Epoch: 6/20... Training Step: 20454... Training loss: 1.6041... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20455... Training loss: 1.6175... 0.0447 sec/batch
Epoch: 6/20... Training Step: 20456... Training loss: 1.5833... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20457... Training loss: 1.5519... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20458... Training loss: 1.7579... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20459... Training loss: 1.7340... 0.0595 sec/batch
Epoch: 6/20... Training Step: 20460... Training loss: 1.6379... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20461... Training loss: 1.7182... 0.0497 sec/batch
Epoch: 6/20... Training Step: 20462... Training loss: 1.5917... 0.0497 sec/batch
Epoch: 6/20... Training Step: 20463... Training loss: 1.6563... 0.0509 sec/batch
Epoch: 6/20... Training Step: 20464... Training loss: 1.7866... 0.0521 sec/batch
Epoch: 6/20... Training Step: 20465... Training loss: 1.6519... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20466... Training loss: 1.6841... 0.0562 sec/batch
Epoch: 6/20... Training Step: 20467... Training loss: 1.5532... 0.0505 sec/batch
Epoch: 6/20... Training Step: 20468... Training loss: 1.5184... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20469... Training loss: 1.5657... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20470... Training loss: 1.5067... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20471... Training loss: 1.5708... 0.0605 sec/batch
Epoch: 6/20... Training Step: 20472... Training loss: 1.6855... 0.0524 sec/batch
Epoch: 6/20... Training Step: 20473... Training loss: 1.6668... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20474... Training loss: 1.6950... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20475... Training loss: 1.5437... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20476... Training loss: 1.5879... 0.0546 sec/batch
Epoch: 6/20... Training Step: 20477... Training loss: 1.6108... 0.0486 sec/batch
Epoch: 6/20... Training Step: 20478... Training loss: 1.5069... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20479... Training loss: 1.6097... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20480... Training loss: 1.7714... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20481... Training loss: 1.6806... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20482... Training loss: 1.5587... 0.0506 sec/batch
Epoch: 6/20... Training Step: 20483... Training loss: 1.6116... 0.0452 sec/batch
Epoch: 6/20... Training Step: 20484... Training loss: 1.6584... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20485... Training loss: 1.7371... 0.0487 sec/batch
Epoch: 6/20... Training Step: 20486... Training loss: 1.7350... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20487... Training loss: 1.6766... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20488... Training loss: 1.6237... 0.0590 sec/batch
Epoch: 6/20... Training Step: 20489... Training loss: 1.5868... 0.0454 sec/batch
Epoch: 6/20... Training Step: 20490... Training loss: 1.4945... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20491... Training loss: 1.6377... 0.0570 sec/batch
Epoch: 6/20... Training Step: 20492... Training loss: 1.5607... 0.0439 sec/batch
Epoch: 6/20... Training Step: 20493... Training loss: 1.5600... 0.0586 sec/batch
Epoch: 6/20... Training Step: 20494... Training loss: 1.4710... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20495... Training loss: 1.6235... 0.0511 sec/batch
Epoch: 6/20... Training Step: 20496... Training loss: 1.7800... 0.0427 sec/batch
Epoch: 6/20... Training Step: 20497... Training loss: 1.5624... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20498... Training loss: 1.6671... 0.0473 sec/batch
Epoch: 6/20... Training Step: 20499... Training loss: 1.5944... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20500... Training loss: 1.4889... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20501... Training loss: 1.6561... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20502... Training loss: 1.8112... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20503... Training loss: 1.7837... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20504... Training loss: 1.5864... 0.0486 sec/batch
Epoch: 6/20... Training Step: 20505... Training loss: 1.6454... 0.0491 sec/batch
Epoch: 6/20... Training Step: 20506... Training loss: 1.6334... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20507... Training loss: 1.6275... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20508... Training loss: 1.7051... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20509... Training loss: 1.7562... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20510... Training loss: 1.6655... 0.0477 sec/batch
Epoch: 6/20... Training Step: 20511... Training loss: 1.6996... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20512... Training loss: 1.5691... 0.0476 sec/batch
Epoch: 6/20... Training Step: 20513... Training loss: 1.7072... 0.0478 sec/batch
Epoch: 6/20... Training Step: 20514... Training loss: 1.7268... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20515... Training loss: 1.7511... 0.0550 sec/batch
Epoch: 6/20... Training Step: 20516... Training loss: 1.6876... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20517... Training loss: 1.6839... 0.0517 sec/batch
Epoch: 6/20... Training Step: 20518... Training loss: 1.5384... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20519... Training loss: 1.5740... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20520... Training loss: 1.6453... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20521... Training loss: 1.6350... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20522... Training loss: 1.6213... 0.0531 sec/batch
Epoch: 6/20... Training Step: 20523... Training loss: 1.7103... 0.0466 sec/batch
Epoch: 6/20... Training Step: 20524... Training loss: 1.7026... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20525... Training loss: 1.6322... 0.0409 sec/batch
Epoch: 6/20... Training Step: 20526... Training loss: 1.6144... 0.0508 sec/batch
Epoch: 6/20... Training Step: 20527... Training loss: 1.7674... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20528... Training loss: 1.6673... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20529... Training loss: 1.6729... 0.0485 sec/batch
Epoch: 6/20... Training Step: 20530... Training loss: 1.6902... 0.0509 sec/batch
Epoch: 6/20... Training Step: 20531... Training loss: 1.6646... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20532... Training loss: 1.6137... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20533... Training loss: 1.6432... 0.0565 sec/batch
Epoch: 6/20... Training Step: 20534... Training loss: 1.6068... 0.0456 sec/batch
Epoch: 6/20... Training Step: 20535... Training loss: 1.5887... 0.0529 sec/batch
Epoch: 6/20... Training Step: 20536... Training loss: 1.6625... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20537... Training loss: 1.6971... 0.0494 sec/batch
Epoch: 6/20... Training Step: 20538... Training loss: 1.5879... 0.0420 sec/batch
Epoch: 6/20... Training Step: 20539... Training loss: 1.5597... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20540... Training loss: 1.5928... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20541... Training loss: 1.7313... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20542... Training loss: 1.7561... 0.0495 sec/batch
Epoch: 6/20... Training Step: 20543... Training loss: 1.7026... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20544... Training loss: 1.6798... 0.0473 sec/batch
Epoch: 6/20... Training Step: 20545... Training loss: 1.6016... 0.0417 sec/batch
Epoch: 6/20... Training Step: 20546... Training loss: 1.7412... 0.0524 sec/batch
Epoch: 6/20... Training Step: 20547... Training loss: 1.6942... 0.0466 sec/batch
Epoch: 6/20... Training Step: 20548... Training loss: 1.6674... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20549... Training loss: 1.6855... 0.0475 sec/batch
Epoch: 6/20... Training Step: 20550... Training loss: 1.6310... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20551... Training loss: 1.7346... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20552... Training loss: 1.6012... 0.0483 sec/batch
Epoch: 6/20... Training Step: 20553... Training loss: 1.6338... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20554... Training loss: 1.6704... 0.0409 sec/batch
Epoch: 6/20... Training Step: 20555... Training loss: 1.6489... 0.0403 sec/batch
Epoch: 6/20... Training Step: 20556... Training loss: 1.6977... 0.0519 sec/batch
Epoch: 6/20... Training Step: 20557... Training loss: 1.7022... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20558... Training loss: 1.6099... 0.0536 sec/batch
Epoch: 6/20... Training Step: 20559... Training loss: 1.6276... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20560... Training loss: 1.6474... 0.0481 sec/batch
Epoch: 6/20... Training Step: 20561... Training loss: 1.6226... 0.0546 sec/batch
Epoch: 6/20... Training Step: 20562... Training loss: 1.6365... 0.0549 sec/batch
Epoch: 6/20... Training Step: 20563... Training loss: 1.7374... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20564... Training loss: 1.5848... 0.0483 sec/batch
Epoch: 6/20... Training Step: 20565... Training loss: 1.5374... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20566... Training loss: 1.5098... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20567... Training loss: 1.6604... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20568... Training loss: 1.5367... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20569... Training loss: 1.5943... 0.0550 sec/batch
Epoch: 6/20... Training Step: 20570... Training loss: 1.6973... 0.0500 sec/batch
Epoch: 6/20... Training Step: 20571... Training loss: 1.6703... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20572... Training loss: 1.5242... 0.0568 sec/batch
Epoch: 6/20... Training Step: 20573... Training loss: 1.5830... 0.0500 sec/batch
Epoch: 6/20... Training Step: 20574... Training loss: 1.6841... 0.0577 sec/batch
Epoch: 6/20... Training Step: 20575... Training loss: 1.6404... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20576... Training loss: 1.5874... 0.0539 sec/batch
Epoch: 6/20... Training Step: 20577... Training loss: 1.6388... 0.0506 sec/batch
Epoch: 6/20... Training Step: 20578... Training loss: 1.5730... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20579... Training loss: 1.6279... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20580... Training loss: 1.5733... 0.0494 sec/batch
Epoch: 6/20... Training Step: 20581... Training loss: 1.6443... 0.0498 sec/batch
Epoch: 6/20... Training Step: 20582... Training loss: 1.4904... 0.0595 sec/batch
Epoch: 6/20... Training Step: 20583... Training loss: 1.6459... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20584... Training loss: 1.5381... 0.0640 sec/batch
Epoch: 6/20... Training Step: 20585... Training loss: 1.4990... 0.0540 sec/batch
Epoch: 6/20... Training Step: 20586... Training loss: 1.6793... 0.0532 sec/batch
Epoch: 6/20... Training Step: 20587... Training loss: 1.6466... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20588... Training loss: 1.6557... 0.0578 sec/batch
Epoch: 6/20... Training Step: 20589... Training loss: 1.6546... 0.0635 sec/batch
Epoch: 6/20... Training Step: 20590... Training loss: 1.6156... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20591... Training loss: 1.6054... 0.0513 sec/batch
Epoch: 6/20... Training Step: 20592... Training loss: 1.4927... 0.0481 sec/batch
Epoch: 6/20... Training Step: 20593... Training loss: 1.5543... 0.0591 sec/batch
Epoch: 6/20... Training Step: 20594... Training loss: 1.6067... 0.0477 sec/batch
Epoch: 6/20... Training Step: 20595... Training loss: 1.5977... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20596... Training loss: 1.7008... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20597... Training loss: 1.6579... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20598... Training loss: 1.6317... 0.0509 sec/batch
Epoch: 6/20... Training Step: 20599... Training loss: 1.5829... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20600... Training loss: 1.6467... 0.0519 sec/batch
Epoch: 6/20... Training Step: 20601... Training loss: 1.6001... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20602... Training loss: 1.8847... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20603... Training loss: 1.6796... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20604... Training loss: 1.5394... 0.0445 sec/batch
Epoch: 6/20... Training Step: 20605... Training loss: 1.5309... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20606... Training loss: 1.5275... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20607... Training loss: 1.6126... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20608... Training loss: 1.6066... 0.0477 sec/batch
Epoch: 6/20... Training Step: 20609... Training loss: 1.6582... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20610... Training loss: 1.5329... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20611... Training loss: 1.4825... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20612... Training loss: 1.8060... 0.0509 sec/batch
Epoch: 6/20... Training Step: 20613... Training loss: 1.6917... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20614... Training loss: 1.6174... 0.0506 sec/batch
Epoch: 6/20... Training Step: 20615... Training loss: 1.7036... 0.0528 sec/batch
Epoch: 6/20... Training Step: 20616... Training loss: 1.6484... 0.0501 sec/batch
Epoch: 6/20... Training Step: 20617... Training loss: 1.6969... 0.0545 sec/batch
Epoch: 6/20... Training Step: 20618... Training loss: 1.6475... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20619... Training loss: 1.7429... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20620... Training loss: 1.7039... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20621... Training loss: 1.6136... 0.0568 sec/batch
Epoch: 6/20... Training Step: 20622... Training loss: 1.6119... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20623... Training loss: 1.6475... 0.0585 sec/batch
Epoch: 6/20... Training Step: 20624... Training loss: 1.7574... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20625... Training loss: 1.6706... 0.0530 sec/batch
Epoch: 6/20... Training Step: 20626... Training loss: 1.7033... 0.0554 sec/batch
Epoch: 6/20... Training Step: 20627... Training loss: 1.8048... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20628... Training loss: 1.7817... 0.0593 sec/batch
Epoch: 6/20... Training Step: 20629... Training loss: 1.6991... 0.0543 sec/batch
Epoch: 6/20... Training Step: 20630... Training loss: 1.7361... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20631... Training loss: 1.5863... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20632... Training loss: 1.6277... 0.0453 sec/batch
Epoch: 6/20... Training Step: 20633... Training loss: 1.6785... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20634... Training loss: 1.5462... 0.0573 sec/batch
Epoch: 6/20... Training Step: 20635... Training loss: 1.6975... 0.0428 sec/batch
Epoch: 6/20... Training Step: 20636... Training loss: 1.7133... 0.0440 sec/batch
Epoch: 6/20... Training Step: 20637... Training loss: 1.5764... 0.0547 sec/batch
Epoch: 6/20... Training Step: 20638... Training loss: 1.4692... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20639... Training loss: 1.6693... 0.0431 sec/batch
Epoch: 6/20... Training Step: 20640... Training loss: 1.6702... 0.0410 sec/batch
Epoch: 6/20... Training Step: 20641... Training loss: 1.6218... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20642... Training loss: 1.5232... 0.0491 sec/batch
Epoch: 6/20... Training Step: 20643... Training loss: 1.6541... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20644... Training loss: 1.6070... 0.0479 sec/batch
Epoch: 6/20... Training Step: 20645... Training loss: 1.5577... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20646... Training loss: 1.5074... 0.0432 sec/batch
Epoch: 6/20... Training Step: 20647... Training loss: 1.6641... 0.0485 sec/batch
Epoch: 6/20... Training Step: 20648... Training loss: 1.5330... 0.0461 sec/batch
Epoch: 6/20... Training Step: 20649... Training loss: 1.7004... 0.0487 sec/batch
Epoch: 6/20... Training Step: 20650... Training loss: 1.5970... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20651... Training loss: 1.6560... 0.0473 sec/batch
Epoch: 6/20... Training Step: 20652... Training loss: 1.5126... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20653... Training loss: 1.6976... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20654... Training loss: 1.8007... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20655... Training loss: 1.7673... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20656... Training loss: 1.7133... 0.0438 sec/batch
Epoch: 6/20... Training Step: 20657... Training loss: 1.6374... 0.0464 sec/batch
Epoch: 6/20... Training Step: 20658... Training loss: 1.5826... 0.0416 sec/batch
Epoch: 6/20... Training Step: 20659... Training loss: 1.5603... 0.0446 sec/batch
Epoch: 6/20... Training Step: 20660... Training loss: 1.5528... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20661... Training loss: 1.5734... 0.0414 sec/batch
Epoch: 6/20... Training Step: 20662... Training loss: 1.6036... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20663... Training loss: 1.6427... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20664... Training loss: 1.5486... 0.0485 sec/batch
Epoch: 6/20... Training Step: 20665... Training loss: 1.6531... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20666... Training loss: 1.5892... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20667... Training loss: 1.5014... 0.0546 sec/batch
Epoch: 6/20... Training Step: 20668... Training loss: 1.6241... 0.0473 sec/batch
Epoch: 6/20... Training Step: 20669... Training loss: 1.5901... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20670... Training loss: 1.6978... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20671... Training loss: 1.6802... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20672... Training loss: 1.6552... 0.0548 sec/batch
Epoch: 6/20... Training Step: 20673... Training loss: 1.5921... 0.0514 sec/batch
Epoch: 6/20... Training Step: 20674... Training loss: 1.5609... 0.0411 sec/batch
Epoch: 6/20... Training Step: 20675... Training loss: 1.5285... 0.0484 sec/batch
Epoch: 6/20... Training Step: 20676... Training loss: 1.5989... 0.0466 sec/batch
Epoch: 6/20... Training Step: 20677... Training loss: 1.6516... 0.0501 sec/batch
Epoch: 6/20... Training Step: 20678... Training loss: 1.6346... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20679... Training loss: 1.5938... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20680... Training loss: 1.5975... 0.0455 sec/batch
Epoch: 6/20... Training Step: 20681... Training loss: 1.6928... 0.0497 sec/batch
Epoch: 6/20... Training Step: 20682... Training loss: 1.6121... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20683... Training loss: 1.5755... 0.0475 sec/batch
Epoch: 6/20... Training Step: 20684... Training loss: 1.5902... 0.0446 sec/batch
Epoch: 6/20... Training Step: 20685... Training loss: 1.7152... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20686... Training loss: 1.4460... 0.0574 sec/batch
Epoch: 6/20... Training Step: 20687... Training loss: 1.6205... 0.0490 sec/batch
Epoch: 6/20... Training Step: 20688... Training loss: 1.5059... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20689... Training loss: 1.6757... 0.0508 sec/batch
Epoch: 6/20... Training Step: 20690... Training loss: 1.6106... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20691... Training loss: 1.6425... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20692... Training loss: 1.6148... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20693... Training loss: 1.5905... 0.0560 sec/batch
Epoch: 6/20... Training Step: 20694... Training loss: 1.6328... 0.0434 sec/batch
Epoch: 6/20... Training Step: 20695... Training loss: 1.6119... 0.0560 sec/batch
Epoch: 6/20... Training Step: 20696... Training loss: 1.6415... 0.0500 sec/batch
Epoch: 6/20... Training Step: 20697... Training loss: 1.7296... 0.0429 sec/batch
Epoch: 6/20... Training Step: 20698... Training loss: 1.6105... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20699... Training loss: 1.6383... 0.0425 sec/batch
Epoch: 6/20... Training Step: 20700... Training loss: 1.7886... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20701... Training loss: 1.7333... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20702... Training loss: 1.4349... 0.0533 sec/batch
Epoch: 6/20... Training Step: 20703... Training loss: 1.6082... 0.0449 sec/batch
Epoch: 6/20... Training Step: 20704... Training loss: 1.5437... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20705... Training loss: 1.7296... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20706... Training loss: 1.6597... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20707... Training loss: 1.7536... 0.0538 sec/batch
Epoch: 6/20... Training Step: 20708... Training loss: 1.6151... 0.0447 sec/batch
Epoch: 6/20... Training Step: 20709... Training loss: 1.6311... 0.0617 sec/batch
Epoch: 6/20... Training Step: 20710... Training loss: 1.7971... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20711... Training loss: 1.6469... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20712... Training loss: 1.6787... 0.0411 sec/batch
Epoch: 6/20... Training Step: 20713... Training loss: 1.7922... 0.0476 sec/batch
Epoch: 6/20... Training Step: 20714... Training loss: 1.6484... 0.0515 sec/batch
Epoch: 6/20... Training Step: 20715... Training loss: 1.6986... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20716... Training loss: 1.9003... 0.0448 sec/batch
Epoch: 6/20... Training Step: 20717... Training loss: 1.7152... 0.0430 sec/batch
Epoch: 6/20... Training Step: 20718... Training loss: 1.7153... 0.0411 sec/batch
Epoch: 6/20... Training Step: 20719... Training loss: 1.7538... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20720... Training loss: 1.7143... 0.0463 sec/batch
Epoch: 6/20... Training Step: 20721... Training loss: 1.6105... 0.0435 sec/batch
Epoch: 6/20... Training Step: 20722... Training loss: 1.5641... 0.0501 sec/batch
Epoch: 6/20... Training Step: 20723... Training loss: 1.6842... 0.0507 sec/batch
Epoch: 6/20... Training Step: 20724... Training loss: 1.6666... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20725... Training loss: 1.7496... 0.0608 sec/batch
Epoch: 6/20... Training Step: 20726... Training loss: 1.6505... 0.0552 sec/batch
Epoch: 6/20... Training Step: 20727... Training loss: 1.6356... 0.0478 sec/batch
Epoch: 6/20... Training Step: 20728... Training loss: 1.7736... 0.0564 sec/batch
Epoch: 6/20... Training Step: 20729... Training loss: 1.5824... 0.0443 sec/batch
Epoch: 6/20... Training Step: 20730... Training loss: 1.5500... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20731... Training loss: 1.8820... 0.0458 sec/batch
Epoch: 6/20... Training Step: 20732... Training loss: 1.8533... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20733... Training loss: 1.7124... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20734... Training loss: 1.7171... 0.0503 sec/batch
Epoch: 6/20... Training Step: 20735... Training loss: 1.5497... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20736... Training loss: 1.6167... 0.0505 sec/batch
Epoch: 6/20... Training Step: 20737... Training loss: 1.6527... 0.0502 sec/batch
Epoch: 6/20... Training Step: 20738... Training loss: 1.5956... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20739... Training loss: 1.6449... 0.0649 sec/batch
Epoch: 6/20... Training Step: 20740... Training loss: 1.6208... 0.0516 sec/batch
Epoch: 6/20... Training Step: 20741... Training loss: 1.6005... 0.0455 sec/batch
Epoch: 6/20... Training Step: 20742... Training loss: 1.6912... 0.0564 sec/batch
Epoch: 6/20... Training Step: 20743... Training loss: 1.7614... 0.0508 sec/batch
Epoch: 6/20... Training Step: 20744... Training loss: 1.7027... 0.0586 sec/batch
Epoch: 6/20... Training Step: 20745... Training loss: 1.7240... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20746... Training loss: 1.6623... 0.0521 sec/batch
Epoch: 6/20... Training Step: 20747... Training loss: 1.8210... 0.0441 sec/batch
Epoch: 6/20... Training Step: 20748... Training loss: 1.5314... 0.0504 sec/batch
Epoch: 6/20... Training Step: 20749... Training loss: 1.6701... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20750... Training loss: 1.6437... 0.0525 sec/batch
Epoch: 6/20... Training Step: 20751... Training loss: 1.4898... 0.0457 sec/batch
Epoch: 6/20... Training Step: 20752... Training loss: 1.5090... 0.0428 sec/batch
Epoch: 6/20... Training Step: 20753... Training loss: 1.5845... 0.0491 sec/batch
Epoch: 6/20... Training Step: 20754... Training loss: 1.5824... 0.0524 sec/batch
Epoch: 6/20... Training Step: 20755... Training loss: 1.6332... 0.0445 sec/batch
Epoch: 6/20... Training Step: 20756... Training loss: 1.7381... 0.0468 sec/batch
Epoch: 6/20... Training Step: 20757... Training loss: 1.7963... 0.0505 sec/batch
Epoch: 6/20... Training Step: 20758... Training loss: 1.7274... 0.0501 sec/batch
Epoch: 6/20... Training Step: 20759... Training loss: 1.7623... 0.0496 sec/batch
Epoch: 6/20... Training Step: 20760... Training loss: 1.7306... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20761... Training loss: 1.6532... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20762... Training loss: 1.6821... 0.0480 sec/batch
Epoch: 6/20... Training Step: 20763... Training loss: 1.7425... 0.0441 sec/batch
Epoch: 6/20... Training Step: 20764... Training loss: 1.6400... 0.0472 sec/batch
Epoch: 6/20... Training Step: 20765... Training loss: 1.6372... 0.0445 sec/batch
Epoch: 6/20... Training Step: 20766... Training loss: 1.6503... 0.0532 sec/batch
Epoch: 6/20... Training Step: 20767... Training loss: 1.7000... 0.0514 sec/batch
Epoch: 6/20... Training Step: 20768... Training loss: 1.5293... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20769... Training loss: 1.6269... 0.0488 sec/batch
Epoch: 6/20... Training Step: 20770... Training loss: 1.5903... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20771... Training loss: 1.5473... 0.0509 sec/batch
Epoch: 6/20... Training Step: 20772... Training loss: 1.7230... 0.0578 sec/batch
Epoch: 6/20... Training Step: 20773... Training loss: 1.7395... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20774... Training loss: 1.6111... 0.0510 sec/batch
Epoch: 6/20... Training Step: 20775... Training loss: 1.7090... 0.0499 sec/batch
Epoch: 6/20... Training Step: 20776... Training loss: 1.6760... 0.0462 sec/batch
Epoch: 6/20... Training Step: 20777... Training loss: 1.6533... 0.0535 sec/batch
Epoch: 6/20... Training Step: 20778... Training loss: 1.6066... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20779... Training loss: 1.5681... 0.0467 sec/batch
Epoch: 6/20... Training Step: 20780... Training loss: 1.6786... 0.0524 sec/batch
Epoch: 6/20... Training Step: 20781... Training loss: 1.6164... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20782... Training loss: 1.5360... 0.0532 sec/batch
Epoch: 6/20... Training Step: 20783... Training loss: 1.5961... 0.0543 sec/batch
Epoch: 6/20... Training Step: 20784... Training loss: 1.6105... 0.0600 sec/batch
Epoch: 6/20... Training Step: 20785... Training loss: 1.6162... 0.0482 sec/batch
Epoch: 6/20... Training Step: 20786... Training loss: 1.6167... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20787... Training loss: 1.5055... 0.0514 sec/batch
Epoch: 6/20... Training Step: 20788... Training loss: 1.5982... 0.0450 sec/batch
Epoch: 6/20... Training Step: 20789... Training loss: 1.5664... 0.0441 sec/batch
Epoch: 6/20... Training Step: 20790... Training loss: 1.5908... 0.0541 sec/batch
Epoch: 6/20... Training Step: 20791... Training loss: 1.4673... 0.0513 sec/batch
Epoch: 6/20... Training Step: 20792... Training loss: 1.5027... 0.0513 sec/batch
Epoch: 6/20... Training Step: 20793... Training loss: 1.5853... 0.0589 sec/batch
Epoch: 6/20... Training Step: 20794... Training loss: 1.7605... 0.0522 sec/batch
Epoch: 6/20... Training Step: 20795... Training loss: 1.6318... 0.0483 sec/batch
Epoch: 6/20... Training Step: 20796... Training loss: 1.5964... 0.0492 sec/batch
Epoch: 6/20... Training Step: 20797... Training loss: 1.5189... 0.0433 sec/batch
Epoch: 6/20... Training Step: 20798... Training loss: 1.6127... 0.0512 sec/batch
Epoch: 6/20... Training Step: 20799... Training loss: 1.6334... 0.0519 sec/batch
Epoch: 6/20... Training Step: 20800... Training loss: 1.6886... 0.0471 sec/batch
Epoch: 6/20... Training Step: 20801... Training loss: 1.6464... 0.0636 sec/batch
Epoch: 6/20... Training Step: 20802... Training loss: 1.6166... 0.0474 sec/batch
Epoch: 6/20... Training Step: 20803... Training loss: 1.7437... 0.0528 sec/batch
Epoch: 6/20... Training Step: 20804... Training loss: 1.6006... 0.0526 sec/batch
Epoch: 6/20... Training Step: 20805... Training loss: 1.6288... 0.0481 sec/batch
Epoch: 6/20... Training Step: 20806... Training loss: 1.5674... 0.0442 sec/batch
Epoch: 6/20... Training Step: 20807... Training loss: 1.7227... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20808... Training loss: 1.6192... 0.0424 sec/batch
Epoch: 6/20... Training Step: 20809... Training loss: 1.5196... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20810... Training loss: 1.7235... 0.0465 sec/batch
Epoch: 6/20... Training Step: 20811... Training loss: 1.5749... 0.0460 sec/batch
Epoch: 6/20... Training Step: 20812... Training loss: 1.6527... 0.0485 sec/batch
Epoch: 6/20... Training Step: 20813... Training loss: 1.5264... 0.0418 sec/batch
Epoch: 6/20... Training Step: 20814... Training loss: 1.5558... 0.0426 sec/batch
Epoch: 6/20... Training Step: 20815... Training loss: 1.7222... 0.0439 sec/batch
Epoch: 6/20... Training Step: 20816... Training loss: 1.6893... 0.0529 sec/batch
Epoch: 6/20... Training Step: 20817... Training loss: 1.5602... 0.0436 sec/batch
Epoch: 6/20... Training Step: 20818... Training loss: 1.6438... 0.0419 sec/batch
Epoch: 6/20... Training Step: 20819... Training loss: 1.6880... 0.0491 sec/batch
Epoch: 6/20... Training Step: 20820... Training loss: 1.5288... 0.0574 sec/batch
Epoch: 6/20... Training Step: 20821... Training loss: 1.5485... 0.0470 sec/batch
Epoch: 6/20... Training Step: 20822... Training loss: 1.6752... 0.0469 sec/batch
Epoch: 6/20... Training Step: 20823... Training loss: 1.5796... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20824... Training loss: 1.5897... 0.0405 sec/batch
Epoch: 6/20... Training Step: 20825... Training loss: 1.5406... 0.0459 sec/batch
Epoch: 6/20... Training Step: 20826... Training loss: 1.7171... 0.0500 sec/batch
---------------------------------------------------------------------------
KeyboardInterrupt Traceback (most recent call last)
<ipython-input-50-ad76a1595262> in <module>()
28 model.final_state,
29 model.optimizer],
---> 30 feed_dict=feed)
31
32 end = time.time()
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
765 try:
766 result = self._run(None, fetches, feed_dict, options_ptr,
--> 767 run_metadata_ptr)
768 if run_metadata:
769 proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
963 if final_fetches or final_targets:
964 results = self._do_run(handle, final_targets, final_fetches,
--> 965 feed_dict_string, options, run_metadata)
966 else:
967 results = []
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_run(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)
1013 if handle is None:
1014 return self._do_call(_run_fn, self._session, feed_dict, fetch_list,
-> 1015 target_list, options, run_metadata)
1016 else:
1017 return self._do_call(_prun_fn, self._session, handle, feed_dict,
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1020 def _do_call(self, fn, *args):
1021 try:
-> 1022 return fn(*args)
1023 except errors.OpError as e:
1024 message = compat.as_text(e.message)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run_fn(session, feed_dict, fetch_list, target_list, options, run_metadata)
1002 return tf_session.TF_Run(session, options,
1003 feed_dict, fetch_list, target_list,
-> 1004 status, run_metadata)
1005
1006 def _prun_fn(session, handle, feed_dict, fetch_list):
KeyboardInterrupt:
Read up on saving and loading checkpoints here: https://www.tensorflow.org/programmers_guide/variables
In [51]:
tf.train.get_checkpoint_state('checkpoints')
Out[51]:
model_checkpoint_path: "checkpoints/i20800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i1000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i1200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i1400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i1600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i1800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i2000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i2200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i2400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i2600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i2800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i3000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i3200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i3400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i3600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i3800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i4000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i4200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i4400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i4600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i4800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i5000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i5200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i5400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i5600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i5800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i6000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i6200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i6400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i6600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i6800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i7000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i7200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i7400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i7600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i7800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i8000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i8200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i8400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i8600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i8800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i9000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i9200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i9400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i9600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i9800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i10000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i10200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i10400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i10600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i10800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i11000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i11200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i11400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i11600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i11800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i12000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i12200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i12400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i12600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i12800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i13000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i13200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i13400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i13600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i13800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i14000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i14200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i14400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i14600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i14800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i15000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i15200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i15400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i15600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i15800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i16000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i16200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i16400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i16600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i16800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i17000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i17200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i17400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i17600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i17800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i18000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i18200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i18400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i18600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i18800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i19000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i19200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i19400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i19600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i19800_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i20000_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i20200_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i20400_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i20600_l128.ckpt"
all_model_checkpoint_paths: "checkpoints/i20800_l128.ckpt"
Now that the network is trained, we'll can use it to generate new text. The idea is that we pass in a character, then the network will predict the next character. We can use the new one, to predict the next one. And we keep doing this to generate all new text. I also included some functionality to prime the network with some text by passing in a string and building up a state from that.
The network gives us predictions for each character. To reduce noise and make things a little less random, I'm going to only choose a new character from the top N most likely characters.
In [52]:
def pick_top_n(preds, vocab_size, top_n=5):
p = np.squeeze(preds)
p[np.argsort(p)[:-top_n]] = 0
p = p / np.sum(p)
c = np.random.choice(vocab_size, 1, p=p)[0]
return c
In [53]:
def sample(checkpoint, n_samples, lstm_size, vocab_size, prime="The "):
samples = [c for c in prime]
model = CharRNN(len(vocab), lstm_size=lstm_size, sampling=True)
saver = tf.train.Saver()
with tf.Session() as sess:
saver.restore(sess, checkpoint)
new_state = sess.run(model.initial_state)
for c in prime:
x = np.zeros((1, 1))
x[0,0] = vocab_to_int[c]
feed = {model.inputs: x,
model.keep_prob: 1.,
model.initial_state: new_state}
preds, new_state = sess.run([model.prediction, model.final_state],
feed_dict=feed)
c = pick_top_n(preds, len(vocab))
samples.append(int_to_vocab[c])
for i in range(n_samples):
x[0,0] = c
feed = {model.inputs: x,
model.keep_prob: 1.,
model.initial_state: new_state}
preds, new_state = sess.run([model.prediction, model.final_state],
feed_dict=feed)
c = pick_top_n(preds, len(vocab))
samples.append(int_to_vocab[c])
return ''.join(samples)
Here, pass in the path to a checkpoint and sample from the network.
In [54]:
tf.train.latest_checkpoint('checkpoints')
Out[54]:
'checkpoints/i20800_l128.ckpt'
In [55]:
checkpoint = tf.train.latest_checkpoint('checkpoints')
samp = sample(checkpoint, 2000, lstm_size, len(vocab), prime="Far")
print(samp)
Fare all the mashe
through the differents.
"To day, she meating, and the whoer or hand.
I did not
drawer that a crush."
"What were the mare to see the more and a dropped
if your, things to the shirt, and the mind at his shoting, and some marraes instant where tell."
"I did the some tomertiem,
time was, that we despair there serene into the some to be saying, I'm so stair."
"You're shrow the conspitions with morithing and setter."
"Yes!"
"I did the misses, and she's a money of sowant and and stord and a mugh on their servants."
"It's not the money what she
was she had not become it was
anyone that
all hear of and servales was a condoncy, they had not a few minute where a doctor was something. I will have to say," he healthed him, the come on the wait, who had an impristed him. He was that he felt saw was to sone and with she went to her
sorry at the corsance had the sounds.
"I've been to shall be shanges, and a children and thing," he
said, and went to his felt
the meeting, that he was
see over a feast, a there were a ming had to take hers how to be
to her hot and went up in the call, but he
was shoot, and she had not she was a moment.
"No, I have so him. I did you could not
think," said Anna was
a self-torders triated on the cheel it. Anywhen
a much at
the compres a something with the consider to
her hand, and still had
a same so he consisted and he do at her. The countess, to say of something with a sontation of that staid and had they seemed at the men and wanters at the stort, and the catter.
"It's time."
"No, where I'll talk the
tanges.
You was she went up."
"I did not children work the termice, that I have stocked his himserious wet them of them. I can
must be any hard a day, something on
alove or all went to, this
wonders to decertated on has won't ask a sound."
"Well, as a facing the money, I did
that.
I had stept, what they say all and her offers." He wanted the wart of stees without their sat oving her to be.
"You want.
You could steement,
I've se
In [56]:
checkpoint = 'checkpoints/i200_l512.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
print(samp)
---------------------------------------------------------------------------
NotFoundError Traceback (most recent call last)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1021 try:
-> 1022 return fn(*args)
1023 except errors.OpError as e:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run_fn(session, feed_dict, fetch_list, target_list, options, run_metadata)
1003 feed_dict, fetch_list, target_list,
-> 1004 status, run_metadata)
1005
/home/luo/anaconda2/envs/dlnd/lib/python3.6/contextlib.py in __exit__(self, type, value, traceback)
88 try:
---> 89 next(self.gen)
90 except StopIteration:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/errors_impl.py in raise_exception_on_not_ok_status()
465 compat.as_text(pywrap_tensorflow.TF_Message(status)),
--> 466 pywrap_tensorflow.TF_GetCode(status))
467 finally:
NotFoundError: Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i200_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_7/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_173_save/RestoreV2_7", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
During handling of the above exception, another exception occurred:
NotFoundError Traceback (most recent call last)
<ipython-input-56-d1a22da3ba8d> in <module>()
1 checkpoint = 'checkpoints/i200_l512.ckpt'
----> 2 samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
3 print(samp)
<ipython-input-53-8eb787ae9642> in sample(checkpoint, n_samples, lstm_size, vocab_size, prime)
4 saver = tf.train.Saver()
5 with tf.Session() as sess:
----> 6 saver.restore(sess, checkpoint)
7 new_state = sess.run(model.initial_state)
8 for c in prime:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py in restore(self, sess, save_path)
1426 return
1427 sess.run(self.saver_def.restore_op_name,
-> 1428 {self.saver_def.filename_tensor_name: save_path})
1429
1430 @staticmethod
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
765 try:
766 result = self._run(None, fetches, feed_dict, options_ptr,
--> 767 run_metadata_ptr)
768 if run_metadata:
769 proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
963 if final_fetches or final_targets:
964 results = self._do_run(handle, final_targets, final_fetches,
--> 965 feed_dict_string, options, run_metadata)
966 else:
967 results = []
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_run(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)
1013 if handle is None:
1014 return self._do_call(_run_fn, self._session, feed_dict, fetch_list,
-> 1015 target_list, options, run_metadata)
1016 else:
1017 return self._do_call(_prun_fn, self._session, handle, feed_dict,
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1033 except KeyError:
1034 pass
-> 1035 raise type(e)(node_def, op, message)
1036
1037 def _extend_graph(self):
NotFoundError: Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i200_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_7/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_173_save/RestoreV2_7", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
Caused by op 'save/RestoreV2', defined at:
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"__main__", mod_spec)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/__main__.py", line 3, in <module>
app.launch_new_instance()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/traitlets/config/application.py", line 658, in launch_instance
app.start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelapp.py", line 474, in start
ioloop.IOLoop.instance().start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/ioloop.py", line 177, in start
super(ZMQIOLoop, self).start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/ioloop.py", line 887, in start
handler_func(fd_obj, events)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/stack_context.py", line 275, in null_wrapper
return fn(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 440, in _handle_events
self._handle_recv()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 472, in _handle_recv
self._run_callback(callback, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 414, in _run_callback
callback(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/stack_context.py", line 275, in null_wrapper
return fn(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 276, in dispatcher
return self.dispatch_shell(stream, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 228, in dispatch_shell
handler(stream, idents, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 390, in execute_request
user_expressions, allow_stdin)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/ipkernel.py", line 196, in do_execute
res = shell.run_cell(code, store_history=store_history, silent=silent)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/zmqshell.py", line 501, in run_cell
return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2717, in run_cell
interactivity=interactivity, compiler=compiler, result=result)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2821, in run_ast_nodes
if self.run_code(code, result):
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2881, in run_code
exec(code_obj, self.user_global_ns, self.user_ns)
File "<ipython-input-56-d1a22da3ba8d>", line 2, in <module>
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
File "<ipython-input-53-8eb787ae9642>", line 4, in sample
saver = tf.train.Saver()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1040, in __init__
self.build()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1070, in build
restore_sequentially=self._restore_sequentially)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 675, in build
restore_sequentially, reshape)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 402, in _AddRestoreOps
tensors = self.restore_op(filename_tensor, saveable, preferred_shard)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 242, in restore_op
[spec.tensor.dtype])[0])
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/ops/gen_io_ops.py", line 668, in restore_v2
dtypes=dtypes, name=name)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 763, in apply_op
op_def=op_def)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 2327, in create_op
original_op=self._default_original_op, op_def=op_def)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1226, in __init__
self._traceback = _extract_stack()
NotFoundError (see above for traceback): Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i200_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_7/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_173_save/RestoreV2_7", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
In [57]:
checkpoint = 'checkpoints/i600_l512.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
print(samp)
---------------------------------------------------------------------------
NotFoundError Traceback (most recent call last)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1021 try:
-> 1022 return fn(*args)
1023 except errors.OpError as e:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run_fn(session, feed_dict, fetch_list, target_list, options, run_metadata)
1003 feed_dict, fetch_list, target_list,
-> 1004 status, run_metadata)
1005
/home/luo/anaconda2/envs/dlnd/lib/python3.6/contextlib.py in __exit__(self, type, value, traceback)
88 try:
---> 89 next(self.gen)
90 except StopIteration:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/errors_impl.py in raise_exception_on_not_ok_status()
465 compat.as_text(pywrap_tensorflow.TF_Message(status)),
--> 466 pywrap_tensorflow.TF_GetCode(status))
467 finally:
NotFoundError: Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i600_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_14/_29 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_158_save/RestoreV2_14", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
During handling of the above exception, another exception occurred:
NotFoundError Traceback (most recent call last)
<ipython-input-57-7c4e18bbddc1> in <module>()
1 checkpoint = 'checkpoints/i600_l512.ckpt'
----> 2 samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
3 print(samp)
<ipython-input-53-8eb787ae9642> in sample(checkpoint, n_samples, lstm_size, vocab_size, prime)
4 saver = tf.train.Saver()
5 with tf.Session() as sess:
----> 6 saver.restore(sess, checkpoint)
7 new_state = sess.run(model.initial_state)
8 for c in prime:
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py in restore(self, sess, save_path)
1426 return
1427 sess.run(self.saver_def.restore_op_name,
-> 1428 {self.saver_def.filename_tensor_name: save_path})
1429
1430 @staticmethod
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
765 try:
766 result = self._run(None, fetches, feed_dict, options_ptr,
--> 767 run_metadata_ptr)
768 if run_metadata:
769 proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
963 if final_fetches or final_targets:
964 results = self._do_run(handle, final_targets, final_fetches,
--> 965 feed_dict_string, options, run_metadata)
966 else:
967 results = []
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_run(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)
1013 if handle is None:
1014 return self._do_call(_run_fn, self._session, feed_dict, fetch_list,
-> 1015 target_list, options, run_metadata)
1016 else:
1017 return self._do_call(_prun_fn, self._session, handle, feed_dict,
/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1033 except KeyError:
1034 pass
-> 1035 raise type(e)(node_def, op, message)
1036
1037 def _extend_graph(self):
NotFoundError: Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i600_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_14/_29 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_158_save/RestoreV2_14", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
Caused by op 'save/RestoreV2', defined at:
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/runpy.py", line 193, in _run_module_as_main
"__main__", mod_spec)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/__main__.py", line 3, in <module>
app.launch_new_instance()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/traitlets/config/application.py", line 658, in launch_instance
app.start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelapp.py", line 474, in start
ioloop.IOLoop.instance().start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/ioloop.py", line 177, in start
super(ZMQIOLoop, self).start()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/ioloop.py", line 887, in start
handler_func(fd_obj, events)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/stack_context.py", line 275, in null_wrapper
return fn(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 440, in _handle_events
self._handle_recv()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 472, in _handle_recv
self._run_callback(callback, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 414, in _run_callback
callback(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tornado/stack_context.py", line 275, in null_wrapper
return fn(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 276, in dispatcher
return self.dispatch_shell(stream, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 228, in dispatch_shell
handler(stream, idents, msg)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 390, in execute_request
user_expressions, allow_stdin)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/ipkernel.py", line 196, in do_execute
res = shell.run_cell(code, store_history=store_history, silent=silent)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/ipykernel/zmqshell.py", line 501, in run_cell
return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2717, in run_cell
interactivity=interactivity, compiler=compiler, result=result)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2821, in run_ast_nodes
if self.run_code(code, result):
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2881, in run_code
exec(code_obj, self.user_global_ns, self.user_ns)
File "<ipython-input-57-7c4e18bbddc1>", line 2, in <module>
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
File "<ipython-input-53-8eb787ae9642>", line 4, in sample
saver = tf.train.Saver()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1040, in __init__
self.build()
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1070, in build
restore_sequentially=self._restore_sequentially)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 675, in build
restore_sequentially, reshape)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 402, in _AddRestoreOps
tensors = self.restore_op(filename_tensor, saveable, preferred_shard)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 242, in restore_op
[spec.tensor.dtype])[0])
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/ops/gen_io_ops.py", line 668, in restore_v2
dtypes=dtypes, name=name)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 763, in apply_op
op_def=op_def)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 2327, in create_op
original_op=self._default_original_op, op_def=op_def)
File "/home/luo/anaconda2/envs/dlnd/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1226, in __init__
self._traceback = _extract_stack()
NotFoundError (see above for traceback): Unsuccessful TensorSliceReader constructor: Failed to find any matching files for checkpoints/i600_l512.ckpt
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_FLOAT], _device="/job:localhost/replica:0/task:0/cpu:0"](_recv_save/Const_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2_14/_29 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/gpu:0", send_device="/job:localhost/replica:0/task:0/cpu:0", send_device_incarnation=1, tensor_name="edge_158_save/RestoreV2_14", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/gpu:0"]()]]
In [ ]:
checkpoint = 'checkpoints/i1200_l512.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Far")
print(samp)
Content source: luofan18/deep-learning
Similar notebooks: