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 = 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[:200]
Out[3]:
"Chapter 1\n\n\nHappy families are all alike; every unhappy family is unhappy in its own\nway.\n\nEverything was in confusion in the Oblonskys' house. The wife had\ndiscovered that the husband was carrying on"
And we can see the characters encoded as integers.
In [4]:
encoded[:200]
Out[4]:
array([55, 56, 0, 77, 44, 34, 75, 25, 81, 9, 9, 9, 43, 0, 77, 77, 28,
25, 1, 0, 35, 42, 62, 42, 34, 2, 25, 0, 75, 34, 25, 0, 62, 62,
25, 0, 62, 42, 3, 34, 6, 25, 34, 23, 34, 75, 28, 25, 67, 18, 56,
0, 77, 77, 28, 25, 1, 0, 35, 42, 62, 28, 25, 42, 2, 25, 67, 18,
56, 0, 77, 77, 28, 25, 42, 18, 25, 42, 44, 2, 25, 50, 45, 18, 9,
45, 0, 28, 53, 9, 9, 72, 23, 34, 75, 28, 44, 56, 42, 18, 15, 25,
45, 0, 2, 25, 42, 18, 25, 13, 50, 18, 1, 67, 2, 42, 50, 18, 25,
42, 18, 25, 44, 56, 34, 25, 39, 74, 62, 50, 18, 2, 3, 28, 2, 5,
25, 56, 50, 67, 2, 34, 53, 25, 73, 56, 34, 25, 45, 42, 1, 34, 25,
56, 0, 68, 9, 68, 42, 2, 13, 50, 23, 34, 75, 34, 68, 25, 44, 56,
0, 44, 25, 44, 56, 34, 25, 56, 67, 2, 74, 0, 18, 68, 25, 45, 0,
2, 25, 13, 0, 75, 75, 28, 42, 18, 15, 25, 50, 18], 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
In [6]:
' '.join(sorted(vocab))
Out[6]:
'\n ! " $ % & \' ( ) * , - . / 0 1 2 3 4 5 6 7 8 9 : ; ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z'
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 [7]:
5//3
Out[7]:
1
In [8]:
np.reshape([1,2,3,4,5,6], [2,-1])
Out[8]:
array([[1, 2, 3],
[4, 5, 6]])
In [9]:
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], y[:, -1] = x[:, 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 [10]:
batches = get_batches(encoded, n_seqs=10, n_steps=50)
x, y = next(batches)
In [11]:
print('x\n', x[:, :23])
print('\ny\n', y[:, :23])
x
[[55 56 0 77 44 34 75 25 81 9 9 9 43 0 77 77 28 25 1 0 35 42 62]
[25 0 35 25 18 50 44 25 15 50 42 18 15 25 44 50 25 2 44 0 28 76 14]
[23 42 18 53 9 9 14 20 34 2 76 25 42 44 5 2 25 2 34 44 44 62 34]
[18 25 68 67 75 42 18 15 25 56 42 2 25 13 50 18 23 34 75 2 0 44 42]
[25 42 44 25 42 2 76 25 2 42 75 46 14 25 2 0 42 68 25 44 56 34 25]
[25 78 44 25 45 0 2 9 50 18 62 28 25 45 56 34 18 25 44 56 34 25 2]
[56 34 18 25 13 50 35 34 25 1 50 75 25 35 34 76 14 25 2 56 34 25 2]
[ 6 25 74 67 44 25 18 50 45 25 2 56 34 25 45 50 67 62 68 25 75 34 0]
[44 25 42 2 18 5 44 53 25 73 56 34 28 5 75 34 25 77 75 50 77 75 42]
[25 2 0 42 68 25 44 50 25 56 34 75 2 34 62 1 76 25 0 18 68 25 74]]
y
[[56 0 77 44 34 75 25 81 9 9 9 43 0 77 77 28 25 1 0 35 42 62 42]
[ 0 35 25 18 50 44 25 15 50 42 18 15 25 44 50 25 2 44 0 28 76 14 25]
[42 18 53 9 9 14 20 34 2 76 25 42 44 5 2 25 2 34 44 44 62 34 68]
[25 68 67 75 42 18 15 25 56 42 2 25 13 50 18 23 34 75 2 0 44 42 50]
[42 44 25 42 2 76 25 2 42 75 46 14 25 2 0 42 68 25 44 56 34 25 50]
[78 44 25 45 0 2 9 50 18 62 28 25 45 56 34 18 25 44 56 34 25 2 0]
[34 18 25 13 50 35 34 25 1 50 75 25 35 34 76 14 25 2 56 34 25 2 0]
[25 74 67 44 25 18 50 45 25 2 56 34 25 45 50 67 62 68 25 75 34 0 68]
[25 42 2 18 5 44 53 25 73 56 34 28 5 75 34 25 77 75 50 77 75 42 34]
[ 2 0 42 68 25 44 50 25 56 34 75 2 34 62 1 76 25 0 18 68 25 74 34]]
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 [12]:
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 [ ]:
tf.__version__
In [20]:
def build_cell(num_units, keep_prob):
# Use a basic LSTM cell
lstm = tf.contrib.rnn.BasicLSTMCell(num_units)
# Add dropout to the cell outputs
drop = tf.contrib.rnn.DropoutWrapper(lstm, output_keep_prob=keep_prob)
return drop
def build_lstm(lstm_size, num_layers, batch_size, keep_prob):
''' Build LSTM cell.
Arguments
---------
lstm_size: Size of the hidden layers in the LSTM cells
num_layers: Number of LSTM layers
batch_size: Batch size
keep_prob: Scalar tensor (tf.placeholder) for the dropout keep probability
'''
### Build the LSTM Cell
# Stack up multiple LSTM layers, for deep learning
cell = tf.contrib.rnn.MultiRNNCell([build_cell(lstm_size, keep_prob) for _ in range(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 [27]:
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)
# Reshape seq_output to a 2D tensor with lstm_size columns
x = tf.reshape(tf.concat(lstm_output, axis=1), [-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))
softmax_b = tf.Variable(tf.zeros(out_size))
# 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 [33]:
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, logits.get_shape())
# 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 [34]:
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 [35]:
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 [46]:
batch_size = 20 # Sequences per batch
num_steps = 100 # Number of sequence steps per batch
lstm_size = 256 # Size of hidden layers in LSTMs
num_layers = 2 # Number of LSTM layers
learning_rate = 0.001 # Learning rate
keep_prob = 0.6 # 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 [47]:
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: 4.4175... 0.1048 sec/batch
Epoch: 1/20... Training Step: 2... Training loss: 4.3795... 0.0760 sec/batch
Epoch: 1/20... Training Step: 3... Training loss: 4.2966... 0.0777 sec/batch
Epoch: 1/20... Training Step: 4... Training loss: 4.0193... 0.0771 sec/batch
Epoch: 1/20... Training Step: 5... Training loss: 3.5692... 0.0741 sec/batch
Epoch: 1/20... Training Step: 6... Training loss: 3.5057... 0.0777 sec/batch
Epoch: 1/20... Training Step: 7... Training loss: 3.3735... 0.0802 sec/batch
Epoch: 1/20... Training Step: 8... Training loss: 3.4853... 0.0776 sec/batch
Epoch: 1/20... Training Step: 9... Training loss: 3.4063... 0.0762 sec/batch
Epoch: 1/20... Training Step: 10... Training loss: 3.3436... 0.0811 sec/batch
Epoch: 1/20... Training Step: 11... Training loss: 3.3227... 0.0749 sec/batch
Epoch: 1/20... Training Step: 12... Training loss: 3.3494... 0.0786 sec/batch
Epoch: 1/20... Training Step: 13... Training loss: 3.2998... 0.0781 sec/batch
Epoch: 1/20... Training Step: 14... Training loss: 3.3138... 0.0740 sec/batch
Epoch: 1/20... Training Step: 15... Training loss: 3.2944... 0.0797 sec/batch
Epoch: 1/20... Training Step: 16... Training loss: 3.3078... 0.0780 sec/batch
Epoch: 1/20... Training Step: 17... Training loss: 3.2421... 0.0772 sec/batch
Epoch: 1/20... Training Step: 18... Training loss: 3.2815... 0.0758 sec/batch
Epoch: 1/20... Training Step: 19... Training loss: 3.2235... 0.0764 sec/batch
Epoch: 1/20... Training Step: 20... Training loss: 3.2061... 0.0764 sec/batch
Epoch: 1/20... Training Step: 21... Training loss: 3.2582... 0.0765 sec/batch
Epoch: 1/20... Training Step: 22... Training loss: 3.2046... 0.0766 sec/batch
Epoch: 1/20... Training Step: 23... Training loss: 3.2400... 0.0793 sec/batch
Epoch: 1/20... Training Step: 24... Training loss: 3.2132... 0.0764 sec/batch
Epoch: 1/20... Training Step: 25... Training loss: 3.2359... 0.0778 sec/batch
Epoch: 1/20... Training Step: 26... Training loss: 3.1432... 0.0752 sec/batch
Epoch: 1/20... Training Step: 27... Training loss: 3.1834... 0.0754 sec/batch
Epoch: 1/20... Training Step: 28... Training loss: 3.2181... 0.0790 sec/batch
Epoch: 1/20... Training Step: 29... Training loss: 3.1900... 0.0775 sec/batch
Epoch: 1/20... Training Step: 30... Training loss: 3.1802... 0.0765 sec/batch
Epoch: 1/20... Training Step: 31... Training loss: 3.1986... 0.0764 sec/batch
Epoch: 1/20... Training Step: 32... Training loss: 3.1565... 0.0766 sec/batch
Epoch: 1/20... Training Step: 33... Training loss: 3.1662... 0.0772 sec/batch
Epoch: 1/20... Training Step: 34... Training loss: 3.1405... 0.0786 sec/batch
Epoch: 1/20... Training Step: 35... Training loss: 3.1897... 0.0801 sec/batch
Epoch: 1/20... Training Step: 36... Training loss: 3.1738... 0.0792 sec/batch
Epoch: 1/20... Training Step: 37... Training loss: 3.1909... 0.0759 sec/batch
Epoch: 1/20... Training Step: 38... Training loss: 3.1983... 0.0772 sec/batch
Epoch: 1/20... Training Step: 39... Training loss: 3.0788... 0.0786 sec/batch
Epoch: 1/20... Training Step: 40... Training loss: 3.1481... 0.0792 sec/batch
Epoch: 1/20... Training Step: 41... Training loss: 3.1650... 0.0764 sec/batch
Epoch: 1/20... Training Step: 42... Training loss: 3.1595... 0.0791 sec/batch
Epoch: 1/20... Training Step: 43... Training loss: 3.1548... 0.0785 sec/batch
Epoch: 1/20... Training Step: 44... Training loss: 3.1340... 0.0761 sec/batch
Epoch: 1/20... Training Step: 45... Training loss: 3.1656... 0.0762 sec/batch
Epoch: 1/20... Training Step: 46... Training loss: 3.1870... 0.0757 sec/batch
Epoch: 1/20... Training Step: 47... Training loss: 3.2087... 0.0813 sec/batch
Epoch: 1/20... Training Step: 48... Training loss: 3.2082... 0.0833 sec/batch
Epoch: 1/20... Training Step: 49... Training loss: 3.2166... 0.0813 sec/batch
Epoch: 1/20... Training Step: 50... Training loss: 3.1635... 0.0760 sec/batch
Epoch: 1/20... Training Step: 51... Training loss: 3.1331... 0.0774 sec/batch
Epoch: 1/20... Training Step: 52... Training loss: 3.1671... 0.0811 sec/batch
Epoch: 1/20... Training Step: 53... Training loss: 3.1275... 0.0762 sec/batch
Epoch: 1/20... Training Step: 54... Training loss: 3.1190... 0.0745 sec/batch
Epoch: 1/20... Training Step: 55... Training loss: 3.1420... 0.0776 sec/batch
Epoch: 1/20... Training Step: 56... Training loss: 3.1283... 0.0785 sec/batch
Epoch: 1/20... Training Step: 57... Training loss: 3.1526... 0.0767 sec/batch
Epoch: 1/20... Training Step: 58... Training loss: 3.1607... 0.0790 sec/batch
Epoch: 1/20... Training Step: 59... Training loss: 3.1510... 0.0779 sec/batch
Epoch: 1/20... Training Step: 60... Training loss: 3.1487... 0.0794 sec/batch
Epoch: 1/20... Training Step: 61... Training loss: 3.1565... 0.0813 sec/batch
Epoch: 1/20... Training Step: 62... Training loss: 3.1602... 0.0781 sec/batch
Epoch: 1/20... Training Step: 63... Training loss: 3.1847... 0.0781 sec/batch
Epoch: 1/20... Training Step: 64... Training loss: 3.1190... 0.0754 sec/batch
Epoch: 1/20... Training Step: 65... Training loss: 3.1251... 0.0792 sec/batch
Epoch: 1/20... Training Step: 66... Training loss: 3.1871... 0.0837 sec/batch
Epoch: 1/20... Training Step: 67... Training loss: 3.1185... 0.0887 sec/batch
Epoch: 1/20... Training Step: 68... Training loss: 3.1121... 0.0830 sec/batch
Epoch: 1/20... Training Step: 69... Training loss: 3.1527... 0.0826 sec/batch
Epoch: 1/20... Training Step: 70... Training loss: 3.1148... 0.0802 sec/batch
Epoch: 1/20... Training Step: 71... Training loss: 3.1653... 0.0776 sec/batch
Epoch: 1/20... Training Step: 72... Training loss: 3.1350... 0.0793 sec/batch
Epoch: 1/20... Training Step: 73... Training loss: 3.1087... 0.0757 sec/batch
Epoch: 1/20... Training Step: 74... Training loss: 3.1349... 0.0760 sec/batch
Epoch: 1/20... Training Step: 75... Training loss: 3.1209... 0.0760 sec/batch
Epoch: 1/20... Training Step: 76... Training loss: 3.1285... 0.0792 sec/batch
Epoch: 1/20... Training Step: 77... Training loss: 3.1856... 0.0746 sec/batch
Epoch: 1/20... Training Step: 78... Training loss: 3.1504... 0.0771 sec/batch
Epoch: 1/20... Training Step: 79... Training loss: 3.1501... 0.0770 sec/batch
Epoch: 1/20... Training Step: 80... Training loss: 3.1033... 0.0769 sec/batch
Epoch: 1/20... Training Step: 81... Training loss: 3.1107... 0.0804 sec/batch
Epoch: 1/20... Training Step: 82... Training loss: 3.1172... 0.0780 sec/batch
Epoch: 1/20... Training Step: 83... Training loss: 3.1190... 0.0760 sec/batch
Epoch: 1/20... Training Step: 84... Training loss: 3.1420... 0.0785 sec/batch
Epoch: 1/20... Training Step: 85... Training loss: 3.1822... 0.0758 sec/batch
Epoch: 1/20... Training Step: 86... Training loss: 3.1418... 0.0793 sec/batch
Epoch: 1/20... Training Step: 87... Training loss: 3.1405... 0.0768 sec/batch
Epoch: 1/20... Training Step: 88... Training loss: 3.1126... 0.0765 sec/batch
Epoch: 1/20... Training Step: 89... Training loss: 3.1654... 0.0750 sec/batch
Epoch: 1/20... Training Step: 90... Training loss: 3.1468... 0.0776 sec/batch
Epoch: 1/20... Training Step: 91... Training loss: 3.1321... 0.0795 sec/batch
Epoch: 1/20... Training Step: 92... Training loss: 3.0831... 0.0771 sec/batch
Epoch: 1/20... Training Step: 93... Training loss: 3.0810... 0.0757 sec/batch
Epoch: 1/20... Training Step: 94... Training loss: 3.0951... 0.0779 sec/batch
Epoch: 1/20... Training Step: 95... Training loss: 3.0726... 0.0771 sec/batch
Epoch: 1/20... Training Step: 96... Training loss: 3.1384... 0.0799 sec/batch
Epoch: 1/20... Training Step: 97... Training loss: 3.1707... 0.0759 sec/batch
Epoch: 1/20... Training Step: 98... Training loss: 3.1591... 0.0765 sec/batch
Epoch: 1/20... Training Step: 99... Training loss: 3.1117... 0.0757 sec/batch
Epoch: 1/20... Training Step: 100... Training loss: 3.1578... 0.0799 sec/batch
Epoch: 1/20... Training Step: 101... Training loss: 3.1486... 0.0767 sec/batch
Epoch: 1/20... Training Step: 102... Training loss: 3.1946... 0.0749 sec/batch
Epoch: 1/20... Training Step: 103... Training loss: 3.0991... 0.0787 sec/batch
Epoch: 1/20... Training Step: 104... Training loss: 3.1508... 0.0786 sec/batch
Epoch: 1/20... Training Step: 105... Training loss: 3.1511... 0.0770 sec/batch
Epoch: 1/20... Training Step: 106... Training loss: 3.0918... 0.0789 sec/batch
Epoch: 1/20... Training Step: 107... Training loss: 3.1220... 0.0756 sec/batch
Epoch: 1/20... Training Step: 108... Training loss: 3.1680... 0.0792 sec/batch
Epoch: 1/20... Training Step: 109... Training loss: 3.1022... 0.0781 sec/batch
Epoch: 1/20... Training Step: 110... Training loss: 3.0621... 0.0796 sec/batch
Epoch: 1/20... Training Step: 111... Training loss: 3.1066... 0.0792 sec/batch
Epoch: 1/20... Training Step: 112... Training loss: 3.0760... 0.0766 sec/batch
Epoch: 1/20... Training Step: 113... Training loss: 3.1274... 0.0775 sec/batch
Epoch: 1/20... Training Step: 114... Training loss: 3.1101... 0.0773 sec/batch
Epoch: 1/20... Training Step: 115... Training loss: 3.1297... 0.0779 sec/batch
Epoch: 1/20... Training Step: 116... Training loss: 3.1066... 0.0793 sec/batch
Epoch: 1/20... Training Step: 117... Training loss: 3.0938... 0.0776 sec/batch
Epoch: 1/20... Training Step: 118... Training loss: 3.1235... 0.0752 sec/batch
Epoch: 1/20... Training Step: 119... Training loss: 3.0710... 0.0757 sec/batch
Epoch: 1/20... Training Step: 120... Training loss: 3.1047... 0.0755 sec/batch
Epoch: 1/20... Training Step: 121... Training loss: 3.1268... 0.0777 sec/batch
Epoch: 1/20... Training Step: 122... Training loss: 3.1694... 0.0750 sec/batch
Epoch: 1/20... Training Step: 123... Training loss: 3.0857... 0.0799 sec/batch
Epoch: 1/20... Training Step: 124... Training loss: 3.1184... 0.0769 sec/batch
Epoch: 1/20... Training Step: 125... Training loss: 3.1328... 0.0764 sec/batch
Epoch: 1/20... Training Step: 126... Training loss: 3.1058... 0.0762 sec/batch
Epoch: 1/20... Training Step: 127... Training loss: 3.0955... 0.0807 sec/batch
Epoch: 1/20... Training Step: 128... Training loss: 3.1026... 0.0758 sec/batch
Epoch: 1/20... Training Step: 129... Training loss: 3.0751... 0.0839 sec/batch
Epoch: 1/20... Training Step: 130... Training loss: 3.0803... 0.0776 sec/batch
Epoch: 1/20... Training Step: 131... Training loss: 3.0425... 0.0791 sec/batch
Epoch: 1/20... Training Step: 132... Training loss: 3.0560... 0.0759 sec/batch
Epoch: 1/20... Training Step: 133... Training loss: 3.0229... 0.0763 sec/batch
Epoch: 1/20... Training Step: 134... Training loss: 3.0098... 0.0771 sec/batch
Epoch: 1/20... Training Step: 135... Training loss: 3.0572... 0.0776 sec/batch
Epoch: 1/20... Training Step: 136... Training loss: 3.0253... 0.0774 sec/batch
Epoch: 1/20... Training Step: 137... Training loss: 3.0100... 0.0799 sec/batch
Epoch: 1/20... Training Step: 138... Training loss: 3.0101... 0.0777 sec/batch
Epoch: 1/20... Training Step: 139... Training loss: 3.0634... 0.0783 sec/batch
Epoch: 1/20... Training Step: 140... Training loss: 2.9695... 0.0767 sec/batch
Epoch: 1/20... Training Step: 141... Training loss: 3.0212... 0.0755 sec/batch
Epoch: 1/20... Training Step: 142... Training loss: 2.9809... 0.0792 sec/batch
Epoch: 1/20... Training Step: 143... Training loss: 2.9953... 0.0767 sec/batch
Epoch: 1/20... Training Step: 144... Training loss: 2.9367... 0.0754 sec/batch
Epoch: 1/20... Training Step: 145... Training loss: 2.9803... 0.0758 sec/batch
Epoch: 1/20... Training Step: 146... Training loss: 2.9805... 0.0768 sec/batch
Epoch: 1/20... Training Step: 147... Training loss: 2.9410... 0.0742 sec/batch
Epoch: 1/20... Training Step: 148... Training loss: 2.9153... 0.0756 sec/batch
Epoch: 1/20... Training Step: 149... Training loss: 2.9390... 0.0780 sec/batch
Epoch: 1/20... Training Step: 150... Training loss: 2.9936... 0.0775 sec/batch
Epoch: 1/20... Training Step: 151... Training loss: 2.8905... 0.0881 sec/batch
Epoch: 1/20... Training Step: 152... Training loss: 2.9390... 0.0793 sec/batch
Epoch: 1/20... Training Step: 153... Training loss: 2.9023... 0.0772 sec/batch
Epoch: 1/20... Training Step: 154... Training loss: 2.8841... 0.0784 sec/batch
Epoch: 1/20... Training Step: 155... Training loss: 2.9095... 0.0781 sec/batch
Epoch: 1/20... Training Step: 156... Training loss: 2.8360... 0.0761 sec/batch
Epoch: 1/20... Training Step: 157... Training loss: 2.8928... 0.0759 sec/batch
Epoch: 1/20... Training Step: 158... Training loss: 2.8488... 0.0778 sec/batch
Epoch: 1/20... Training Step: 159... Training loss: 2.8748... 0.0771 sec/batch
Epoch: 1/20... Training Step: 160... Training loss: 2.9012... 0.0789 sec/batch
Epoch: 1/20... Training Step: 161... Training loss: 2.8182... 0.0758 sec/batch
Epoch: 1/20... Training Step: 162... Training loss: 2.8028... 0.0754 sec/batch
Epoch: 1/20... Training Step: 163... Training loss: 2.8375... 0.0777 sec/batch
Epoch: 1/20... Training Step: 164... Training loss: 2.8448... 0.0746 sec/batch
Epoch: 1/20... Training Step: 165... Training loss: 2.8172... 0.0778 sec/batch
Epoch: 1/20... Training Step: 166... Training loss: 2.8314... 0.0753 sec/batch
Epoch: 1/20... Training Step: 167... Training loss: 2.7756... 0.0743 sec/batch
Epoch: 1/20... Training Step: 168... Training loss: 2.8774... 0.0766 sec/batch
Epoch: 1/20... Training Step: 169... Training loss: 2.8429... 0.0764 sec/batch
Epoch: 1/20... Training Step: 170... Training loss: 2.8123... 0.0756 sec/batch
Epoch: 1/20... Training Step: 171... Training loss: 2.8385... 0.0768 sec/batch
Epoch: 1/20... Training Step: 172... Training loss: 2.7600... 0.0794 sec/batch
Epoch: 1/20... Training Step: 173... Training loss: 2.7238... 0.0774 sec/batch
Epoch: 1/20... Training Step: 174... Training loss: 2.8109... 0.0761 sec/batch
Epoch: 1/20... Training Step: 175... Training loss: 2.8110... 0.0784 sec/batch
Epoch: 1/20... Training Step: 176... Training loss: 2.7568... 0.0767 sec/batch
Epoch: 1/20... Training Step: 177... Training loss: 2.7574... 0.0774 sec/batch
Epoch: 1/20... Training Step: 178... Training loss: 2.8165... 0.0791 sec/batch
Epoch: 1/20... Training Step: 179... Training loss: 2.8089... 0.0770 sec/batch
Epoch: 1/20... Training Step: 180... Training loss: 2.7603... 0.0753 sec/batch
Epoch: 1/20... Training Step: 181... Training loss: 2.8194... 0.0797 sec/batch
Epoch: 1/20... Training Step: 182... Training loss: 2.7258... 0.0767 sec/batch
Epoch: 1/20... Training Step: 183... Training loss: 2.6714... 0.0761 sec/batch
Epoch: 1/20... Training Step: 184... Training loss: 2.6889... 0.0855 sec/batch
Epoch: 1/20... Training Step: 185... Training loss: 2.6862... 0.0747 sec/batch
Epoch: 1/20... Training Step: 186... Training loss: 2.7496... 0.0753 sec/batch
Epoch: 1/20... Training Step: 187... Training loss: 2.7379... 0.0792 sec/batch
Epoch: 1/20... Training Step: 188... Training loss: 2.6834... 0.0777 sec/batch
Epoch: 1/20... Training Step: 189... Training loss: 2.6288... 0.0797 sec/batch
Epoch: 1/20... Training Step: 190... Training loss: 2.6904... 0.0787 sec/batch
Epoch: 1/20... Training Step: 191... Training loss: 2.6529... 0.0800 sec/batch
Epoch: 1/20... Training Step: 192... Training loss: 2.7353... 0.0757 sec/batch
Epoch: 1/20... Training Step: 193... Training loss: 2.6392... 0.0758 sec/batch
Epoch: 1/20... Training Step: 194... Training loss: 2.6881... 0.0750 sec/batch
Epoch: 1/20... Training Step: 195... Training loss: 2.6821... 0.0785 sec/batch
Epoch: 1/20... Training Step: 196... Training loss: 2.6533... 0.0771 sec/batch
Epoch: 1/20... Training Step: 197... Training loss: 2.6349... 0.0764 sec/batch
Epoch: 1/20... Training Step: 198... Training loss: 2.6573... 0.0746 sec/batch
Epoch: 1/20... Training Step: 199... Training loss: 2.6192... 0.0771 sec/batch
Epoch: 1/20... Training Step: 200... Training loss: 2.6514... 0.0784 sec/batch
Epoch: 1/20... Training Step: 201... Training loss: 2.6225... 0.0783 sec/batch
Epoch: 1/20... Training Step: 202... Training loss: 2.6553... 0.0772 sec/batch
Epoch: 1/20... Training Step: 203... Training loss: 2.5763... 0.0889 sec/batch
Epoch: 1/20... Training Step: 204... Training loss: 2.6485... 0.0806 sec/batch
Epoch: 1/20... Training Step: 205... Training loss: 2.6356... 0.0794 sec/batch
Epoch: 1/20... Training Step: 206... Training loss: 2.6947... 0.0785 sec/batch
Epoch: 1/20... Training Step: 207... Training loss: 2.5938... 0.0781 sec/batch
Epoch: 1/20... Training Step: 208... Training loss: 2.5485... 0.0811 sec/batch
Epoch: 1/20... Training Step: 209... Training loss: 2.6304... 0.0806 sec/batch
Epoch: 1/20... Training Step: 210... Training loss: 2.5931... 0.0796 sec/batch
Epoch: 1/20... Training Step: 211... Training loss: 2.5870... 0.0765 sec/batch
Epoch: 1/20... Training Step: 212... Training loss: 2.6036... 0.0776 sec/batch
Epoch: 1/20... Training Step: 213... Training loss: 2.5965... 0.0778 sec/batch
Epoch: 1/20... Training Step: 214... Training loss: 2.5868... 0.0759 sec/batch
Epoch: 1/20... Training Step: 215... Training loss: 2.6020... 0.0746 sec/batch
Epoch: 1/20... Training Step: 216... Training loss: 2.5926... 0.0774 sec/batch
Epoch: 1/20... Training Step: 217... Training loss: 2.5322... 0.0770 sec/batch
Epoch: 1/20... Training Step: 218... Training loss: 2.6037... 0.0783 sec/batch
Epoch: 1/20... Training Step: 219... Training loss: 2.5806... 0.0853 sec/batch
Epoch: 1/20... Training Step: 220... Training loss: 2.5522... 0.0750 sec/batch
Epoch: 1/20... Training Step: 221... Training loss: 2.5025... 0.0747 sec/batch
Epoch: 1/20... Training Step: 222... Training loss: 2.5830... 0.0788 sec/batch
Epoch: 1/20... Training Step: 223... Training loss: 2.5428... 0.0784 sec/batch
Epoch: 1/20... Training Step: 224... Training loss: 2.5732... 0.0785 sec/batch
Epoch: 1/20... Training Step: 225... Training loss: 2.6109... 0.0783 sec/batch
Epoch: 1/20... Training Step: 226... Training loss: 2.5544... 0.0749 sec/batch
Epoch: 1/20... Training Step: 227... Training loss: 2.5615... 0.0788 sec/batch
Epoch: 1/20... Training Step: 228... Training loss: 2.4971... 0.0880 sec/batch
Epoch: 1/20... Training Step: 229... Training loss: 2.5457... 0.0785 sec/batch
Epoch: 1/20... Training Step: 230... Training loss: 2.6229... 0.0785 sec/batch
Epoch: 1/20... Training Step: 231... Training loss: 2.5310... 0.0830 sec/batch
Epoch: 1/20... Training Step: 232... Training loss: 2.5518... 0.0793 sec/batch
Epoch: 1/20... Training Step: 233... Training loss: 2.5303... 0.0757 sec/batch
Epoch: 1/20... Training Step: 234... Training loss: 2.5745... 0.0786 sec/batch
Epoch: 1/20... Training Step: 235... Training loss: 2.5042... 0.0737 sec/batch
Epoch: 1/20... Training Step: 236... Training loss: 2.5508... 0.0772 sec/batch
Epoch: 1/20... Training Step: 237... Training loss: 2.4566... 0.0752 sec/batch
Epoch: 1/20... Training Step: 238... Training loss: 2.5588... 0.0789 sec/batch
Epoch: 1/20... Training Step: 239... Training loss: 2.5308... 0.0759 sec/batch
Epoch: 1/20... Training Step: 240... Training loss: 2.5540... 0.0767 sec/batch
Epoch: 1/20... Training Step: 241... Training loss: 2.5498... 0.0812 sec/batch
Epoch: 1/20... Training Step: 242... Training loss: 2.4394... 0.0774 sec/batch
Epoch: 1/20... Training Step: 243... Training loss: 2.4988... 0.0860 sec/batch
Epoch: 1/20... Training Step: 244... Training loss: 2.5576... 0.0759 sec/batch
Epoch: 1/20... Training Step: 245... Training loss: 2.5195... 0.0757 sec/batch
Epoch: 1/20... Training Step: 246... Training loss: 2.5614... 0.0792 sec/batch
Epoch: 1/20... Training Step: 247... Training loss: 2.5549... 0.0779 sec/batch
Epoch: 1/20... Training Step: 248... Training loss: 2.4939... 0.0769 sec/batch
Epoch: 1/20... Training Step: 249... Training loss: 2.4820... 0.0778 sec/batch
Epoch: 1/20... Training Step: 250... Training loss: 2.4648... 0.0811 sec/batch
Epoch: 1/20... Training Step: 251... Training loss: 2.5340... 0.0766 sec/batch
Epoch: 1/20... Training Step: 252... Training loss: 2.5388... 0.0790 sec/batch
Epoch: 1/20... Training Step: 253... Training loss: 2.4956... 0.0764 sec/batch
Epoch: 1/20... Training Step: 254... Training loss: 2.5231... 0.0792 sec/batch
Epoch: 1/20... Training Step: 255... Training loss: 2.5978... 0.0777 sec/batch
Epoch: 1/20... Training Step: 256... Training loss: 2.5574... 0.0761 sec/batch
Epoch: 1/20... Training Step: 257... Training loss: 2.5522... 0.0769 sec/batch
Epoch: 1/20... Training Step: 258... Training loss: 2.5003... 0.0767 sec/batch
Epoch: 1/20... Training Step: 259... Training loss: 2.4989... 0.0764 sec/batch
Epoch: 1/20... Training Step: 260... Training loss: 2.5101... 0.0743 sec/batch
Epoch: 1/20... Training Step: 261... Training loss: 2.5443... 0.0796 sec/batch
Epoch: 1/20... Training Step: 262... Training loss: 2.5228... 0.0772 sec/batch
Epoch: 1/20... Training Step: 263... Training loss: 2.5076... 0.0766 sec/batch
Epoch: 1/20... Training Step: 264... Training loss: 2.4987... 0.0788 sec/batch
Epoch: 1/20... Training Step: 265... Training loss: 2.5194... 0.0771 sec/batch
Epoch: 1/20... Training Step: 266... Training loss: 2.5266... 0.0764 sec/batch
Epoch: 1/20... Training Step: 267... Training loss: 2.4207... 0.0786 sec/batch
Epoch: 1/20... Training Step: 268... Training loss: 2.4562... 0.0750 sec/batch
Epoch: 1/20... Training Step: 269... Training loss: 2.4554... 0.0750 sec/batch
Epoch: 1/20... Training Step: 270... Training loss: 2.4736... 0.0776 sec/batch
Epoch: 1/20... Training Step: 271... Training loss: 2.4606... 0.0768 sec/batch
Epoch: 1/20... Training Step: 272... Training loss: 2.4525... 0.0752 sec/batch
Epoch: 1/20... Training Step: 273... Training loss: 2.4406... 0.0777 sec/batch
Epoch: 1/20... Training Step: 274... Training loss: 2.4401... 0.0765 sec/batch
Epoch: 1/20... Training Step: 275... Training loss: 2.4524... 0.0769 sec/batch
Epoch: 1/20... Training Step: 276... Training loss: 2.4546... 0.0801 sec/batch
Epoch: 1/20... Training Step: 277... Training loss: 2.4800... 0.0777 sec/batch
Epoch: 1/20... Training Step: 278... Training loss: 2.4618... 0.0773 sec/batch
Epoch: 1/20... Training Step: 279... Training loss: 2.5144... 0.0779 sec/batch
Epoch: 1/20... Training Step: 280... Training loss: 2.5052... 0.0814 sec/batch
Epoch: 1/20... Training Step: 281... Training loss: 2.4553... 0.0760 sec/batch
Epoch: 1/20... Training Step: 282... Training loss: 2.5899... 0.0796 sec/batch
Epoch: 1/20... Training Step: 283... Training loss: 2.4380... 0.0774 sec/batch
Epoch: 1/20... Training Step: 284... Training loss: 2.4233... 0.0743 sec/batch
Epoch: 1/20... Training Step: 285... Training loss: 2.5519... 0.0788 sec/batch
Epoch: 1/20... Training Step: 286... Training loss: 2.4401... 0.0759 sec/batch
Epoch: 1/20... Training Step: 287... Training loss: 2.4432... 0.0790 sec/batch
Epoch: 1/20... Training Step: 288... Training loss: 2.3777... 0.0783 sec/batch
Epoch: 1/20... Training Step: 289... Training loss: 2.4869... 0.0773 sec/batch
Epoch: 1/20... Training Step: 290... Training loss: 2.4706... 0.0818 sec/batch
Epoch: 1/20... Training Step: 291... Training loss: 2.4633... 0.0787 sec/batch
Epoch: 1/20... Training Step: 292... Training loss: 2.4519... 0.0762 sec/batch
Epoch: 1/20... Training Step: 293... Training loss: 2.4859... 0.0751 sec/batch
Epoch: 1/20... Training Step: 294... Training loss: 2.4506... 0.0769 sec/batch
Epoch: 1/20... Training Step: 295... Training loss: 2.4684... 0.0765 sec/batch
Epoch: 1/20... Training Step: 296... Training loss: 2.4740... 0.0748 sec/batch
Epoch: 1/20... Training Step: 297... Training loss: 2.4591... 0.0806 sec/batch
Epoch: 1/20... Training Step: 298... Training loss: 2.4425... 0.0775 sec/batch
Epoch: 1/20... Training Step: 299... Training loss: 2.4440... 0.0766 sec/batch
Epoch: 1/20... Training Step: 300... Training loss: 2.4731... 0.0781 sec/batch
Epoch: 1/20... Training Step: 301... Training loss: 2.4834... 0.0764 sec/batch
Epoch: 1/20... Training Step: 302... Training loss: 2.5026... 0.0743 sec/batch
Epoch: 1/20... Training Step: 303... Training loss: 2.4253... 0.0803 sec/batch
Epoch: 1/20... Training Step: 304... Training loss: 2.4319... 0.0778 sec/batch
Epoch: 1/20... Training Step: 305... Training loss: 2.4924... 0.0779 sec/batch
Epoch: 1/20... Training Step: 306... Training loss: 2.5118... 0.0800 sec/batch
Epoch: 1/20... Training Step: 307... Training loss: 2.4384... 0.0774 sec/batch
Epoch: 1/20... Training Step: 308... Training loss: 2.3789... 0.0777 sec/batch
Epoch: 1/20... Training Step: 309... Training loss: 2.4622... 0.0769 sec/batch
Epoch: 1/20... Training Step: 310... Training loss: 2.4214... 0.0758 sec/batch
Epoch: 1/20... Training Step: 311... Training loss: 2.4423... 0.0760 sec/batch
Epoch: 1/20... Training Step: 312... Training loss: 2.4374... 0.0764 sec/batch
Epoch: 1/20... Training Step: 313... Training loss: 2.4198... 0.0746 sec/batch
Epoch: 1/20... Training Step: 314... Training loss: 2.4124... 0.0769 sec/batch
Epoch: 1/20... Training Step: 315... Training loss: 2.4161... 0.0793 sec/batch
Epoch: 1/20... Training Step: 316... Training loss: 2.3509... 0.0766 sec/batch
Epoch: 1/20... Training Step: 317... Training loss: 2.3723... 0.0764 sec/batch
Epoch: 1/20... Training Step: 318... Training loss: 2.4239... 0.0798 sec/batch
Epoch: 1/20... Training Step: 319... Training loss: 2.3689... 0.0764 sec/batch
Epoch: 1/20... Training Step: 320... Training loss: 2.4227... 0.0744 sec/batch
Epoch: 1/20... Training Step: 321... Training loss: 2.4504... 0.0795 sec/batch
Epoch: 1/20... Training Step: 322... Training loss: 2.3739... 0.0758 sec/batch
Epoch: 1/20... Training Step: 323... Training loss: 2.3774... 0.0767 sec/batch
Epoch: 1/20... Training Step: 324... Training loss: 2.3637... 0.0790 sec/batch
Epoch: 1/20... Training Step: 325... Training loss: 2.3632... 0.0755 sec/batch
Epoch: 1/20... Training Step: 326... Training loss: 2.3730... 0.0790 sec/batch
Epoch: 1/20... Training Step: 327... Training loss: 2.3874... 0.0760 sec/batch
Epoch: 1/20... Training Step: 328... Training loss: 2.3425... 0.0772 sec/batch
Epoch: 1/20... Training Step: 329... Training loss: 2.3770... 0.0750 sec/batch
Epoch: 1/20... Training Step: 330... Training loss: 2.4142... 0.0766 sec/batch
Epoch: 1/20... Training Step: 331... Training loss: 2.3655... 0.0790 sec/batch
Epoch: 1/20... Training Step: 332... Training loss: 2.4043... 0.0779 sec/batch
Epoch: 1/20... Training Step: 333... Training loss: 2.4355... 0.0780 sec/batch
Epoch: 1/20... Training Step: 334... Training loss: 2.4112... 0.0749 sec/batch
Epoch: 1/20... Training Step: 335... Training loss: 2.4418... 0.0784 sec/batch
Epoch: 1/20... Training Step: 336... Training loss: 2.3850... 0.0767 sec/batch
Epoch: 1/20... Training Step: 337... Training loss: 2.4763... 0.0780 sec/batch
Epoch: 1/20... Training Step: 338... Training loss: 2.3668... 0.0767 sec/batch
Epoch: 1/20... Training Step: 339... Training loss: 2.4846... 0.0783 sec/batch
Epoch: 1/20... Training Step: 340... Training loss: 2.4006... 0.0763 sec/batch
Epoch: 1/20... Training Step: 341... Training loss: 2.4011... 0.0760 sec/batch
Epoch: 1/20... Training Step: 342... Training loss: 2.3713... 0.0790 sec/batch
Epoch: 1/20... Training Step: 343... Training loss: 2.3157... 0.0796 sec/batch
Epoch: 1/20... Training Step: 344... Training loss: 2.3921... 0.0771 sec/batch
Epoch: 1/20... Training Step: 345... Training loss: 2.3814... 0.0768 sec/batch
Epoch: 1/20... Training Step: 346... Training loss: 2.4342... 0.0757 sec/batch
Epoch: 1/20... Training Step: 347... Training loss: 2.3940... 0.0745 sec/batch
Epoch: 1/20... Training Step: 348... Training loss: 2.3711... 0.0763 sec/batch
Epoch: 1/20... Training Step: 349... Training loss: 2.3960... 0.0744 sec/batch
Epoch: 1/20... Training Step: 350... Training loss: 2.4733... 0.0756 sec/batch
Epoch: 1/20... Training Step: 351... Training loss: 2.3654... 0.0760 sec/batch
Epoch: 1/20... Training Step: 352... Training loss: 2.3486... 0.0773 sec/batch
Epoch: 1/20... Training Step: 353... Training loss: 2.3615... 0.0767 sec/batch
Epoch: 1/20... Training Step: 354... Training loss: 2.3955... 0.0773 sec/batch
Epoch: 1/20... Training Step: 355... Training loss: 2.3187... 0.0775 sec/batch
Epoch: 1/20... Training Step: 356... Training loss: 2.3781... 0.0758 sec/batch
Epoch: 1/20... Training Step: 357... Training loss: 2.3796... 0.0750 sec/batch
Epoch: 1/20... Training Step: 358... Training loss: 2.3800... 0.0781 sec/batch
Epoch: 1/20... Training Step: 359... Training loss: 2.3766... 0.0813 sec/batch
Epoch: 1/20... Training Step: 360... Training loss: 2.3536... 0.0817 sec/batch
Epoch: 1/20... Training Step: 361... Training loss: 2.4011... 0.0799 sec/batch
Epoch: 1/20... Training Step: 362... Training loss: 2.3964... 0.0803 sec/batch
Epoch: 1/20... Training Step: 363... Training loss: 2.3951... 0.0777 sec/batch
Epoch: 1/20... Training Step: 364... Training loss: 2.4099... 0.0798 sec/batch
Epoch: 1/20... Training Step: 365... Training loss: 2.3147... 0.0802 sec/batch
Epoch: 1/20... Training Step: 366... Training loss: 2.3550... 0.0772 sec/batch
Epoch: 1/20... Training Step: 367... Training loss: 2.3444... 0.0748 sec/batch
Epoch: 1/20... Training Step: 368... Training loss: 2.3166... 0.0768 sec/batch
Epoch: 1/20... Training Step: 369... Training loss: 2.3934... 0.0868 sec/batch
Epoch: 1/20... Training Step: 370... Training loss: 2.3762... 0.0749 sec/batch
Epoch: 1/20... Training Step: 371... Training loss: 2.3463... 0.0751 sec/batch
Epoch: 1/20... Training Step: 372... Training loss: 2.3409... 0.0800 sec/batch
Epoch: 1/20... Training Step: 373... Training loss: 2.3736... 0.0779 sec/batch
Epoch: 1/20... Training Step: 374... Training loss: 2.3532... 0.0749 sec/batch
Epoch: 1/20... Training Step: 375... Training loss: 2.3755... 0.0765 sec/batch
Epoch: 1/20... Training Step: 376... Training loss: 2.3024... 0.0787 sec/batch
Epoch: 1/20... Training Step: 377... Training loss: 2.3148... 0.0797 sec/batch
Epoch: 1/20... Training Step: 378... Training loss: 2.3888... 0.0775 sec/batch
Epoch: 1/20... Training Step: 379... Training loss: 2.3990... 0.0755 sec/batch
Epoch: 1/20... Training Step: 380... Training loss: 2.3323... 0.0781 sec/batch
Epoch: 1/20... Training Step: 381... Training loss: 2.3801... 0.0787 sec/batch
Epoch: 1/20... Training Step: 382... Training loss: 2.3162... 0.0822 sec/batch
Epoch: 1/20... Training Step: 383... Training loss: 2.3225... 0.0778 sec/batch
Epoch: 1/20... Training Step: 384... Training loss: 2.4307... 0.0760 sec/batch
Epoch: 1/20... Training Step: 385... Training loss: 2.3459... 0.0732 sec/batch
Epoch: 1/20... Training Step: 386... Training loss: 2.3352... 0.0750 sec/batch
Epoch: 1/20... Training Step: 387... Training loss: 2.3763... 0.0791 sec/batch
Epoch: 1/20... Training Step: 388... Training loss: 2.3061... 0.0758 sec/batch
Epoch: 1/20... Training Step: 389... Training loss: 2.3929... 0.0758 sec/batch
Epoch: 1/20... Training Step: 390... Training loss: 2.3211... 0.0746 sec/batch
Epoch: 1/20... Training Step: 391... Training loss: 2.3164... 0.0783 sec/batch
Epoch: 1/20... Training Step: 392... Training loss: 2.3140... 0.0747 sec/batch
Epoch: 1/20... Training Step: 393... Training loss: 2.3907... 0.0790 sec/batch
Epoch: 1/20... Training Step: 394... Training loss: 2.2676... 0.0744 sec/batch
Epoch: 1/20... Training Step: 395... Training loss: 2.3304... 0.0743 sec/batch
Epoch: 1/20... Training Step: 396... Training loss: 2.4324... 0.0778 sec/batch
Epoch: 1/20... Training Step: 397... Training loss: 2.3256... 0.0741 sec/batch
Epoch: 1/20... Training Step: 398... Training loss: 2.3197... 0.0771 sec/batch
Epoch: 1/20... Training Step: 399... Training loss: 2.3901... 0.0786 sec/batch
Epoch: 1/20... Training Step: 400... Training loss: 2.3096... 0.0755 sec/batch
Epoch: 1/20... Training Step: 401... Training loss: 2.3040... 0.0758 sec/batch
Epoch: 1/20... Training Step: 402... Training loss: 2.3376... 0.0812 sec/batch
Epoch: 1/20... Training Step: 403... Training loss: 2.3071... 0.0792 sec/batch
Epoch: 1/20... Training Step: 404... Training loss: 2.3705... 0.0762 sec/batch
Epoch: 1/20... Training Step: 405... Training loss: 2.3018... 0.0771 sec/batch
Epoch: 1/20... Training Step: 406... Training loss: 2.2836... 0.0738 sec/batch
Epoch: 1/20... Training Step: 407... Training loss: 2.2666... 0.0792 sec/batch
Epoch: 1/20... Training Step: 408... Training loss: 2.3726... 0.0744 sec/batch
Epoch: 1/20... Training Step: 409... Training loss: 2.2948... 0.0785 sec/batch
Epoch: 1/20... Training Step: 410... Training loss: 2.3183... 0.0779 sec/batch
Epoch: 1/20... Training Step: 411... Training loss: 2.2771... 0.0742 sec/batch
Epoch: 1/20... Training Step: 412... Training loss: 2.2753... 0.0739 sec/batch
Epoch: 1/20... Training Step: 413... Training loss: 2.2838... 0.0869 sec/batch
Epoch: 1/20... Training Step: 414... Training loss: 2.2896... 0.0761 sec/batch
Epoch: 1/20... Training Step: 415... Training loss: 2.2209... 0.0748 sec/batch
Epoch: 1/20... Training Step: 416... Training loss: 2.3424... 0.0756 sec/batch
Epoch: 1/20... Training Step: 417... Training loss: 2.3230... 0.0778 sec/batch
Epoch: 1/20... Training Step: 418... Training loss: 2.2861... 0.0793 sec/batch
Epoch: 1/20... Training Step: 419... Training loss: 2.2860... 0.0842 sec/batch
Epoch: 1/20... Training Step: 420... Training loss: 2.3546... 0.0774 sec/batch
Epoch: 1/20... Training Step: 421... Training loss: 2.3141... 0.0764 sec/batch
Epoch: 1/20... Training Step: 422... Training loss: 2.2931... 0.0778 sec/batch
Epoch: 1/20... Training Step: 423... Training loss: 2.2723... 0.0754 sec/batch
Epoch: 1/20... Training Step: 424... Training loss: 2.2797... 0.0741 sec/batch
Epoch: 1/20... Training Step: 425... Training loss: 2.3314... 0.0764 sec/batch
Epoch: 1/20... Training Step: 426... Training loss: 2.3009... 0.0770 sec/batch
Epoch: 1/20... Training Step: 427... Training loss: 2.3325... 0.0755 sec/batch
Epoch: 1/20... Training Step: 428... Training loss: 2.3326... 0.0745 sec/batch
Epoch: 1/20... Training Step: 429... Training loss: 2.2661... 0.0756 sec/batch
Epoch: 1/20... Training Step: 430... Training loss: 2.3291... 0.0740 sec/batch
Epoch: 1/20... Training Step: 431... Training loss: 2.3305... 0.0781 sec/batch
Epoch: 1/20... Training Step: 432... Training loss: 2.3374... 0.0742 sec/batch
Epoch: 1/20... Training Step: 433... Training loss: 2.3647... 0.0753 sec/batch
Epoch: 1/20... Training Step: 434... Training loss: 2.3425... 0.0768 sec/batch
Epoch: 1/20... Training Step: 435... Training loss: 2.3061... 0.0787 sec/batch
Epoch: 1/20... Training Step: 436... Training loss: 2.2994... 0.0755 sec/batch
Epoch: 1/20... Training Step: 437... Training loss: 2.3205... 0.0760 sec/batch
Epoch: 1/20... Training Step: 438... Training loss: 2.3159... 0.0738 sec/batch
Epoch: 1/20... Training Step: 439... Training loss: 2.2647... 0.0743 sec/batch
Epoch: 1/20... Training Step: 440... Training loss: 2.2748... 0.0785 sec/batch
Epoch: 1/20... Training Step: 441... Training loss: 2.2742... 0.0759 sec/batch
Epoch: 1/20... Training Step: 442... Training loss: 2.2747... 0.0745 sec/batch
Epoch: 1/20... Training Step: 443... Training loss: 2.2952... 0.0762 sec/batch
Epoch: 1/20... Training Step: 444... Training loss: 2.2787... 0.0748 sec/batch
Epoch: 1/20... Training Step: 445... Training loss: 2.2380... 0.0763 sec/batch
Epoch: 1/20... Training Step: 446... Training loss: 2.2505... 0.0749 sec/batch
Epoch: 1/20... Training Step: 447... Training loss: 2.2529... 0.0762 sec/batch
Epoch: 1/20... Training Step: 448... Training loss: 2.2017... 0.0753 sec/batch
Epoch: 1/20... Training Step: 449... Training loss: 2.2524... 0.0766 sec/batch
Epoch: 1/20... Training Step: 450... Training loss: 2.3611... 0.0797 sec/batch
Epoch: 1/20... Training Step: 451... Training loss: 2.2927... 0.0751 sec/batch
Epoch: 1/20... Training Step: 452... Training loss: 2.2752... 0.0756 sec/batch
Epoch: 1/20... Training Step: 453... Training loss: 2.2499... 0.0748 sec/batch
Epoch: 1/20... Training Step: 454... Training loss: 2.2957... 0.0758 sec/batch
Epoch: 1/20... Training Step: 455... Training loss: 2.3007... 0.0767 sec/batch
Epoch: 1/20... Training Step: 456... Training loss: 2.2925... 0.0763 sec/batch
Epoch: 1/20... Training Step: 457... Training loss: 2.1598... 0.0751 sec/batch
Epoch: 1/20... Training Step: 458... Training loss: 2.2064... 0.0757 sec/batch
Epoch: 1/20... Training Step: 459... Training loss: 2.3031... 0.0764 sec/batch
Epoch: 1/20... Training Step: 460... Training loss: 2.2967... 0.0755 sec/batch
Epoch: 1/20... Training Step: 461... Training loss: 2.2960... 0.0772 sec/batch
Epoch: 1/20... Training Step: 462... Training loss: 2.1797... 0.0756 sec/batch
Epoch: 1/20... Training Step: 463... Training loss: 2.1976... 0.0754 sec/batch
Epoch: 1/20... Training Step: 464... Training loss: 2.1855... 0.0776 sec/batch
Epoch: 1/20... Training Step: 465... Training loss: 2.1922... 0.0742 sec/batch
Epoch: 1/20... Training Step: 466... Training loss: 2.2258... 0.0742 sec/batch
Epoch: 1/20... Training Step: 467... Training loss: 2.1698... 0.0765 sec/batch
Epoch: 1/20... Training Step: 468... Training loss: 2.2642... 0.0780 sec/batch
Epoch: 1/20... Training Step: 469... Training loss: 2.2435... 0.0748 sec/batch
Epoch: 1/20... Training Step: 470... Training loss: 2.2746... 0.0753 sec/batch
Epoch: 1/20... Training Step: 471... Training loss: 2.2084... 0.0767 sec/batch
Epoch: 1/20... Training Step: 472... Training loss: 2.2270... 0.0816 sec/batch
Epoch: 1/20... Training Step: 473... Training loss: 2.1842... 0.0744 sec/batch
Epoch: 1/20... Training Step: 474... Training loss: 2.2314... 0.0757 sec/batch
Epoch: 1/20... Training Step: 475... Training loss: 2.1831... 0.0760 sec/batch
Epoch: 1/20... Training Step: 476... Training loss: 2.2104... 0.0793 sec/batch
Epoch: 1/20... Training Step: 477... Training loss: 2.1940... 0.0788 sec/batch
Epoch: 1/20... Training Step: 478... Training loss: 2.2176... 0.0829 sec/batch
Epoch: 1/20... Training Step: 479... Training loss: 2.1931... 0.0780 sec/batch
Epoch: 1/20... Training Step: 480... Training loss: 2.1925... 0.0799 sec/batch
Epoch: 1/20... Training Step: 481... Training loss: 2.1704... 0.0755 sec/batch
Epoch: 1/20... Training Step: 482... Training loss: 2.2491... 0.0748 sec/batch
Epoch: 1/20... Training Step: 483... Training loss: 2.2131... 0.0749 sec/batch
Epoch: 1/20... Training Step: 484... Training loss: 2.2272... 0.0780 sec/batch
Epoch: 1/20... Training Step: 485... Training loss: 2.2088... 0.0760 sec/batch
Epoch: 1/20... Training Step: 486... Training loss: 2.2325... 0.0781 sec/batch
Epoch: 1/20... Training Step: 487... Training loss: 2.2455... 0.0768 sec/batch
Epoch: 1/20... Training Step: 488... Training loss: 2.2416... 0.0747 sec/batch
Epoch: 1/20... Training Step: 489... Training loss: 2.2370... 0.0750 sec/batch
Epoch: 1/20... Training Step: 490... Training loss: 2.2675... 0.0753 sec/batch
Epoch: 1/20... Training Step: 491... Training loss: 2.2055... 0.0759 sec/batch
Epoch: 1/20... Training Step: 492... Training loss: 2.1677... 0.0753 sec/batch
Epoch: 1/20... Training Step: 493... Training loss: 2.2705... 0.0739 sec/batch
Epoch: 1/20... Training Step: 494... Training loss: 2.2302... 0.0760 sec/batch
Epoch: 1/20... Training Step: 495... Training loss: 2.1827... 0.0737 sec/batch
Epoch: 1/20... Training Step: 496... Training loss: 2.2133... 0.0746 sec/batch
Epoch: 1/20... Training Step: 497... Training loss: 2.2597... 0.0798 sec/batch
Epoch: 1/20... Training Step: 498... Training loss: 2.2187... 0.0741 sec/batch
Epoch: 1/20... Training Step: 499... Training loss: 2.2815... 0.0738 sec/batch
Epoch: 1/20... Training Step: 500... Training loss: 2.2498... 0.0799 sec/batch
Epoch: 1/20... Training Step: 501... Training loss: 2.2264... 0.0756 sec/batch
Epoch: 1/20... Training Step: 502... Training loss: 2.2220... 0.0803 sec/batch
Epoch: 1/20... Training Step: 503... Training loss: 2.2269... 0.0789 sec/batch
Epoch: 1/20... Training Step: 504... Training loss: 2.2196... 0.0755 sec/batch
Epoch: 1/20... Training Step: 505... Training loss: 2.2495... 0.0739 sec/batch
Epoch: 1/20... Training Step: 506... Training loss: 2.1906... 0.0774 sec/batch
Epoch: 1/20... Training Step: 507... Training loss: 2.2402... 0.0745 sec/batch
Epoch: 1/20... Training Step: 508... Training loss: 2.2586... 0.0752 sec/batch
Epoch: 1/20... Training Step: 509... Training loss: 2.2264... 0.0777 sec/batch
Epoch: 1/20... Training Step: 510... Training loss: 2.2051... 0.0774 sec/batch
Epoch: 1/20... Training Step: 511... Training loss: 2.2599... 0.0738 sec/batch
Epoch: 1/20... Training Step: 512... Training loss: 2.2659... 0.0759 sec/batch
Epoch: 1/20... Training Step: 513... Training loss: 2.2813... 0.0744 sec/batch
Epoch: 1/20... Training Step: 514... Training loss: 2.2793... 0.0754 sec/batch
Epoch: 1/20... Training Step: 515... Training loss: 2.2591... 0.0751 sec/batch
Epoch: 1/20... Training Step: 516... Training loss: 2.1409... 0.0733 sec/batch
Epoch: 1/20... Training Step: 517... Training loss: 2.1803... 0.0766 sec/batch
Epoch: 1/20... Training Step: 518... Training loss: 2.1749... 0.0788 sec/batch
Epoch: 1/20... Training Step: 519... Training loss: 2.2479... 0.0787 sec/batch
Epoch: 1/20... Training Step: 520... Training loss: 2.1998... 0.0811 sec/batch
Epoch: 1/20... Training Step: 521... Training loss: 2.2159... 0.0749 sec/batch
Epoch: 1/20... Training Step: 522... Training loss: 2.2007... 0.0781 sec/batch
Epoch: 1/20... Training Step: 523... Training loss: 2.1915... 0.0754 sec/batch
Epoch: 1/20... Training Step: 524... Training loss: 2.1693... 0.0780 sec/batch
Epoch: 1/20... Training Step: 525... Training loss: 2.1902... 0.0777 sec/batch
Epoch: 1/20... Training Step: 526... Training loss: 2.2083... 0.0791 sec/batch
Epoch: 1/20... Training Step: 527... Training loss: 2.2103... 0.0781 sec/batch
Epoch: 1/20... Training Step: 528... Training loss: 2.1098... 0.0786 sec/batch
Epoch: 1/20... Training Step: 529... Training loss: 2.1556... 0.0772 sec/batch
Epoch: 1/20... Training Step: 530... Training loss: 2.2228... 0.0766 sec/batch
Epoch: 1/20... Training Step: 531... Training loss: 2.2694... 0.0795 sec/batch
Epoch: 1/20... Training Step: 532... Training loss: 2.2060... 0.0809 sec/batch
Epoch: 1/20... Training Step: 533... Training loss: 2.2355... 0.0801 sec/batch
Epoch: 1/20... Training Step: 534... Training loss: 2.2055... 0.0817 sec/batch
Epoch: 1/20... Training Step: 535... Training loss: 2.2157... 0.0788 sec/batch
Epoch: 1/20... Training Step: 536... Training loss: 2.2032... 0.0822 sec/batch
Epoch: 1/20... Training Step: 537... Training loss: 2.2280... 0.0783 sec/batch
Epoch: 1/20... Training Step: 538... Training loss: 2.1617... 0.0788 sec/batch
Epoch: 1/20... Training Step: 539... Training loss: 2.1314... 0.0816 sec/batch
Epoch: 1/20... Training Step: 540... Training loss: 2.1637... 0.0807 sec/batch
Epoch: 1/20... Training Step: 541... Training loss: 2.2897... 0.0794 sec/batch
Epoch: 1/20... Training Step: 542... Training loss: 2.1499... 0.0773 sec/batch
Epoch: 1/20... Training Step: 543... Training loss: 2.1369... 0.0758 sec/batch
Epoch: 1/20... Training Step: 544... Training loss: 2.1711... 0.0741 sec/batch
Epoch: 1/20... Training Step: 545... Training loss: 2.1987... 0.0771 sec/batch
Epoch: 1/20... Training Step: 546... Training loss: 2.1555... 0.0768 sec/batch
Epoch: 1/20... Training Step: 547... Training loss: 2.2058... 0.0775 sec/batch
Epoch: 1/20... Training Step: 548... Training loss: 2.2545... 0.0841 sec/batch
Epoch: 1/20... Training Step: 549... Training loss: 2.1462... 0.0789 sec/batch
Epoch: 1/20... Training Step: 550... Training loss: 2.1515... 0.0832 sec/batch
Epoch: 1/20... Training Step: 551... Training loss: 2.1941... 0.0780 sec/batch
Epoch: 1/20... Training Step: 552... Training loss: 2.1560... 0.0739 sec/batch
Epoch: 1/20... Training Step: 553... Training loss: 2.1365... 0.0797 sec/batch
Epoch: 1/20... Training Step: 554... Training loss: 2.1899... 0.0785 sec/batch
Epoch: 1/20... Training Step: 555... Training loss: 2.1782... 0.0771 sec/batch
Epoch: 1/20... Training Step: 556... Training loss: 2.1369... 0.0781 sec/batch
Epoch: 1/20... Training Step: 557... Training loss: 2.1501... 0.0794 sec/batch
Epoch: 1/20... Training Step: 558... Training loss: 2.2195... 0.0775 sec/batch
Epoch: 1/20... Training Step: 559... Training loss: 2.2010... 0.0762 sec/batch
Epoch: 1/20... Training Step: 560... Training loss: 2.1477... 0.0761 sec/batch
Epoch: 1/20... Training Step: 561... Training loss: 2.1983... 0.0741 sec/batch
Epoch: 1/20... Training Step: 562... Training loss: 2.1780... 0.0742 sec/batch
Epoch: 1/20... Training Step: 563... Training loss: 2.1334... 0.0756 sec/batch
Epoch: 1/20... Training Step: 564... Training loss: 2.1313... 0.0757 sec/batch
Epoch: 1/20... Training Step: 565... Training loss: 2.1511... 0.0754 sec/batch
Epoch: 1/20... Training Step: 566... Training loss: 2.1901... 0.0769 sec/batch
Epoch: 1/20... Training Step: 567... Training loss: 2.1859... 0.0744 sec/batch
Epoch: 1/20... Training Step: 568... Training loss: 2.1466... 0.0757 sec/batch
Epoch: 1/20... Training Step: 569... Training loss: 2.1540... 0.0759 sec/batch
Epoch: 1/20... Training Step: 570... Training loss: 2.1374... 0.0772 sec/batch
Epoch: 1/20... Training Step: 571... Training loss: 2.1523... 0.0737 sec/batch
Epoch: 1/20... Training Step: 572... Training loss: 2.1097... 0.0749 sec/batch
Epoch: 1/20... Training Step: 573... Training loss: 2.1939... 0.0744 sec/batch
Epoch: 1/20... Training Step: 574... Training loss: 2.1882... 0.0759 sec/batch
Epoch: 1/20... Training Step: 575... Training loss: 2.1652... 0.0766 sec/batch
Epoch: 1/20... Training Step: 576... Training loss: 2.1671... 0.0766 sec/batch
Epoch: 1/20... Training Step: 577... Training loss: 2.1701... 0.0755 sec/batch
Epoch: 1/20... Training Step: 578... Training loss: 2.1933... 0.0780 sec/batch
Epoch: 1/20... Training Step: 579... Training loss: 2.1923... 0.0776 sec/batch
Epoch: 1/20... Training Step: 580... Training loss: 2.1679... 0.0751 sec/batch
Epoch: 1/20... Training Step: 581... Training loss: 2.1377... 0.0783 sec/batch
Epoch: 1/20... Training Step: 582... Training loss: 2.1536... 0.0761 sec/batch
Epoch: 1/20... Training Step: 583... Training loss: 2.2172... 0.0796 sec/batch
Epoch: 1/20... Training Step: 584... Training loss: 2.0968... 0.0805 sec/batch
Epoch: 1/20... Training Step: 585... Training loss: 2.1432... 0.0789 sec/batch
Epoch: 1/20... Training Step: 586... Training loss: 2.1851... 0.0765 sec/batch
Epoch: 1/20... Training Step: 587... Training loss: 2.1613... 0.0759 sec/batch
Epoch: 1/20... Training Step: 588... Training loss: 2.1792... 0.0783 sec/batch
Epoch: 1/20... Training Step: 589... Training loss: 2.1274... 0.0788 sec/batch
Epoch: 1/20... Training Step: 590... Training loss: 2.1695... 0.0792 sec/batch
Epoch: 1/20... Training Step: 591... Training loss: 2.1492... 0.0754 sec/batch
Epoch: 1/20... Training Step: 592... Training loss: 2.2622... 0.0763 sec/batch
Epoch: 1/20... Training Step: 593... Training loss: 2.1441... 0.0765 sec/batch
Epoch: 1/20... Training Step: 594... Training loss: 2.1363... 0.0748 sec/batch
Epoch: 1/20... Training Step: 595... Training loss: 2.1360... 0.0757 sec/batch
Epoch: 1/20... Training Step: 596... Training loss: 2.1620... 0.0765 sec/batch
Epoch: 1/20... Training Step: 597... Training loss: 2.0977... 0.0762 sec/batch
Epoch: 1/20... Training Step: 598... Training loss: 2.0791... 0.0760 sec/batch
Epoch: 1/20... Training Step: 599... Training loss: 2.1105... 0.0776 sec/batch
Epoch: 1/20... Training Step: 600... Training loss: 2.1540... 0.0757 sec/batch
Epoch: 1/20... Training Step: 601... Training loss: 2.1934... 0.0731 sec/batch
Epoch: 1/20... Training Step: 602... Training loss: 2.1039... 0.0735 sec/batch
Epoch: 1/20... Training Step: 603... Training loss: 2.1021... 0.0740 sec/batch
Epoch: 1/20... Training Step: 604... Training loss: 2.1376... 0.0748 sec/batch
Epoch: 1/20... Training Step: 605... Training loss: 2.1017... 0.0736 sec/batch
Epoch: 1/20... Training Step: 606... Training loss: 2.1670... 0.0767 sec/batch
Epoch: 1/20... Training Step: 607... Training loss: 2.1981... 0.0780 sec/batch
Epoch: 1/20... Training Step: 608... Training loss: 2.1911... 0.0764 sec/batch
Epoch: 1/20... Training Step: 609... Training loss: 2.1578... 0.0747 sec/batch
Epoch: 1/20... Training Step: 610... Training loss: 2.1801... 0.0785 sec/batch
Epoch: 1/20... Training Step: 611... Training loss: 2.1156... 0.0737 sec/batch
Epoch: 1/20... Training Step: 612... Training loss: 2.1378... 0.0737 sec/batch
Epoch: 1/20... Training Step: 613... Training loss: 2.1921... 0.0767 sec/batch
Epoch: 1/20... Training Step: 614... Training loss: 2.1273... 0.0748 sec/batch
Epoch: 1/20... Training Step: 615... Training loss: 2.1593... 0.0736 sec/batch
Epoch: 1/20... Training Step: 616... Training loss: 2.1459... 0.0778 sec/batch
Epoch: 1/20... Training Step: 617... Training loss: 2.2269... 0.0766 sec/batch
Epoch: 1/20... Training Step: 618... Training loss: 2.1305... 0.0735 sec/batch
Epoch: 1/20... Training Step: 619... Training loss: 2.0680... 0.0761 sec/batch
Epoch: 1/20... Training Step: 620... Training loss: 2.1143... 0.0737 sec/batch
Epoch: 1/20... Training Step: 621... Training loss: 2.1180... 0.0740 sec/batch
Epoch: 1/20... Training Step: 622... Training loss: 2.0413... 0.0764 sec/batch
Epoch: 1/20... Training Step: 623... Training loss: 2.1553... 0.0822 sec/batch
Epoch: 1/20... Training Step: 624... Training loss: 2.1114... 0.0751 sec/batch
Epoch: 1/20... Training Step: 625... Training loss: 2.1307... 0.0769 sec/batch
Epoch: 1/20... Training Step: 626... Training loss: 2.0892... 0.0767 sec/batch
Epoch: 1/20... Training Step: 627... Training loss: 2.1229... 0.0736 sec/batch
Epoch: 1/20... Training Step: 628... Training loss: 2.1450... 0.0758 sec/batch
Epoch: 1/20... Training Step: 629... Training loss: 1.9890... 0.0771 sec/batch
Epoch: 1/20... Training Step: 630... Training loss: 2.1203... 0.0739 sec/batch
Epoch: 1/20... Training Step: 631... Training loss: 2.0523... 0.0760 sec/batch
Epoch: 1/20... Training Step: 632... Training loss: 2.1055... 0.0735 sec/batch
Epoch: 1/20... Training Step: 633... Training loss: 2.0315... 0.0734 sec/batch
Epoch: 1/20... Training Step: 634... Training loss: 2.1144... 0.0758 sec/batch
Epoch: 1/20... Training Step: 635... Training loss: 2.1106... 0.0773 sec/batch
Epoch: 1/20... Training Step: 636... Training loss: 2.1014... 0.0737 sec/batch
Epoch: 1/20... Training Step: 637... Training loss: 2.1234... 0.0758 sec/batch
Epoch: 1/20... Training Step: 638... Training loss: 2.0769... 0.0739 sec/batch
Epoch: 1/20... Training Step: 639... Training loss: 2.1325... 0.0751 sec/batch
Epoch: 1/20... Training Step: 640... Training loss: 2.1480... 0.0769 sec/batch
Epoch: 1/20... Training Step: 641... Training loss: 2.1461... 0.0743 sec/batch
Epoch: 1/20... Training Step: 642... Training loss: 2.1135... 0.0735 sec/batch
Epoch: 1/20... Training Step: 643... Training loss: 2.0762... 0.0780 sec/batch
Epoch: 1/20... Training Step: 644... Training loss: 2.1318... 0.0737 sec/batch
Epoch: 1/20... Training Step: 645... Training loss: 2.0873... 0.0736 sec/batch
Epoch: 1/20... Training Step: 646... Training loss: 2.1960... 0.0769 sec/batch
Epoch: 1/20... Training Step: 647... Training loss: 2.1269... 0.0739 sec/batch
Epoch: 1/20... Training Step: 648... Training loss: 2.0380... 0.0734 sec/batch
Epoch: 1/20... Training Step: 649... Training loss: 2.1354... 0.0793 sec/batch
Epoch: 1/20... Training Step: 650... Training loss: 2.0516... 0.0762 sec/batch
Epoch: 1/20... Training Step: 651... Training loss: 2.0864... 0.0749 sec/batch
Epoch: 1/20... Training Step: 652... Training loss: 2.1502... 0.0741 sec/batch
Epoch: 1/20... Training Step: 653... Training loss: 2.1335... 0.0735 sec/batch
Epoch: 1/20... Training Step: 654... Training loss: 2.1202... 0.0767 sec/batch
Epoch: 1/20... Training Step: 655... Training loss: 2.1512... 0.0755 sec/batch
Epoch: 1/20... Training Step: 656... Training loss: 2.1773... 0.0734 sec/batch
Epoch: 1/20... Training Step: 657... Training loss: 2.0657... 0.0744 sec/batch
Epoch: 1/20... Training Step: 658... Training loss: 2.0748... 0.0755 sec/batch
Epoch: 1/20... Training Step: 659... Training loss: 2.1569... 0.0745 sec/batch
Epoch: 1/20... Training Step: 660... Training loss: 2.1630... 0.0750 sec/batch
Epoch: 1/20... Training Step: 661... Training loss: 2.1046... 0.0756 sec/batch
Epoch: 1/20... Training Step: 662... Training loss: 2.1301... 0.0739 sec/batch
Epoch: 1/20... Training Step: 663... Training loss: 2.0785... 0.0766 sec/batch
Epoch: 1/20... Training Step: 664... Training loss: 2.2097... 0.0754 sec/batch
Epoch: 1/20... Training Step: 665... Training loss: 2.1224... 0.0758 sec/batch
Epoch: 1/20... Training Step: 666... Training loss: 2.0563... 0.0735 sec/batch
Epoch: 1/20... Training Step: 667... Training loss: 2.0952... 0.0756 sec/batch
Epoch: 1/20... Training Step: 668... Training loss: 2.1335... 0.0740 sec/batch
Epoch: 1/20... Training Step: 669... Training loss: 2.1063... 0.0770 sec/batch
Epoch: 1/20... Training Step: 670... Training loss: 2.0451... 0.0759 sec/batch
Epoch: 1/20... Training Step: 671... Training loss: 2.0948... 0.0736 sec/batch
Epoch: 1/20... Training Step: 672... Training loss: 2.0481... 0.0736 sec/batch
Epoch: 1/20... Training Step: 673... Training loss: 2.1051... 0.0773 sec/batch
Epoch: 1/20... Training Step: 674... Training loss: 2.0977... 0.0759 sec/batch
Epoch: 1/20... Training Step: 675... Training loss: 2.0639... 0.0755 sec/batch
Epoch: 1/20... Training Step: 676... Training loss: 2.0842... 0.0783 sec/batch
Epoch: 1/20... Training Step: 677... Training loss: 2.1224... 0.0759 sec/batch
Epoch: 1/20... Training Step: 678... Training loss: 2.0547... 0.0745 sec/batch
Epoch: 1/20... Training Step: 679... Training loss: 2.0353... 0.0767 sec/batch
Epoch: 1/20... Training Step: 680... Training loss: 2.0690... 0.0759 sec/batch
Epoch: 1/20... Training Step: 681... Training loss: 2.0809... 0.0757 sec/batch
Epoch: 1/20... Training Step: 682... Training loss: 2.0254... 0.0767 sec/batch
Epoch: 1/20... Training Step: 683... Training loss: 2.0682... 0.0731 sec/batch
Epoch: 1/20... Training Step: 684... Training loss: 2.0853... 0.0735 sec/batch
Epoch: 1/20... Training Step: 685... Training loss: 2.1447... 0.0760 sec/batch
Epoch: 1/20... Training Step: 686... Training loss: 2.1098... 0.0757 sec/batch
Epoch: 1/20... Training Step: 687... Training loss: 2.0677... 0.0739 sec/batch
Epoch: 1/20... Training Step: 688... Training loss: 2.0798... 0.0781 sec/batch
Epoch: 1/20... Training Step: 689... Training loss: 2.0551... 0.0751 sec/batch
Epoch: 1/20... Training Step: 690... Training loss: 2.1181... 0.0739 sec/batch
Epoch: 1/20... Training Step: 691... Training loss: 2.0588... 0.0752 sec/batch
Epoch: 1/20... Training Step: 692... Training loss: 2.0846... 0.0735 sec/batch
Epoch: 1/20... Training Step: 693... Training loss: 2.0070... 0.0765 sec/batch
Epoch: 1/20... Training Step: 694... Training loss: 2.0457... 0.0776 sec/batch
Epoch: 1/20... Training Step: 695... Training loss: 2.0890... 0.0752 sec/batch
Epoch: 1/20... Training Step: 696... Training loss: 2.0373... 0.0749 sec/batch
Epoch: 1/20... Training Step: 697... Training loss: 2.0821... 0.0823 sec/batch
Epoch: 1/20... Training Step: 698... Training loss: 2.1026... 0.0736 sec/batch
Epoch: 1/20... Training Step: 699... Training loss: 2.0308... 0.0754 sec/batch
Epoch: 1/20... Training Step: 700... Training loss: 2.0671... 0.0747 sec/batch
Epoch: 1/20... Training Step: 701... Training loss: 2.0516... 0.0749 sec/batch
Epoch: 1/20... Training Step: 702... Training loss: 2.0237... 0.0796 sec/batch
Epoch: 1/20... Training Step: 703... Training loss: 2.0843... 0.0755 sec/batch
Epoch: 1/20... Training Step: 704... Training loss: 1.9991... 0.0738 sec/batch
Epoch: 1/20... Training Step: 705... Training loss: 1.9784... 0.0741 sec/batch
Epoch: 1/20... Training Step: 706... Training loss: 2.0331... 0.0756 sec/batch
Epoch: 1/20... Training Step: 707... Training loss: 2.0806... 0.0750 sec/batch
Epoch: 1/20... Training Step: 708... Training loss: 2.0815... 0.0740 sec/batch
Epoch: 1/20... Training Step: 709... Training loss: 2.1279... 0.0752 sec/batch
Epoch: 1/20... Training Step: 710... Training loss: 2.0589... 0.0768 sec/batch
Epoch: 1/20... Training Step: 711... Training loss: 2.0539... 0.0758 sec/batch
Epoch: 1/20... Training Step: 712... Training loss: 2.1009... 0.0751 sec/batch
Epoch: 1/20... Training Step: 713... Training loss: 2.0295... 0.0776 sec/batch
Epoch: 1/20... Training Step: 714... Training loss: 2.0243... 0.0736 sec/batch
Epoch: 1/20... Training Step: 715... Training loss: 2.0546... 0.0842 sec/batch
Epoch: 1/20... Training Step: 716... Training loss: 2.0605... 0.0754 sec/batch
Epoch: 1/20... Training Step: 717... Training loss: 2.0805... 0.0767 sec/batch
Epoch: 1/20... Training Step: 718... Training loss: 2.0619... 0.0762 sec/batch
Epoch: 1/20... Training Step: 719... Training loss: 2.0705... 0.0747 sec/batch
Epoch: 1/20... Training Step: 720... Training loss: 2.1191... 0.0756 sec/batch
Epoch: 1/20... Training Step: 721... Training loss: 2.0445... 0.0756 sec/batch
Epoch: 1/20... Training Step: 722... Training loss: 2.0145... 0.0733 sec/batch
Epoch: 1/20... Training Step: 723... Training loss: 2.0599... 0.0733 sec/batch
Epoch: 1/20... Training Step: 724... Training loss: 2.0601... 0.0772 sec/batch
Epoch: 1/20... Training Step: 725... Training loss: 2.1377... 0.0742 sec/batch
Epoch: 1/20... Training Step: 726... Training loss: 2.2144... 0.0737 sec/batch
Epoch: 1/20... Training Step: 727... Training loss: 2.0575... 0.0828 sec/batch
Epoch: 1/20... Training Step: 728... Training loss: 2.0625... 0.0737 sec/batch
Epoch: 1/20... Training Step: 729... Training loss: 2.0770... 0.0748 sec/batch
Epoch: 1/20... Training Step: 730... Training loss: 2.1099... 0.0764 sec/batch
Epoch: 1/20... Training Step: 731... Training loss: 2.0456... 0.0774 sec/batch
Epoch: 1/20... Training Step: 732... Training loss: 2.0392... 0.0768 sec/batch
Epoch: 1/20... Training Step: 733... Training loss: 1.9797... 0.0767 sec/batch
Epoch: 1/20... Training Step: 734... Training loss: 2.1320... 0.0765 sec/batch
Epoch: 1/20... Training Step: 735... Training loss: 2.0079... 0.0734 sec/batch
Epoch: 1/20... Training Step: 736... Training loss: 2.0837... 0.0760 sec/batch
Epoch: 1/20... Training Step: 737... Training loss: 1.9680... 0.0767 sec/batch
Epoch: 1/20... Training Step: 738... Training loss: 2.0256... 0.0749 sec/batch
Epoch: 1/20... Training Step: 739... Training loss: 2.0347... 0.0756 sec/batch
Epoch: 1/20... Training Step: 740... Training loss: 2.0374... 0.0758 sec/batch
Epoch: 1/20... Training Step: 741... Training loss: 2.0715... 0.0757 sec/batch
Epoch: 1/20... Training Step: 742... Training loss: 2.0284... 0.0747 sec/batch
Epoch: 1/20... Training Step: 743... Training loss: 1.9850... 0.0754 sec/batch
Epoch: 1/20... Training Step: 744... Training loss: 2.0575... 0.0757 sec/batch
Epoch: 1/20... Training Step: 745... Training loss: 2.0335... 0.0825 sec/batch
Epoch: 1/20... Training Step: 746... Training loss: 2.0280... 0.0757 sec/batch
Epoch: 1/20... Training Step: 747... Training loss: 2.1032... 0.0737 sec/batch
Epoch: 1/20... Training Step: 748... Training loss: 2.0888... 0.0772 sec/batch
Epoch: 1/20... Training Step: 749... Training loss: 2.0426... 0.0746 sec/batch
Epoch: 1/20... Training Step: 750... Training loss: 1.9827... 0.0737 sec/batch
Epoch: 1/20... Training Step: 751... Training loss: 2.0977... 0.0757 sec/batch
Epoch: 1/20... Training Step: 752... Training loss: 1.9710... 0.0767 sec/batch
Epoch: 1/20... Training Step: 753... Training loss: 2.0049... 0.0748 sec/batch
Epoch: 1/20... Training Step: 754... Training loss: 2.0326... 0.0760 sec/batch
Epoch: 1/20... Training Step: 755... Training loss: 2.0037... 0.0798 sec/batch
Epoch: 1/20... Training Step: 756... Training loss: 2.0210... 0.0750 sec/batch
Epoch: 1/20... Training Step: 757... Training loss: 2.0478... 0.0756 sec/batch
Epoch: 1/20... Training Step: 758... Training loss: 2.1301... 0.0746 sec/batch
Epoch: 1/20... Training Step: 759... Training loss: 2.0519... 0.0747 sec/batch
Epoch: 1/20... Training Step: 760... Training loss: 2.1309... 0.0771 sec/batch
Epoch: 1/20... Training Step: 761... Training loss: 2.0595... 0.0753 sec/batch
Epoch: 1/20... Training Step: 762... Training loss: 1.9760... 0.0736 sec/batch
Epoch: 1/20... Training Step: 763... Training loss: 2.0487... 0.0763 sec/batch
Epoch: 1/20... Training Step: 764... Training loss: 2.0112... 0.0743 sec/batch
Epoch: 1/20... Training Step: 765... Training loss: 2.0214... 0.0741 sec/batch
Epoch: 1/20... Training Step: 766... Training loss: 2.0280... 0.0774 sec/batch
Epoch: 1/20... Training Step: 767... Training loss: 2.0393... 0.0760 sec/batch
Epoch: 1/20... Training Step: 768... Training loss: 2.0008... 0.0739 sec/batch
Epoch: 1/20... Training Step: 769... Training loss: 1.9864... 0.0770 sec/batch
Epoch: 1/20... Training Step: 770... Training loss: 2.0295... 0.0743 sec/batch
Epoch: 1/20... Training Step: 771... Training loss: 2.0476... 0.0759 sec/batch
Epoch: 1/20... Training Step: 772... Training loss: 2.1509... 0.0785 sec/batch
Epoch: 1/20... Training Step: 773... Training loss: 2.0901... 0.0737 sec/batch
Epoch: 1/20... Training Step: 774... Training loss: 2.0811... 0.0740 sec/batch
Epoch: 1/20... Training Step: 775... Training loss: 2.0519... 0.0757 sec/batch
Epoch: 1/20... Training Step: 776... Training loss: 1.9695... 0.0751 sec/batch
Epoch: 1/20... Training Step: 777... Training loss: 2.0062... 0.0743 sec/batch
Epoch: 1/20... Training Step: 778... Training loss: 1.9746... 0.0759 sec/batch
Epoch: 1/20... Training Step: 779... Training loss: 1.9965... 0.0755 sec/batch
Epoch: 1/20... Training Step: 780... Training loss: 2.0306... 0.0737 sec/batch
Epoch: 1/20... Training Step: 781... Training loss: 2.0253... 0.0844 sec/batch
Epoch: 1/20... Training Step: 782... Training loss: 2.0129... 0.0738 sec/batch
Epoch: 1/20... Training Step: 783... Training loss: 1.9961... 0.0743 sec/batch
Epoch: 1/20... Training Step: 784... Training loss: 2.0271... 0.0771 sec/batch
Epoch: 1/20... Training Step: 785... Training loss: 2.0544... 0.0749 sec/batch
Epoch: 1/20... Training Step: 786... Training loss: 2.0104... 0.0768 sec/batch
Epoch: 1/20... Training Step: 787... Training loss: 2.0498... 0.0742 sec/batch
Epoch: 1/20... Training Step: 788... Training loss: 2.0316... 0.0756 sec/batch
Epoch: 1/20... Training Step: 789... Training loss: 2.0153... 0.0735 sec/batch
Epoch: 1/20... Training Step: 790... Training loss: 2.0202... 0.0754 sec/batch
Epoch: 1/20... Training Step: 791... Training loss: 2.0204... 0.0733 sec/batch
Epoch: 1/20... Training Step: 792... Training loss: 1.9402... 0.0739 sec/batch
Epoch: 1/20... Training Step: 793... Training loss: 2.0015... 0.0765 sec/batch
Epoch: 1/20... Training Step: 794... Training loss: 2.0244... 0.0759 sec/batch
Epoch: 1/20... Training Step: 795... Training loss: 1.9354... 0.0739 sec/batch
Epoch: 1/20... Training Step: 796... Training loss: 1.9483... 0.0762 sec/batch
Epoch: 1/20... Training Step: 797... Training loss: 1.9883... 0.0743 sec/batch
Epoch: 1/20... Training Step: 798... Training loss: 2.0147... 0.0739 sec/batch
Epoch: 1/20... Training Step: 799... Training loss: 2.0525... 0.0768 sec/batch
Epoch: 1/20... Training Step: 800... Training loss: 1.9690... 0.0825 sec/batch
Epoch: 1/20... Training Step: 801... Training loss: 2.0726... 0.0760 sec/batch
Epoch: 1/20... Training Step: 802... Training loss: 2.0433... 0.0740 sec/batch
Epoch: 1/20... Training Step: 803... Training loss: 2.0623... 0.0752 sec/batch
Epoch: 1/20... Training Step: 804... Training loss: 2.0319... 0.0770 sec/batch
Epoch: 1/20... Training Step: 805... Training loss: 1.9689... 0.0741 sec/batch
Epoch: 1/20... Training Step: 806... Training loss: 2.0231... 0.0737 sec/batch
Epoch: 1/20... Training Step: 807... Training loss: 1.9862... 0.0758 sec/batch
Epoch: 1/20... Training Step: 808... Training loss: 2.0271... 0.0768 sec/batch
Epoch: 1/20... Training Step: 809... Training loss: 2.0844... 0.0755 sec/batch
Epoch: 1/20... Training Step: 810... Training loss: 2.0764... 0.0776 sec/batch
Epoch: 1/20... Training Step: 811... Training loss: 2.0633... 0.0737 sec/batch
Epoch: 1/20... Training Step: 812... Training loss: 2.0385... 0.0734 sec/batch
Epoch: 1/20... Training Step: 813... Training loss: 2.0088... 0.0791 sec/batch
Epoch: 1/20... Training Step: 814... Training loss: 2.1199... 0.0734 sec/batch
Epoch: 1/20... Training Step: 815... Training loss: 2.0508... 0.0738 sec/batch
Epoch: 1/20... Training Step: 816... Training loss: 2.1094... 0.0831 sec/batch
Epoch: 1/20... Training Step: 817... Training loss: 2.0738... 0.0741 sec/batch
Epoch: 1/20... Training Step: 818... Training loss: 2.0823... 0.0772 sec/batch
Epoch: 1/20... Training Step: 819... Training loss: 2.0700... 0.0761 sec/batch
Epoch: 1/20... Training Step: 820... Training loss: 2.0782... 0.0737 sec/batch
Epoch: 1/20... Training Step: 821... Training loss: 2.0431... 0.0756 sec/batch
Epoch: 1/20... Training Step: 822... Training loss: 1.9195... 0.0772 sec/batch
Epoch: 1/20... Training Step: 823... Training loss: 1.9858... 0.0738 sec/batch
Epoch: 1/20... Training Step: 824... Training loss: 1.9815... 0.0746 sec/batch
Epoch: 1/20... Training Step: 825... Training loss: 1.9849... 0.0762 sec/batch
Epoch: 1/20... Training Step: 826... Training loss: 2.0041... 0.0734 sec/batch
Epoch: 1/20... Training Step: 827... Training loss: 2.0319... 0.0757 sec/batch
Epoch: 1/20... Training Step: 828... Training loss: 2.0690... 0.0753 sec/batch
Epoch: 1/20... Training Step: 829... Training loss: 2.0186... 0.0735 sec/batch
Epoch: 1/20... Training Step: 830... Training loss: 2.0222... 0.0765 sec/batch
Epoch: 1/20... Training Step: 831... Training loss: 2.0868... 0.0764 sec/batch
Epoch: 1/20... Training Step: 832... Training loss: 2.1339... 0.0751 sec/batch
Epoch: 1/20... Training Step: 833... Training loss: 2.1394... 0.0752 sec/batch
Epoch: 1/20... Training Step: 834... Training loss: 1.9928... 0.0753 sec/batch
Epoch: 1/20... Training Step: 835... Training loss: 1.9268... 0.0770 sec/batch
Epoch: 1/20... Training Step: 836... Training loss: 2.0523... 0.0744 sec/batch
Epoch: 1/20... Training Step: 837... Training loss: 2.0693... 0.0745 sec/batch
Epoch: 1/20... Training Step: 838... Training loss: 2.0446... 0.0738 sec/batch
Epoch: 1/20... Training Step: 839... Training loss: 2.1092... 0.0763 sec/batch
Epoch: 1/20... Training Step: 840... Training loss: 2.0133... 0.0754 sec/batch
Epoch: 1/20... Training Step: 841... Training loss: 2.0640... 0.0732 sec/batch
Epoch: 1/20... Training Step: 842... Training loss: 2.0832... 0.0754 sec/batch
Epoch: 1/20... Training Step: 843... Training loss: 2.0714... 0.0823 sec/batch
Epoch: 1/20... Training Step: 844... Training loss: 2.0507... 0.0753 sec/batch
Epoch: 1/20... Training Step: 845... Training loss: 2.1635... 0.0735 sec/batch
Epoch: 1/20... Training Step: 846... Training loss: 1.9431... 0.0840 sec/batch
Epoch: 1/20... Training Step: 847... Training loss: 1.9560... 0.0767 sec/batch
Epoch: 1/20... Training Step: 848... Training loss: 2.0154... 0.0741 sec/batch
Epoch: 1/20... Training Step: 849... Training loss: 1.9897... 0.0756 sec/batch
Epoch: 1/20... Training Step: 850... Training loss: 1.9980... 0.0782 sec/batch
Epoch: 1/20... Training Step: 851... Training loss: 1.9845... 0.0771 sec/batch
Epoch: 1/20... Training Step: 852... Training loss: 2.0711... 0.0745 sec/batch
Epoch: 1/20... Training Step: 853... Training loss: 2.0399... 0.0744 sec/batch
Epoch: 1/20... Training Step: 854... Training loss: 2.0065... 0.0749 sec/batch
Epoch: 1/20... Training Step: 855... Training loss: 2.0560... 0.0767 sec/batch
Epoch: 1/20... Training Step: 856... Training loss: 2.0108... 0.0738 sec/batch
Epoch: 1/20... Training Step: 857... Training loss: 1.9935... 0.0753 sec/batch
Epoch: 1/20... Training Step: 858... Training loss: 1.9790... 0.0754 sec/batch
Epoch: 1/20... Training Step: 859... Training loss: 2.0412... 0.0740 sec/batch
Epoch: 1/20... Training Step: 860... Training loss: 2.0225... 0.0743 sec/batch
Epoch: 1/20... Training Step: 861... Training loss: 2.0107... 0.0756 sec/batch
Epoch: 1/20... Training Step: 862... Training loss: 1.9680... 0.0753 sec/batch
Epoch: 1/20... Training Step: 863... Training loss: 1.9712... 0.0743 sec/batch
Epoch: 1/20... Training Step: 864... Training loss: 2.0288... 0.0763 sec/batch
Epoch: 1/20... Training Step: 865... Training loss: 1.9800... 0.0756 sec/batch
Epoch: 1/20... Training Step: 866... Training loss: 2.0009... 0.0759 sec/batch
Epoch: 1/20... Training Step: 867... Training loss: 1.9573... 0.0755 sec/batch
Epoch: 1/20... Training Step: 868... Training loss: 2.0967... 0.0750 sec/batch
Epoch: 1/20... Training Step: 869... Training loss: 2.0514... 0.0746 sec/batch
Epoch: 1/20... Training Step: 870... Training loss: 2.0322... 0.0770 sec/batch
Epoch: 1/20... Training Step: 871... Training loss: 1.9969... 0.0751 sec/batch
Epoch: 1/20... Training Step: 872... Training loss: 2.0283... 0.0780 sec/batch
Epoch: 1/20... Training Step: 873... Training loss: 2.0538... 0.0754 sec/batch
Epoch: 1/20... Training Step: 874... Training loss: 2.0132... 0.0762 sec/batch
Epoch: 1/20... Training Step: 875... Training loss: 2.0328... 0.0763 sec/batch
Epoch: 1/20... Training Step: 876... Training loss: 2.0482... 0.0760 sec/batch
Epoch: 1/20... Training Step: 877... Training loss: 2.0469... 0.0779 sec/batch
Epoch: 1/20... Training Step: 878... Training loss: 2.0613... 0.0764 sec/batch
Epoch: 1/20... Training Step: 879... Training loss: 1.9781... 0.0751 sec/batch
Epoch: 1/20... Training Step: 880... Training loss: 1.9731... 0.0735 sec/batch
Epoch: 1/20... Training Step: 881... Training loss: 2.0177... 0.0765 sec/batch
Epoch: 1/20... Training Step: 882... Training loss: 1.9495... 0.0789 sec/batch
Epoch: 1/20... Training Step: 883... Training loss: 1.9857... 0.0745 sec/batch
Epoch: 1/20... Training Step: 884... Training loss: 2.0294... 0.0757 sec/batch
Epoch: 1/20... Training Step: 885... Training loss: 1.9778... 0.0775 sec/batch
Epoch: 1/20... Training Step: 886... Training loss: 2.0664... 0.0826 sec/batch
Epoch: 1/20... Training Step: 887... Training loss: 1.9821... 0.0801 sec/batch
Epoch: 1/20... Training Step: 888... Training loss: 2.1084... 0.0759 sec/batch
Epoch: 1/20... Training Step: 889... Training loss: 1.9715... 0.0768 sec/batch
Epoch: 1/20... Training Step: 890... Training loss: 2.0641... 0.0764 sec/batch
Epoch: 1/20... Training Step: 891... Training loss: 2.0331... 0.0769 sec/batch
Epoch: 1/20... Training Step: 892... Training loss: 1.9246... 0.0735 sec/batch
Epoch: 1/20... Training Step: 893... Training loss: 1.9919... 0.0737 sec/batch
Epoch: 1/20... Training Step: 894... Training loss: 1.9621... 0.0757 sec/batch
Epoch: 1/20... Training Step: 895... Training loss: 2.0686... 0.0750 sec/batch
Epoch: 1/20... Training Step: 896... Training loss: 2.0268... 0.0774 sec/batch
Epoch: 1/20... Training Step: 897... Training loss: 1.9923... 0.0845 sec/batch
Epoch: 1/20... Training Step: 898... Training loss: 2.0455... 0.0754 sec/batch
Epoch: 1/20... Training Step: 899... Training loss: 1.9878... 0.0785 sec/batch
Epoch: 1/20... Training Step: 900... Training loss: 1.9766... 0.0798 sec/batch
Epoch: 1/20... Training Step: 901... Training loss: 2.0445... 0.0782 sec/batch
Epoch: 1/20... Training Step: 902... Training loss: 1.9121... 0.0836 sec/batch
Epoch: 1/20... Training Step: 903... Training loss: 1.9165... 0.0781 sec/batch
Epoch: 1/20... Training Step: 904... Training loss: 1.9296... 0.0753 sec/batch
Epoch: 1/20... Training Step: 905... Training loss: 1.9486... 0.0761 sec/batch
Epoch: 1/20... Training Step: 906... Training loss: 1.9697... 0.0757 sec/batch
Epoch: 1/20... Training Step: 907... Training loss: 1.9262... 0.0737 sec/batch
Epoch: 1/20... Training Step: 908... Training loss: 2.0175... 0.0758 sec/batch
Epoch: 1/20... Training Step: 909... Training loss: 1.9880... 0.0807 sec/batch
Epoch: 1/20... Training Step: 910... Training loss: 1.9393... 0.0790 sec/batch
Epoch: 1/20... Training Step: 911... Training loss: 1.9731... 0.0736 sec/batch
Epoch: 1/20... Training Step: 912... Training loss: 1.9478... 0.0789 sec/batch
Epoch: 1/20... Training Step: 913... Training loss: 2.0224... 0.0759 sec/batch
Epoch: 1/20... Training Step: 914... Training loss: 1.9780... 0.0760 sec/batch
Epoch: 1/20... Training Step: 915... Training loss: 1.9800... 0.0756 sec/batch
Epoch: 1/20... Training Step: 916... Training loss: 1.9596... 0.0740 sec/batch
Epoch: 1/20... Training Step: 917... Training loss: 1.9773... 0.0733 sec/batch
Epoch: 1/20... Training Step: 918... Training loss: 1.9819... 0.0776 sec/batch
Epoch: 1/20... Training Step: 919... Training loss: 1.9452... 0.0746 sec/batch
Epoch: 1/20... Training Step: 920... Training loss: 1.9856... 0.0768 sec/batch
Epoch: 1/20... Training Step: 921... Training loss: 2.1537... 0.0764 sec/batch
Epoch: 1/20... Training Step: 922... Training loss: 1.9939... 0.0762 sec/batch
Epoch: 1/20... Training Step: 923... Training loss: 2.0019... 0.0780 sec/batch
Epoch: 1/20... Training Step: 924... Training loss: 2.0043... 0.0759 sec/batch
Epoch: 1/20... Training Step: 925... Training loss: 2.0051... 0.0757 sec/batch
Epoch: 1/20... Training Step: 926... Training loss: 2.0738... 0.0752 sec/batch
Epoch: 1/20... Training Step: 927... Training loss: 2.0228... 0.0765 sec/batch
Epoch: 1/20... Training Step: 928... Training loss: 2.0208... 0.0809 sec/batch
Epoch: 1/20... Training Step: 929... Training loss: 2.0852... 0.0746 sec/batch
Epoch: 1/20... Training Step: 930... Training loss: 2.0835... 0.0752 sec/batch
Epoch: 1/20... Training Step: 931... Training loss: 1.9785... 0.0788 sec/batch
Epoch: 1/20... Training Step: 932... Training loss: 2.0514... 0.0786 sec/batch
Epoch: 1/20... Training Step: 933... Training loss: 1.9407... 0.0763 sec/batch
Epoch: 1/20... Training Step: 934... Training loss: 1.9917... 0.0770 sec/batch
Epoch: 1/20... Training Step: 935... Training loss: 1.9994... 0.0768 sec/batch
Epoch: 1/20... Training Step: 936... Training loss: 1.9931... 0.0757 sec/batch
Epoch: 1/20... Training Step: 937... Training loss: 1.9654... 0.0738 sec/batch
Epoch: 1/20... Training Step: 938... Training loss: 1.9986... 0.0759 sec/batch
Epoch: 1/20... Training Step: 939... Training loss: 2.0098... 0.0769 sec/batch
Epoch: 1/20... Training Step: 940... Training loss: 2.0343... 0.0744 sec/batch
Epoch: 1/20... Training Step: 941... Training loss: 1.9903... 0.0747 sec/batch
Epoch: 1/20... Training Step: 942... Training loss: 1.9755... 0.0756 sec/batch
Epoch: 1/20... Training Step: 943... Training loss: 2.0020... 0.0748 sec/batch
Epoch: 1/20... Training Step: 944... Training loss: 1.9927... 0.0740 sec/batch
Epoch: 1/20... Training Step: 945... Training loss: 2.0263... 0.0775 sec/batch
Epoch: 1/20... Training Step: 946... Training loss: 2.0207... 0.0749 sec/batch
Epoch: 1/20... Training Step: 947... Training loss: 1.9690... 0.0745 sec/batch
Epoch: 1/20... Training Step: 948... Training loss: 1.9421... 0.0768 sec/batch
Epoch: 1/20... Training Step: 949... Training loss: 1.9429... 0.0739 sec/batch
Epoch: 1/20... Training Step: 950... Training loss: 1.8973... 0.0737 sec/batch
Epoch: 1/20... Training Step: 951... Training loss: 2.0009... 0.0759 sec/batch
Epoch: 1/20... Training Step: 952... Training loss: 1.9937... 0.0751 sec/batch
Epoch: 1/20... Training Step: 953... Training loss: 2.0124... 0.0784 sec/batch
Epoch: 1/20... Training Step: 954... Training loss: 1.9492... 0.0779 sec/batch
Epoch: 1/20... Training Step: 955... Training loss: 2.0266... 0.0778 sec/batch
Epoch: 1/20... Training Step: 956... Training loss: 2.0367... 0.0747 sec/batch
Epoch: 1/20... Training Step: 957... Training loss: 1.9800... 0.0768 sec/batch
Epoch: 1/20... Training Step: 958... Training loss: 2.0340... 0.0754 sec/batch
Epoch: 1/20... Training Step: 959... Training loss: 1.9227... 0.0757 sec/batch
Epoch: 1/20... Training Step: 960... Training loss: 1.8820... 0.0748 sec/batch
Epoch: 1/20... Training Step: 961... Training loss: 1.9770... 0.0773 sec/batch
Epoch: 1/20... Training Step: 962... Training loss: 1.9905... 0.0765 sec/batch
Epoch: 1/20... Training Step: 963... Training loss: 2.0102... 0.0841 sec/batch
Epoch: 1/20... Training Step: 964... Training loss: 1.9359... 0.0768 sec/batch
Epoch: 1/20... Training Step: 965... Training loss: 2.0586... 0.0761 sec/batch
Epoch: 1/20... Training Step: 966... Training loss: 2.0906... 0.0762 sec/batch
Epoch: 1/20... Training Step: 967... Training loss: 1.9870... 0.0761 sec/batch
Epoch: 1/20... Training Step: 968... Training loss: 1.9802... 0.0821 sec/batch
Epoch: 1/20... Training Step: 969... Training loss: 1.9850... 0.0798 sec/batch
Epoch: 1/20... Training Step: 970... Training loss: 2.1454... 0.0778 sec/batch
Epoch: 1/20... Training Step: 971... Training loss: 1.8818... 0.0771 sec/batch
Epoch: 1/20... Training Step: 972... Training loss: 2.0782... 0.0776 sec/batch
Epoch: 1/20... Training Step: 973... Training loss: 2.0422... 0.0807 sec/batch
Epoch: 1/20... Training Step: 974... Training loss: 1.9489... 0.0794 sec/batch
Epoch: 1/20... Training Step: 975... Training loss: 1.9867... 0.0760 sec/batch
Epoch: 1/20... Training Step: 976... Training loss: 2.0424... 0.0810 sec/batch
Epoch: 1/20... Training Step: 977... Training loss: 2.0224... 0.0785 sec/batch
Epoch: 1/20... Training Step: 978... Training loss: 1.9600... 0.0773 sec/batch
Epoch: 1/20... Training Step: 979... Training loss: 1.9935... 0.0779 sec/batch
Epoch: 1/20... Training Step: 980... Training loss: 1.9768... 0.0780 sec/batch
Epoch: 1/20... Training Step: 981... Training loss: 1.9087... 0.0932 sec/batch
Epoch: 1/20... Training Step: 982... Training loss: 2.0145... 0.0786 sec/batch
Epoch: 1/20... Training Step: 983... Training loss: 1.9898... 0.0758 sec/batch
Epoch: 1/20... Training Step: 984... Training loss: 1.9393... 0.0758 sec/batch
Epoch: 1/20... Training Step: 985... Training loss: 1.9543... 0.0759 sec/batch
Epoch: 1/20... Training Step: 986... Training loss: 1.9472... 0.0761 sec/batch
Epoch: 1/20... Training Step: 987... Training loss: 1.9305... 0.0750 sec/batch
Epoch: 1/20... Training Step: 988... Training loss: 1.9347... 0.0745 sec/batch
Epoch: 1/20... Training Step: 989... Training loss: 1.9159... 0.0898 sec/batch
Epoch: 1/20... Training Step: 990... Training loss: 1.9182... 0.0794 sec/batch
Epoch: 1/20... Training Step: 991... Training loss: 1.9392... 0.0750 sec/batch
Epoch: 1/20... Training Step: 992... Training loss: 1.9836... 0.0774 sec/batch
Epoch: 2/20... Training Step: 993... Training loss: 2.0591... 0.0745 sec/batch
Epoch: 2/20... Training Step: 994... Training loss: 1.9237... 0.0743 sec/batch
Epoch: 2/20... Training Step: 995... Training loss: 1.9902... 0.0761 sec/batch
Epoch: 2/20... Training Step: 996... Training loss: 1.9423... 0.0762 sec/batch
Epoch: 2/20... Training Step: 997... Training loss: 1.8614... 0.0764 sec/batch
Epoch: 2/20... Training Step: 998... Training loss: 1.8804... 0.0771 sec/batch
Epoch: 2/20... Training Step: 999... Training loss: 1.8857... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1000... Training loss: 1.9012... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1001... Training loss: 1.9692... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1002... Training loss: 1.8573... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1003... Training loss: 1.9650... 0.0784 sec/batch
Epoch: 2/20... Training Step: 1004... Training loss: 1.8822... 0.0784 sec/batch
Epoch: 2/20... Training Step: 1005... Training loss: 1.9233... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1006... Training loss: 1.9592... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1007... Training loss: 1.9512... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1008... Training loss: 1.9069... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1009... Training loss: 1.9242... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1010... Training loss: 1.9238... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1011... Training loss: 1.9051... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1012... Training loss: 1.8416... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1013... Training loss: 1.9552... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1014... Training loss: 1.8897... 0.0800 sec/batch
Epoch: 2/20... Training Step: 1015... Training loss: 1.8869... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1016... Training loss: 1.8636... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1017... Training loss: 1.9321... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1018... Training loss: 1.8815... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1019... Training loss: 1.9499... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1020... Training loss: 1.8962... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1021... Training loss: 1.9187... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1022... Training loss: 1.9677... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1023... Training loss: 1.9189... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1024... Training loss: 1.9369... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1025... Training loss: 1.9307... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1026... Training loss: 1.9532... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1027... Training loss: 1.9918... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1028... Training loss: 1.9728... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1029... Training loss: 1.8992... 0.0785 sec/batch
Epoch: 2/20... Training Step: 1030... Training loss: 1.9950... 0.0779 sec/batch
Epoch: 2/20... Training Step: 1031... Training loss: 1.8854... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1032... Training loss: 1.8764... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1033... Training loss: 1.8604... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1034... Training loss: 1.9131... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1035... Training loss: 1.8880... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1036... Training loss: 1.9092... 0.0776 sec/batch
Epoch: 2/20... Training Step: 1037... Training loss: 1.9526... 0.0793 sec/batch
Epoch: 2/20... Training Step: 1038... Training loss: 1.8898... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1039... Training loss: 1.9460... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1040... Training loss: 1.9096... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1041... Training loss: 1.9403... 0.0783 sec/batch
Epoch: 2/20... Training Step: 1042... Training loss: 1.9635... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1043... Training loss: 1.8009... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1044... Training loss: 1.9658... 0.0795 sec/batch
Epoch: 2/20... Training Step: 1045... Training loss: 1.8570... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1046... Training loss: 1.8571... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1047... Training loss: 1.8511... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1048... Training loss: 1.9101... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1049... Training loss: 1.9107... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1050... Training loss: 1.9160... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1051... Training loss: 1.8660... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1052... Training loss: 1.9312... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1053... Training loss: 1.8388... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1054... Training loss: 1.8659... 0.0784 sec/batch
Epoch: 2/20... Training Step: 1055... Training loss: 1.9429... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1056... Training loss: 1.8142... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1057... Training loss: 1.9305... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1058... Training loss: 1.9129... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1059... Training loss: 1.8487... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1060... Training loss: 1.8261... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1061... Training loss: 1.8866... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1062... Training loss: 1.9341... 0.0778 sec/batch
Epoch: 2/20... Training Step: 1063... Training loss: 1.9340... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1064... Training loss: 1.9520... 0.0792 sec/batch
Epoch: 2/20... Training Step: 1065... Training loss: 1.8439... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1066... Training loss: 1.8208... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1067... Training loss: 1.8460... 0.0795 sec/batch
Epoch: 2/20... Training Step: 1068... Training loss: 1.8198... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1069... Training loss: 1.8877... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1070... Training loss: 1.8734... 0.0788 sec/batch
Epoch: 2/20... Training Step: 1071... Training loss: 1.9477... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1072... Training loss: 1.9226... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1073... Training loss: 1.9241... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1074... Training loss: 1.8561... 0.0792 sec/batch
Epoch: 2/20... Training Step: 1075... Training loss: 1.9183... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1076... Training loss: 1.9356... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1077... Training loss: 1.8975... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1078... Training loss: 1.9555... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1079... Training loss: 1.8408... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1080... Training loss: 1.9155... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1081... Training loss: 1.9013... 0.0827 sec/batch
Epoch: 2/20... Training Step: 1082... Training loss: 1.9592... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1083... Training loss: 2.0144... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1084... Training loss: 1.8446... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1085... Training loss: 1.8196... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1086... Training loss: 1.7661... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1087... Training loss: 1.8153... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1088... Training loss: 1.9360... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1089... Training loss: 1.8994... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1090... Training loss: 1.8282... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1091... Training loss: 1.8846... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1092... Training loss: 1.8570... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1093... Training loss: 1.8382... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1094... Training loss: 1.8781... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1095... Training loss: 1.8751... 0.0796 sec/batch
Epoch: 2/20... Training Step: 1096... Training loss: 1.9426... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1097... Training loss: 1.8828... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1098... Training loss: 1.8733... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1099... Training loss: 1.9053... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1100... Training loss: 1.9434... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1101... Training loss: 1.8955... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1102... Training loss: 1.8957... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1103... Training loss: 1.8675... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1104... Training loss: 1.8429... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1105... Training loss: 2.0175... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1106... Training loss: 1.8912... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1107... Training loss: 1.8492... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1108... Training loss: 1.9996... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1109... Training loss: 1.9121... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1110... Training loss: 1.9656... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1111... Training loss: 1.8463... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1112... Training loss: 1.8854... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1113... Training loss: 1.9806... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1114... Training loss: 1.9129... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1115... Training loss: 1.8927... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1116... Training loss: 1.8526... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1117... Training loss: 1.9608... 0.0730 sec/batch
Epoch: 2/20... Training Step: 1118... Training loss: 1.9663... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1119... Training loss: 1.9144... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1120... Training loss: 1.9035... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1121... Training loss: 1.8973... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1122... Training loss: 1.8857... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1123... Training loss: 1.8484... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1124... Training loss: 1.9305... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1125... Training loss: 1.8952... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1126... Training loss: 1.8604... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1127... Training loss: 1.8675... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1128... Training loss: 1.9079... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1129... Training loss: 1.8761... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1130... Training loss: 1.9788... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1131... Training loss: 1.9517... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1132... Training loss: 1.8797... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1133... Training loss: 1.8760... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1134... Training loss: 1.8578... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1135... Training loss: 1.8372... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1136... Training loss: 1.8732... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1137... Training loss: 1.7997... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1138... Training loss: 1.9085... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1139... Training loss: 1.8705... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1140... Training loss: 1.7822... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1141... Training loss: 1.8616... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1142... Training loss: 1.8610... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1143... Training loss: 1.7531... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1144... Training loss: 1.8882... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1145... Training loss: 1.8377... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1146... Training loss: 1.8461... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1147... Training loss: 1.8084... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1148... Training loss: 1.8273... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1149... Training loss: 1.8651... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1150... Training loss: 1.8271... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1151... Training loss: 1.8534... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1152... Training loss: 1.9029... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1153... Training loss: 1.8349... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1154... Training loss: 1.8129... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1155... Training loss: 1.8452... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1156... Training loss: 1.9086... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1157... Training loss: 1.8828... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1158... Training loss: 1.9319... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1159... Training loss: 1.8674... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1160... Training loss: 1.8711... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1161... Training loss: 1.9039... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1162... Training loss: 1.8867... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1163... Training loss: 1.9406... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1164... Training loss: 1.7905... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1165... Training loss: 1.8172... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1166... Training loss: 1.8832... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1167... Training loss: 1.8530... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1168... Training loss: 1.8694... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1169... Training loss: 1.9012... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1170... Training loss: 1.9654... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1171... Training loss: 1.9423... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1172... Training loss: 1.8382... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1173... Training loss: 1.9369... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1174... Training loss: 1.9524... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1175... Training loss: 1.8188... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1176... Training loss: 1.8576... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1177... Training loss: 1.8847... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1178... Training loss: 1.9003... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1179... Training loss: 1.9169... 0.0782 sec/batch
Epoch: 2/20... Training Step: 1180... Training loss: 1.8674... 0.0782 sec/batch
Epoch: 2/20... Training Step: 1181... Training loss: 1.8606... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1182... Training loss: 1.9113... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1183... Training loss: 1.9339... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1184... Training loss: 1.9521... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1185... Training loss: 1.9251... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1186... Training loss: 1.9289... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1187... Training loss: 1.8540... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1188... Training loss: 1.8235... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1189... Training loss: 1.9150... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1190... Training loss: 1.8459... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1191... Training loss: 1.8703... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1192... Training loss: 1.8873... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1193... Training loss: 1.8695... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1194... Training loss: 1.8468... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1195... Training loss: 1.8642... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1196... Training loss: 1.8590... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1197... Training loss: 1.8855... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1198... Training loss: 1.9406... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1199... Training loss: 1.8512... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1200... Training loss: 1.8048... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1201... Training loss: 1.8835... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1202... Training loss: 1.8618... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1203... Training loss: 1.8263... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1204... Training loss: 1.9208... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1205... Training loss: 1.8512... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1206... Training loss: 1.8266... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1207... Training loss: 1.8468... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1208... Training loss: 1.8703... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1209... Training loss: 1.8360... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1210... Training loss: 1.8933... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1211... Training loss: 1.8774... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1212... Training loss: 1.8718... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1213... Training loss: 1.8145... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1214... Training loss: 1.8839... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1215... Training loss: 1.8382... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1216... Training loss: 1.8304... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1217... Training loss: 1.9140... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1218... Training loss: 1.8556... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1219... Training loss: 1.8708... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1220... Training loss: 1.8712... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1221... Training loss: 1.7830... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1222... Training loss: 1.8605... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1223... Training loss: 1.8195... 0.0792 sec/batch
Epoch: 2/20... Training Step: 1224... Training loss: 1.8798... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1225... Training loss: 1.8276... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1226... Training loss: 1.9340... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1227... Training loss: 1.7880... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1228... Training loss: 1.8714... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1229... Training loss: 1.7822... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1230... Training loss: 1.9025... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1231... Training loss: 1.8590... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1232... Training loss: 1.8480... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1233... Training loss: 1.8224... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1234... Training loss: 1.8063... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1235... Training loss: 1.8396... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1236... Training loss: 1.9119... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1237... Training loss: 1.8251... 0.0784 sec/batch
Epoch: 2/20... Training Step: 1238... Training loss: 1.8863... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1239... Training loss: 1.8157... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1240... Training loss: 1.7904... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1241... Training loss: 1.7832... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1242... Training loss: 1.8257... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1243... Training loss: 1.8257... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1244... Training loss: 1.8406... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1245... Training loss: 1.8237... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1246... Training loss: 1.8336... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1247... Training loss: 1.8419... 0.0780 sec/batch
Epoch: 2/20... Training Step: 1248... Training loss: 1.8321... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1249... Training loss: 1.9314... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1250... Training loss: 1.8313... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1251... Training loss: 1.8285... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1252... Training loss: 1.8477... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1253... Training loss: 1.9213... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1254... Training loss: 1.9184... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1255... Training loss: 1.8517... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1256... Training loss: 1.8786... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1257... Training loss: 1.8718... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1258... Training loss: 1.8623... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1259... Training loss: 1.7873... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1260... Training loss: 1.8247... 0.0784 sec/batch
Epoch: 2/20... Training Step: 1261... Training loss: 1.8044... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1262... Training loss: 1.7896... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1263... Training loss: 1.8390... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1264... Training loss: 1.8423... 0.0776 sec/batch
Epoch: 2/20... Training Step: 1265... Training loss: 1.7317... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1266... Training loss: 1.7645... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1267... Training loss: 1.8288... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1268... Training loss: 1.8497... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1269... Training loss: 1.8274... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1270... Training loss: 1.7676... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1271... Training loss: 1.8050... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1272... Training loss: 1.9156... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1273... Training loss: 1.8560... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1274... Training loss: 1.8544... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1275... Training loss: 1.8615... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1276... Training loss: 1.8144... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1277... Training loss: 1.8035... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1278... Training loss: 1.8117... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1279... Training loss: 1.8466... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1280... Training loss: 1.7768... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1281... Training loss: 1.8065... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1282... Training loss: 1.8283... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1283... Training loss: 1.8389... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1284... Training loss: 1.7847... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1285... Training loss: 1.8023... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1286... Training loss: 1.7824... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1287... Training loss: 1.8192... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1288... Training loss: 1.8320... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1289... Training loss: 1.8041... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1290... Training loss: 1.8907... 0.0822 sec/batch
Epoch: 2/20... Training Step: 1291... Training loss: 1.8650... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1292... Training loss: 1.8670... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1293... Training loss: 1.8472... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1294... Training loss: 1.7999... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1295... Training loss: 1.7993... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1296... Training loss: 1.8743... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1297... Training loss: 1.8565... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1298... Training loss: 1.8899... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1299... Training loss: 1.8065... 0.0730 sec/batch
Epoch: 2/20... Training Step: 1300... Training loss: 1.7323... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1301... Training loss: 1.9146... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1302... Training loss: 1.8135... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1303... Training loss: 1.8798... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1304... Training loss: 1.8547... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1305... Training loss: 1.8146... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1306... Training loss: 1.7873... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1307... Training loss: 1.8216... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1308... Training loss: 1.7198... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1309... Training loss: 1.7589... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1310... Training loss: 1.7874... 0.0803 sec/batch
Epoch: 2/20... Training Step: 1311... Training loss: 1.7784... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1312... Training loss: 1.8227... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1313... Training loss: 1.8794... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1314... Training loss: 1.7463... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1315... Training loss: 1.7513... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1316... Training loss: 1.7869... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1317... Training loss: 1.8314... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1318... Training loss: 1.7704... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1319... Training loss: 1.7692... 0.0776 sec/batch
Epoch: 2/20... Training Step: 1320... Training loss: 1.7542... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1321... Training loss: 1.8156... 0.0780 sec/batch
Epoch: 2/20... Training Step: 1322... Training loss: 1.8081... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1323... Training loss: 1.8067... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1324... Training loss: 1.8013... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1325... Training loss: 1.9231... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1326... Training loss: 1.8719... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1327... Training loss: 1.8720... 0.0778 sec/batch
Epoch: 2/20... Training Step: 1328... Training loss: 1.8542... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1329... Training loss: 1.8668... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1330... Training loss: 1.7812... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1331... Training loss: 1.8912... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1332... Training loss: 1.7992... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1333... Training loss: 1.8229... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1334... Training loss: 1.8038... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1335... Training loss: 1.7509... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1336... Training loss: 1.8571... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1337... Training loss: 1.8041... 0.0728 sec/batch
Epoch: 2/20... Training Step: 1338... Training loss: 1.8775... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1339... Training loss: 1.8511... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1340... Training loss: 1.8320... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1341... Training loss: 1.8657... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1342... Training loss: 1.9142... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1343... Training loss: 1.8281... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1344... Training loss: 1.7441... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1345... Training loss: 1.7799... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1346... Training loss: 1.7944... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1347... Training loss: 1.8063... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1348... Training loss: 1.8227... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1349... Training loss: 1.7928... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1350... Training loss: 1.8176... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1351... Training loss: 1.8570... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1352... Training loss: 1.7824... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1353... Training loss: 1.8097... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1354... Training loss: 1.8926... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1355... Training loss: 1.8135... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1356... Training loss: 1.8147... 0.0788 sec/batch
Epoch: 2/20... Training Step: 1357... Training loss: 1.7797... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1358... Training loss: 1.8011... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1359... Training loss: 1.7846... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1360... Training loss: 1.7584... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1361... Training loss: 1.7454... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1362... Training loss: 1.7595... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1363... Training loss: 1.8136... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1364... Training loss: 1.7541... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1365... Training loss: 1.8092... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1366... Training loss: 1.7535... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1367... Training loss: 1.8072... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1368... Training loss: 1.7794... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1369... Training loss: 1.7806... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1370... Training loss: 1.8342... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1371... Training loss: 1.8146... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1372... Training loss: 1.8208... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1373... Training loss: 1.7914... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1374... Training loss: 1.7548... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1375... Training loss: 1.8208... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1376... Training loss: 1.8650... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1377... Training loss: 1.8353... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1378... Training loss: 1.8019... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1379... Training loss: 1.8571... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1380... Training loss: 1.8548... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1381... Training loss: 1.8403... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1382... Training loss: 1.7903... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1383... Training loss: 1.8384... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1384... Training loss: 1.7841... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1385... Training loss: 1.8312... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1386... Training loss: 1.7812... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1387... Training loss: 1.8013... 0.0805 sec/batch
Epoch: 2/20... Training Step: 1388... Training loss: 1.8022... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1389... Training loss: 1.7740... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1390... Training loss: 1.8375... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1391... Training loss: 1.8108... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1392... Training loss: 1.7468... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1393... Training loss: 1.7534... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1394... Training loss: 1.8438... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1395... Training loss: 1.7837... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1396... Training loss: 1.8645... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1397... Training loss: 1.7150... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1398... Training loss: 1.7789... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1399... Training loss: 1.7514... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1400... Training loss: 1.8967... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1401... Training loss: 1.7546... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1402... Training loss: 1.7728... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1403... Training loss: 1.7230... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1404... Training loss: 1.7157... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1405... Training loss: 1.7931... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1406... Training loss: 1.7291... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1407... Training loss: 1.7345... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1408... Training loss: 1.8534... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1409... Training loss: 1.7312... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1410... Training loss: 1.7653... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1411... Training loss: 1.7930... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1412... Training loss: 1.9093... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1413... Training loss: 1.7594... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1414... Training loss: 1.7744... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1415... Training loss: 1.7654... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1416... Training loss: 1.7781... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1417... Training loss: 1.7859... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1418... Training loss: 1.7963... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1419... Training loss: 1.8535... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1420... Training loss: 1.8536... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1421... Training loss: 1.7989... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1422... Training loss: 1.8664... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1423... Training loss: 1.8005... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1424... Training loss: 1.8550... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1425... Training loss: 1.8897... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1426... Training loss: 1.8194... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1427... Training loss: 1.8224... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1428... Training loss: 1.7925... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1429... Training loss: 1.8387... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1430... Training loss: 1.9114... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1431... Training loss: 1.8114... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1432... Training loss: 1.8184... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1433... Training loss: 1.8251... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1434... Training loss: 1.8008... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1435... Training loss: 1.8326... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1436... Training loss: 1.7940... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1437... Training loss: 1.7459... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1438... Training loss: 1.8238... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1439... Training loss: 1.7645... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1440... Training loss: 1.7806... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1441... Training loss: 1.8258... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1442... Training loss: 1.8853... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1443... Training loss: 1.7959... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1444... Training loss: 1.8064... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1445... Training loss: 1.7991... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1446... Training loss: 1.8317... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1447... Training loss: 1.8164... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1448... Training loss: 1.8042... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1449... Training loss: 1.7055... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1450... Training loss: 1.7539... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1451... Training loss: 1.8695... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1452... Training loss: 1.8279... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1453... Training loss: 1.8103... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1454... Training loss: 1.7563... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1455... Training loss: 1.7519... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1456... Training loss: 1.7457... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1457... Training loss: 1.7246... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1458... Training loss: 1.7588... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1459... Training loss: 1.6848... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1460... Training loss: 1.7952... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1461... Training loss: 1.7728... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1462... Training loss: 1.8516... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1463... Training loss: 1.7538... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1464... Training loss: 1.7433... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1465... Training loss: 1.7477... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1466... Training loss: 1.7953... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1467... Training loss: 1.6381... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1468... Training loss: 1.7700... 0.0776 sec/batch
Epoch: 2/20... Training Step: 1469... Training loss: 1.7648... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1470... Training loss: 1.7410... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1471... Training loss: 1.7320... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1472... Training loss: 1.7187... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1473... Training loss: 1.7248... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1474... Training loss: 1.7843... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1475... Training loss: 1.7617... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1476... Training loss: 1.7904... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1477... Training loss: 1.7226... 0.0778 sec/batch
Epoch: 2/20... Training Step: 1478... Training loss: 1.7350... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1479... Training loss: 1.7976... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1480... Training loss: 1.7542... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1481... Training loss: 1.7283... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1482... Training loss: 1.7248... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1483... Training loss: 1.7134... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1484... Training loss: 1.7227... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1485... Training loss: 1.7739... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1486... Training loss: 1.7699... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1487... Training loss: 1.7022... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1488... Training loss: 1.6937... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1489... Training loss: 1.8108... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1490... Training loss: 1.7620... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1491... Training loss: 1.7782... 0.0779 sec/batch
Epoch: 2/20... Training Step: 1492... Training loss: 1.7697... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1493... Training loss: 1.7383... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1494... Training loss: 1.7628... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1495... Training loss: 1.7692... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1496... Training loss: 1.7672... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1497... Training loss: 1.8096... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1498... Training loss: 1.7219... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1499... Training loss: 1.7819... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1500... Training loss: 1.7822... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1501... Training loss: 1.7445... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1502... Training loss: 1.7184... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1503... Training loss: 1.7820... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1504... Training loss: 1.8123... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1505... Training loss: 1.8376... 0.0797 sec/batch
Epoch: 2/20... Training Step: 1506... Training loss: 1.8472... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1507... Training loss: 1.7744... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1508... Training loss: 1.7443... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1509... Training loss: 1.7702... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1510... Training loss: 1.7595... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1511... Training loss: 1.7413... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1512... Training loss: 1.7391... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1513... Training loss: 1.7982... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1514... Training loss: 1.7359... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1515... Training loss: 1.7970... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1516... Training loss: 1.7220... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1517... Training loss: 1.7726... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1518... Training loss: 1.7776... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1519... Training loss: 1.7650... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1520... Training loss: 1.6868... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1521... Training loss: 1.7374... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1522... Training loss: 1.8272... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1523... Training loss: 1.8794... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1524... Training loss: 1.7938... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1525... Training loss: 1.8305... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1526... Training loss: 1.8037... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1527... Training loss: 1.7850... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1528... Training loss: 1.7690... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1529... Training loss: 1.7857... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1530... Training loss: 1.7371... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1531... Training loss: 1.7276... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1532... Training loss: 1.7149... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1533... Training loss: 1.8998... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1534... Training loss: 1.7251... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1535... Training loss: 1.6992... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1536... Training loss: 1.7357... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1537... Training loss: 1.7479... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1538... Training loss: 1.7322... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1539... Training loss: 1.8219... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1540... Training loss: 1.8538... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1541... Training loss: 1.7377... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1542... Training loss: 1.7193... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1543... Training loss: 1.7693... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1544... Training loss: 1.7086... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1545... Training loss: 1.6920... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1546... Training loss: 1.7924... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1547... Training loss: 1.7667... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1548... Training loss: 1.6989... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1549... Training loss: 1.7042... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1550... Training loss: 1.8071... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1551... Training loss: 1.7784... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1552... Training loss: 1.6672... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1553... Training loss: 1.7530... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1554... Training loss: 1.7589... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1555... Training loss: 1.6806... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1556... Training loss: 1.7252... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1557... Training loss: 1.7512... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1558... Training loss: 1.7909... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1559... Training loss: 1.7677... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1560... Training loss: 1.6842... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1561... Training loss: 1.6925... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1562... Training loss: 1.7005... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1563... Training loss: 1.6882... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1564... Training loss: 1.6891... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1565... Training loss: 1.8084... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1566... Training loss: 1.7796... 0.0782 sec/batch
Epoch: 2/20... Training Step: 1567... Training loss: 1.7614... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1568... Training loss: 1.7274... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1569... Training loss: 1.7727... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1570... Training loss: 1.8711... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1571... Training loss: 1.7878... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1572... Training loss: 1.7230... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1573... Training loss: 1.6905... 0.0728 sec/batch
Epoch: 2/20... Training Step: 1574... Training loss: 1.7260... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1575... Training loss: 1.8178... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1576... Training loss: 1.6945... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1577... Training loss: 1.7319... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1578... Training loss: 1.7515... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1579... Training loss: 1.7574... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1580... Training loss: 1.7752... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1581... Training loss: 1.7043... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1582... Training loss: 1.7532... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1583... Training loss: 1.7240... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1584... Training loss: 1.8544... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1585... Training loss: 1.7323... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1586... Training loss: 1.7300... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1587... Training loss: 1.7038... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1588... Training loss: 1.7931... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1589... Training loss: 1.6573... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1590... Training loss: 1.6468... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1591... Training loss: 1.7174... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1592... Training loss: 1.7386... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1593... Training loss: 1.7981... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1594... Training loss: 1.7299... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1595... Training loss: 1.7166... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1596... Training loss: 1.7330... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1597... Training loss: 1.7042... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1598... Training loss: 1.7463... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1599... Training loss: 1.8055... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1600... Training loss: 1.8058... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1601... Training loss: 1.7888... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1602... Training loss: 1.7886... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1603... Training loss: 1.6587... 0.0774 sec/batch
Epoch: 2/20... Training Step: 1604... Training loss: 1.7485... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1605... Training loss: 1.8025... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1606... Training loss: 1.7149... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1607... Training loss: 1.8161... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1608... Training loss: 1.7424... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1609... Training loss: 1.7776... 0.0778 sec/batch
Epoch: 2/20... Training Step: 1610... Training loss: 1.7705... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1611... Training loss: 1.6752... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1612... Training loss: 1.7366... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1613... Training loss: 1.6961... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1614... Training loss: 1.6438... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1615... Training loss: 1.7850... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1616... Training loss: 1.7487... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1617... Training loss: 1.7100... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1618... Training loss: 1.6961... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1619... Training loss: 1.7098... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1620... Training loss: 1.7818... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1621... Training loss: 1.5618... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1622... Training loss: 1.7407... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1623... Training loss: 1.7338... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1624... Training loss: 1.7114... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1625... Training loss: 1.6820... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1626... Training loss: 1.7430... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1627... Training loss: 1.7458... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1628... Training loss: 1.7029... 0.0780 sec/batch
Epoch: 2/20... Training Step: 1629... Training loss: 1.7803... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1630... Training loss: 1.6870... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1631... Training loss: 1.7866... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1632... Training loss: 1.7907... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1633... Training loss: 1.8237... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1634... Training loss: 1.7731... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1635... Training loss: 1.7318... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1636... Training loss: 1.7325... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1637... Training loss: 1.7123... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1638... Training loss: 1.7836... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1639... Training loss: 1.7508... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1640... Training loss: 1.6536... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1641... Training loss: 1.7129... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1642... Training loss: 1.7177... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1643... Training loss: 1.7104... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1644... Training loss: 1.7774... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1645... Training loss: 1.7514... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1646... Training loss: 1.7870... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1647... Training loss: 1.8024... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1648... Training loss: 1.8330... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1649... Training loss: 1.6732... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1650... Training loss: 1.7060... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1651... Training loss: 1.7546... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1652... Training loss: 1.8114... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1653... Training loss: 1.7659... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1654... Training loss: 1.8017... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1655... Training loss: 1.7184... 0.0759 sec/batch
Epoch: 2/20... Training Step: 1656... Training loss: 1.8649... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1657... Training loss: 1.8235... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1658... Training loss: 1.7035... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1659... Training loss: 1.7999... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1660... Training loss: 1.7832... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1661... Training loss: 1.7243... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1662... Training loss: 1.6948... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1663... Training loss: 1.7476... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1664... Training loss: 1.7329... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1665... Training loss: 1.7199... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1666... Training loss: 1.7223... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1667... Training loss: 1.6977... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1668... Training loss: 1.7442... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1669... Training loss: 1.8187... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1670... Training loss: 1.6964... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1671... Training loss: 1.6763... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1672... Training loss: 1.7411... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1673... Training loss: 1.7391... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1674... Training loss: 1.6975... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1675... Training loss: 1.7080... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1676... Training loss: 1.6860... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1677... Training loss: 1.8079... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1678... Training loss: 1.7562... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1679... Training loss: 1.7152... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1680... Training loss: 1.7379... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1681... Training loss: 1.7090... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1682... Training loss: 1.7386... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1683... Training loss: 1.7197... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1684... Training loss: 1.7548... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1685... Training loss: 1.6778... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1686... Training loss: 1.6917... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1687... Training loss: 1.7472... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1688... Training loss: 1.7283... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1689... Training loss: 1.7641... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1690... Training loss: 1.7860... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1691... Training loss: 1.7050... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1692... Training loss: 1.7658... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1693... Training loss: 1.7197... 0.0730 sec/batch
Epoch: 2/20... Training Step: 1694... Training loss: 1.6957... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1695... Training loss: 1.7453... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1696... Training loss: 1.6707... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1697... Training loss: 1.6217... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1698... Training loss: 1.6597... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1699... Training loss: 1.7516... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1700... Training loss: 1.7516... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1701... Training loss: 1.7899... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1702... Training loss: 1.7057... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1703... Training loss: 1.7204... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1704... Training loss: 1.8072... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1705... Training loss: 1.6714... 0.0730 sec/batch
Epoch: 2/20... Training Step: 1706... Training loss: 1.6930... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1707... Training loss: 1.7241... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1708... Training loss: 1.7793... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1709... Training loss: 1.7713... 0.0792 sec/batch
Epoch: 2/20... Training Step: 1710... Training loss: 1.7406... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1711... Training loss: 1.7761... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1712... Training loss: 1.8418... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1713... Training loss: 1.7385... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1714... Training loss: 1.6909... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1715... Training loss: 1.7265... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1716... Training loss: 1.7579... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1717... Training loss: 1.8153... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1718... Training loss: 1.8610... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1719... Training loss: 1.7157... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1720... Training loss: 1.7558... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1721... Training loss: 1.7574... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1722... Training loss: 1.7178... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1723... Training loss: 1.7115... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1724... Training loss: 1.7349... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1725... Training loss: 1.6776... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1726... Training loss: 1.7760... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1727... Training loss: 1.6789... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1728... Training loss: 1.7290... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1729... Training loss: 1.6385... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1730... Training loss: 1.7645... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1731... Training loss: 1.7051... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1732... Training loss: 1.7258... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1733... Training loss: 1.6918... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1734... Training loss: 1.7065... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1735... Training loss: 1.6968... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1736... Training loss: 1.7668... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1737... Training loss: 1.7127... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1738... Training loss: 1.7170... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1739... Training loss: 1.7507... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1740... Training loss: 1.7378... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1741... Training loss: 1.6884... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1742... Training loss: 1.6472... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1743... Training loss: 1.8022... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1744... Training loss: 1.6764... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1745... Training loss: 1.7049... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1746... Training loss: 1.7204... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1747... Training loss: 1.6685... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1748... Training loss: 1.7080... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1749... Training loss: 1.7239... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1750... Training loss: 1.7882... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1751... Training loss: 1.7425... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1752... Training loss: 1.8233... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1753... Training loss: 1.7605... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1754... Training loss: 1.6604... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1755... Training loss: 1.7188... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1756... Training loss: 1.6930... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1757... Training loss: 1.7113... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1758... Training loss: 1.7414... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1759... Training loss: 1.6955... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1760... Training loss: 1.7031... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1761... Training loss: 1.6779... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1762... Training loss: 1.7076... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1763... Training loss: 1.7378... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1764... Training loss: 1.8301... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1765... Training loss: 1.7563... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1766... Training loss: 1.7744... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1767... Training loss: 1.7457... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1768... Training loss: 1.6489... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1769... Training loss: 1.6908... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1770... Training loss: 1.6972... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1771... Training loss: 1.6966... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1772... Training loss: 1.6819... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1773... Training loss: 1.6853... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1774... Training loss: 1.7054... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1775... Training loss: 1.7365... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1776... Training loss: 1.7307... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1777... Training loss: 1.7073... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1778... Training loss: 1.7091... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1779... Training loss: 1.7350... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1780... Training loss: 1.7213... 0.0729 sec/batch
Epoch: 2/20... Training Step: 1781... Training loss: 1.7214... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1782... Training loss: 1.7262... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1783... Training loss: 1.7073... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1784... Training loss: 1.6458... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1785... Training loss: 1.7033... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1786... Training loss: 1.6602... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1787... Training loss: 1.6210... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1788... Training loss: 1.6721... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1789... Training loss: 1.6905... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1790... Training loss: 1.7145... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1791... Training loss: 1.7205... 0.0856 sec/batch
Epoch: 2/20... Training Step: 1792... Training loss: 1.6246... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1793... Training loss: 1.7710... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1794... Training loss: 1.7116... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1795... Training loss: 1.7600... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1796... Training loss: 1.7240... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1797... Training loss: 1.6639... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1798... Training loss: 1.7255... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1799... Training loss: 1.6712... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1800... Training loss: 1.7334... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1801... Training loss: 1.7673... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1802... Training loss: 1.7925... 0.0730 sec/batch
Epoch: 2/20... Training Step: 1803... Training loss: 1.7548... 0.0779 sec/batch
Epoch: 2/20... Training Step: 1804... Training loss: 1.7805... 0.0782 sec/batch
Epoch: 2/20... Training Step: 1805... Training loss: 1.7193... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1806... Training loss: 1.7756... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1807... Training loss: 1.7751... 0.0781 sec/batch
Epoch: 2/20... Training Step: 1808... Training loss: 1.8008... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1809... Training loss: 1.7769... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1810... Training loss: 1.8060... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1811... Training loss: 1.7574... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1812... Training loss: 1.7817... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1813... Training loss: 1.7224... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1814... Training loss: 1.6355... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1815... Training loss: 1.7168... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1816... Training loss: 1.6965... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1817... Training loss: 1.6955... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1818... Training loss: 1.6964... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1819... Training loss: 1.7480... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1820... Training loss: 1.7918... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1821... Training loss: 1.7402... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1822... Training loss: 1.7030... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1823... Training loss: 1.8015... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1824... Training loss: 1.8287... 0.0731 sec/batch
Epoch: 2/20... Training Step: 1825... Training loss: 1.8554... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1826... Training loss: 1.7151... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1827... Training loss: 1.6258... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1828... Training loss: 1.7309... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1829... Training loss: 1.7769... 0.0768 sec/batch
Epoch: 2/20... Training Step: 1830... Training loss: 1.7568... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1831... Training loss: 1.7908... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1832... Training loss: 1.7079... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1833... Training loss: 1.7901... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1834... Training loss: 1.8184... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1835... Training loss: 1.7806... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1836... Training loss: 1.7990... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1837... Training loss: 1.8363... 0.0786 sec/batch
Epoch: 2/20... Training Step: 1838... Training loss: 1.6711... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1839... Training loss: 1.7163... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1840... Training loss: 1.7357... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1841... Training loss: 1.7137... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1842... Training loss: 1.7124... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1843... Training loss: 1.7051... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1844... Training loss: 1.7929... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1845... Training loss: 1.7730... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1846... Training loss: 1.7476... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1847... Training loss: 1.7331... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1848... Training loss: 1.6994... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1849... Training loss: 1.7174... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1850... Training loss: 1.7217... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1851... Training loss: 1.7662... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1852... Training loss: 1.7437... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1853... Training loss: 1.7476... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1854... Training loss: 1.6865... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1855... Training loss: 1.6855... 0.0756 sec/batch
Epoch: 2/20... Training Step: 1856... Training loss: 1.7243... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1857... Training loss: 1.7003... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1858... Training loss: 1.7066... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1859... Training loss: 1.6817... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1860... Training loss: 1.8291... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1861... Training loss: 1.7435... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1862... Training loss: 1.7885... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1863... Training loss: 1.7516... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1864... Training loss: 1.7505... 0.0772 sec/batch
Epoch: 2/20... Training Step: 1865... Training loss: 1.7775... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1866... Training loss: 1.7420... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1867... Training loss: 1.7345... 0.0763 sec/batch
Epoch: 2/20... Training Step: 1868... Training loss: 1.7715... 0.0770 sec/batch
Epoch: 2/20... Training Step: 1869... Training loss: 1.7737... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1870... Training loss: 1.7812... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1871... Training loss: 1.7112... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1872... Training loss: 1.6940... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1873... Training loss: 1.7286... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1874... Training loss: 1.6551... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1875... Training loss: 1.7073... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1876... Training loss: 1.7112... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1877... Training loss: 1.7237... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1878... Training loss: 1.7990... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1879... Training loss: 1.7278... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1880... Training loss: 1.8145... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1881... Training loss: 1.7484... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1882... Training loss: 1.8153... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1883... Training loss: 1.7815... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1884... Training loss: 1.6607... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1885... Training loss: 1.6887... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1886... Training loss: 1.6537... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1887... Training loss: 1.7977... 0.0771 sec/batch
Epoch: 2/20... Training Step: 1888... Training loss: 1.7272... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1889... Training loss: 1.7140... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1890... Training loss: 1.7606... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1891... Training loss: 1.7087... 0.0762 sec/batch
Epoch: 2/20... Training Step: 1892... Training loss: 1.6656... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1893... Training loss: 1.7041... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1894... Training loss: 1.6603... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1895... Training loss: 1.6444... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1896... Training loss: 1.6451... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1897... Training loss: 1.6614... 0.0713 sec/batch
Epoch: 2/20... Training Step: 1898... Training loss: 1.6839... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1899... Training loss: 1.6621... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1900... Training loss: 1.7751... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1901... Training loss: 1.6738... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1902... Training loss: 1.5860... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1903... Training loss: 1.6661... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1904... Training loss: 1.6941... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1905... Training loss: 1.7943... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1906... Training loss: 1.7256... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1907... Training loss: 1.6970... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1908... Training loss: 1.6798... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1909... Training loss: 1.6930... 0.0761 sec/batch
Epoch: 2/20... Training Step: 1910... Training loss: 1.7336... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1911... Training loss: 1.7021... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1912... Training loss: 1.7172... 0.0744 sec/batch
Epoch: 2/20... Training Step: 1913... Training loss: 1.8659... 0.0758 sec/batch
Epoch: 2/20... Training Step: 1914... Training loss: 1.7508... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1915... Training loss: 1.7638... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1916... Training loss: 1.7203... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1917... Training loss: 1.7432... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1918... Training loss: 1.7511... 0.0760 sec/batch
Epoch: 2/20... Training Step: 1919... Training loss: 1.7367... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1920... Training loss: 1.7382... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1921... Training loss: 1.7952... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1922... Training loss: 1.7943... 0.0743 sec/batch
Epoch: 2/20... Training Step: 1923... Training loss: 1.7400... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1924... Training loss: 1.8052... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1925... Training loss: 1.6736... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1926... Training loss: 1.7240... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1927... Training loss: 1.7355... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1928... Training loss: 1.7439... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1929... Training loss: 1.7022... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1930... Training loss: 1.7290... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1931... Training loss: 1.7083... 0.0757 sec/batch
Epoch: 2/20... Training Step: 1932... Training loss: 1.7879... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1933... Training loss: 1.7433... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1934... Training loss: 1.7343... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1935... Training loss: 1.7664... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1936... Training loss: 1.7032... 0.0773 sec/batch
Epoch: 2/20... Training Step: 1937... Training loss: 1.7780... 0.0745 sec/batch
Epoch: 2/20... Training Step: 1938... Training loss: 1.7348... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1939... Training loss: 1.7128... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1940... Training loss: 1.6819... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1941... Training loss: 1.7051... 0.0765 sec/batch
Epoch: 2/20... Training Step: 1942... Training loss: 1.6072... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1943... Training loss: 1.7408... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1944... Training loss: 1.6660... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1945... Training loss: 1.7147... 0.0751 sec/batch
Epoch: 2/20... Training Step: 1946... Training loss: 1.7071... 0.0732 sec/batch
Epoch: 2/20... Training Step: 1947... Training loss: 1.7832... 0.0748 sec/batch
Epoch: 2/20... Training Step: 1948... Training loss: 1.7555... 0.0769 sec/batch
Epoch: 2/20... Training Step: 1949... Training loss: 1.6959... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1950... Training loss: 1.7479... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1951... Training loss: 1.6702... 0.0775 sec/batch
Epoch: 2/20... Training Step: 1952... Training loss: 1.6439... 0.0764 sec/batch
Epoch: 2/20... Training Step: 1953... Training loss: 1.7116... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1954... Training loss: 1.7332... 0.0753 sec/batch
Epoch: 2/20... Training Step: 1955... Training loss: 1.7446... 0.0734 sec/batch
Epoch: 2/20... Training Step: 1956... Training loss: 1.7033... 0.0750 sec/batch
Epoch: 2/20... Training Step: 1957... Training loss: 1.8174... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1958... Training loss: 1.8199... 0.0735 sec/batch
Epoch: 2/20... Training Step: 1959... Training loss: 1.7449... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1960... Training loss: 1.7515... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1961... Training loss: 1.7197... 0.0777 sec/batch
Epoch: 2/20... Training Step: 1962... Training loss: 1.9194... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1963... Training loss: 1.6411... 0.0791 sec/batch
Epoch: 2/20... Training Step: 1964... Training loss: 1.8151... 0.0746 sec/batch
Epoch: 2/20... Training Step: 1965... Training loss: 1.8179... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1966... Training loss: 1.6941... 0.0742 sec/batch
Epoch: 2/20... Training Step: 1967... Training loss: 1.7638... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1968... Training loss: 1.7883... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1969... Training loss: 1.7790... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1970... Training loss: 1.7526... 0.0754 sec/batch
Epoch: 2/20... Training Step: 1971... Training loss: 1.7602... 0.0736 sec/batch
Epoch: 2/20... Training Step: 1972... Training loss: 1.6947... 0.0752 sec/batch
Epoch: 2/20... Training Step: 1973... Training loss: 1.6397... 0.0738 sec/batch
Epoch: 2/20... Training Step: 1974... Training loss: 1.8141... 0.0749 sec/batch
Epoch: 2/20... Training Step: 1975... Training loss: 1.7717... 0.0747 sec/batch
Epoch: 2/20... Training Step: 1976... Training loss: 1.6993... 0.0740 sec/batch
Epoch: 2/20... Training Step: 1977... Training loss: 1.7124... 0.0755 sec/batch
Epoch: 2/20... Training Step: 1978... Training loss: 1.7032... 0.0741 sec/batch
Epoch: 2/20... Training Step: 1979... Training loss: 1.6782... 0.0739 sec/batch
Epoch: 2/20... Training Step: 1980... Training loss: 1.6934... 0.0766 sec/batch
Epoch: 2/20... Training Step: 1981... Training loss: 1.6829... 0.0767 sec/batch
Epoch: 2/20... Training Step: 1982... Training loss: 1.6241... 0.0737 sec/batch
Epoch: 2/20... Training Step: 1983... Training loss: 1.7025... 0.0733 sec/batch
Epoch: 2/20... Training Step: 1984... Training loss: 1.7533... 0.0747 sec/batch
Epoch: 3/20... Training Step: 1985... Training loss: 1.7619... 0.0739 sec/batch
Epoch: 3/20... Training Step: 1986... Training loss: 1.6677... 0.0765 sec/batch
Epoch: 3/20... Training Step: 1987... Training loss: 1.7621... 0.0754 sec/batch
Epoch: 3/20... Training Step: 1988... Training loss: 1.7105... 0.0753 sec/batch
Epoch: 3/20... Training Step: 1989... Training loss: 1.6114... 0.0753 sec/batch
Epoch: 3/20... Training Step: 1990... Training loss: 1.6213... 0.0750 sec/batch
Epoch: 3/20... Training Step: 1991... Training loss: 1.6527... 0.0756 sec/batch
Epoch: 3/20... Training Step: 1992... Training loss: 1.6952... 0.0771 sec/batch
Epoch: 3/20... Training Step: 1993... Training loss: 1.7084... 0.0753 sec/batch
Epoch: 3/20... Training Step: 1994... Training loss: 1.6119... 0.0741 sec/batch
Epoch: 3/20... Training Step: 1995... Training loss: 1.7471... 0.0734 sec/batch
Epoch: 3/20... Training Step: 1996... Training loss: 1.6308... 0.0743 sec/batch
Epoch: 3/20... Training Step: 1997... Training loss: 1.6617... 0.0746 sec/batch
Epoch: 3/20... Training Step: 1998... Training loss: 1.6891... 0.0735 sec/batch
Epoch: 3/20... Training Step: 1999... Training loss: 1.7101... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2000... Training loss: 1.6455... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2001... Training loss: 1.6452... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2002... Training loss: 1.6692... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2003... Training loss: 1.6843... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2004... Training loss: 1.6583... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2005... Training loss: 1.6989... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2006... Training loss: 1.6489... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2007... Training loss: 1.6358... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2008... Training loss: 1.6242... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2009... Training loss: 1.7056... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2010... Training loss: 1.6254... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2011... Training loss: 1.6959... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2012... Training loss: 1.6013... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2013... Training loss: 1.6953... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2014... Training loss: 1.7499... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2015... Training loss: 1.7079... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2016... Training loss: 1.7284... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2017... Training loss: 1.6969... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2018... Training loss: 1.7381... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2019... Training loss: 1.7759... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2020... Training loss: 1.7568... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2021... Training loss: 1.6860... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2022... Training loss: 1.7650... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2023... Training loss: 1.6898... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2024... Training loss: 1.6549... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2025... Training loss: 1.6433... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2026... Training loss: 1.6548... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2027... Training loss: 1.6704... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2028... Training loss: 1.6629... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2029... Training loss: 1.7102... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2030... Training loss: 1.6398... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2031... Training loss: 1.7131... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2032... Training loss: 1.6429... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2033... Training loss: 1.7007... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2034... Training loss: 1.7089... 0.0774 sec/batch
Epoch: 3/20... Training Step: 2035... Training loss: 1.5692... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2036... Training loss: 1.7132... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2037... Training loss: 1.5982... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2038... Training loss: 1.6306... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2039... Training loss: 1.6377... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2040... Training loss: 1.6291... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2041... Training loss: 1.6796... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2042... Training loss: 1.7206... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2043... Training loss: 1.6703... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2044... Training loss: 1.6692... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2045... Training loss: 1.5719... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2046... Training loss: 1.6261... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2047... Training loss: 1.6743... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2048... Training loss: 1.5931... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2049... Training loss: 1.6923... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2050... Training loss: 1.6638... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2051... Training loss: 1.6168... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2052... Training loss: 1.5999... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2053... Training loss: 1.6691... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2054... Training loss: 1.6969... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2055... Training loss: 1.7042... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2056... Training loss: 1.7272... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2057... Training loss: 1.6573... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2058... Training loss: 1.5913... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2059... Training loss: 1.6063... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2060... Training loss: 1.5888... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2061... Training loss: 1.6489... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2062... Training loss: 1.6559... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2063... Training loss: 1.7179... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2064... Training loss: 1.7219... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2065... Training loss: 1.6957... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2066... Training loss: 1.6477... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2067... Training loss: 1.6783... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2068... Training loss: 1.6894... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2069... Training loss: 1.6339... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2070... Training loss: 1.7254... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2071... Training loss: 1.6364... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2072... Training loss: 1.7024... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2073... Training loss: 1.6807... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2074... Training loss: 1.7214... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2075... Training loss: 1.7836... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2076... Training loss: 1.6316... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2077... Training loss: 1.6013... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2078... Training loss: 1.5474... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2079... Training loss: 1.5962... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2080... Training loss: 1.7216... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2081... Training loss: 1.6696... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2082... Training loss: 1.5973... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2083... Training loss: 1.6227... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2084... Training loss: 1.6206... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2085... Training loss: 1.6186... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2086... Training loss: 1.6428... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2087... Training loss: 1.6650... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2088... Training loss: 1.7322... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2089... Training loss: 1.6871... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2090... Training loss: 1.6570... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2091... Training loss: 1.6301... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2092... Training loss: 1.7289... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2093... Training loss: 1.6576... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2094... Training loss: 1.6970... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2095... Training loss: 1.6426... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2096... Training loss: 1.6374... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2097... Training loss: 1.7706... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2098... Training loss: 1.6748... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2099... Training loss: 1.6084... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2100... Training loss: 1.7886... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2101... Training loss: 1.6955... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2102... Training loss: 1.7461... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2103... Training loss: 1.6493... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2104... Training loss: 1.6688... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2105... Training loss: 1.7603... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2106... Training loss: 1.6542... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2107... Training loss: 1.6714... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2108... Training loss: 1.6363... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2109... Training loss: 1.7288... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2110... Training loss: 1.7499... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2111... Training loss: 1.7313... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2112... Training loss: 1.6782... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2113... Training loss: 1.6566... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2114... Training loss: 1.6615... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2115... Training loss: 1.6549... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2116... Training loss: 1.7456... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2117... Training loss: 1.6870... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2118... Training loss: 1.6505... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2119... Training loss: 1.6821... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2120... Training loss: 1.6755... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2121... Training loss: 1.6923... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2122... Training loss: 1.7661... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2123... Training loss: 1.7329... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2124... Training loss: 1.6833... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2125... Training loss: 1.6458... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2126... Training loss: 1.6366... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2127... Training loss: 1.6316... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2128... Training loss: 1.6547... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2129... Training loss: 1.5600... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2130... Training loss: 1.7335... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2131... Training loss: 1.6482... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2132... Training loss: 1.5714... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2133... Training loss: 1.6491... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2134... Training loss: 1.6294... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2135... Training loss: 1.5598... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2136... Training loss: 1.6720... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2137... Training loss: 1.6209... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2138... Training loss: 1.6340... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2139... Training loss: 1.6129... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2140... Training loss: 1.6326... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2141... Training loss: 1.6471... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2142... Training loss: 1.6306... 0.0781 sec/batch
Epoch: 3/20... Training Step: 2143... Training loss: 1.6572... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2144... Training loss: 1.6811... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2145... Training loss: 1.6168... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2146... Training loss: 1.5938... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2147... Training loss: 1.6352... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2148... Training loss: 1.7034... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2149... Training loss: 1.6759... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2150... Training loss: 1.7454... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2151... Training loss: 1.7064... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2152... Training loss: 1.6652... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2153... Training loss: 1.7001... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2154... Training loss: 1.6818... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2155... Training loss: 1.7237... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2156... Training loss: 1.6203... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2157... Training loss: 1.6520... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2158... Training loss: 1.7031... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2159... Training loss: 1.6554... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2160... Training loss: 1.6794... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2161... Training loss: 1.7196... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2162... Training loss: 1.7615... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2163... Training loss: 1.7398... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2164... Training loss: 1.6527... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2165... Training loss: 1.6735... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2166... Training loss: 1.7411... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2167... Training loss: 1.6290... 0.0804 sec/batch
Epoch: 3/20... Training Step: 2168... Training loss: 1.6712... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2169... Training loss: 1.6996... 0.0775 sec/batch
Epoch: 3/20... Training Step: 2170... Training loss: 1.6995... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2171... Training loss: 1.7102... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2172... Training loss: 1.6523... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2173... Training loss: 1.6491... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2174... Training loss: 1.7148... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2175... Training loss: 1.7277... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2176... Training loss: 1.7468... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2177... Training loss: 1.7245... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2178... Training loss: 1.7800... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2179... Training loss: 1.6667... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2180... Training loss: 1.6108... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2181... Training loss: 1.7236... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2182... Training loss: 1.6284... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2183... Training loss: 1.6860... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2184... Training loss: 1.6930... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2185... Training loss: 1.6573... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2186... Training loss: 1.6545... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2187... Training loss: 1.6470... 0.0781 sec/batch
Epoch: 3/20... Training Step: 2188... Training loss: 1.6520... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2189... Training loss: 1.6749... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2190... Training loss: 1.7290... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2191... Training loss: 1.6441... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2192... Training loss: 1.6214... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2193... Training loss: 1.7038... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2194... Training loss: 1.6539... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2195... Training loss: 1.6183... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2196... Training loss: 1.7542... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2197... Training loss: 1.6521... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2198... Training loss: 1.6706... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2199... Training loss: 1.6478... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2200... Training loss: 1.6691... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2201... Training loss: 1.6264... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2202... Training loss: 1.6864... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2203... Training loss: 1.6874... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2204... Training loss: 1.6623... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2205... Training loss: 1.6107... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2206... Training loss: 1.7448... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2207... Training loss: 1.6643... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2208... Training loss: 1.6541... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2209... Training loss: 1.7223... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2210... Training loss: 1.6390... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2211... Training loss: 1.6596... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2212... Training loss: 1.6634... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2213... Training loss: 1.5690... 0.0777 sec/batch
Epoch: 3/20... Training Step: 2214... Training loss: 1.6450... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2215... Training loss: 1.6510... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2216... Training loss: 1.6703... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2217... Training loss: 1.6577... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2218... Training loss: 1.7587... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2219... Training loss: 1.6292... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2220... Training loss: 1.6745... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2221... Training loss: 1.6005... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2222... Training loss: 1.6932... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2223... Training loss: 1.6907... 0.0877 sec/batch
Epoch: 3/20... Training Step: 2224... Training loss: 1.6357... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2225... Training loss: 1.6581... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2226... Training loss: 1.6406... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2227... Training loss: 1.6669... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2228... Training loss: 1.7191... 0.0725 sec/batch
Epoch: 3/20... Training Step: 2229... Training loss: 1.6303... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2230... Training loss: 1.6938... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2231... Training loss: 1.6221... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2232... Training loss: 1.5828... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2233... Training loss: 1.6181... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2234... Training loss: 1.6536... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2235... Training loss: 1.6258... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2236... Training loss: 1.6430... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2237... Training loss: 1.6199... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2238... Training loss: 1.6512... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2239... Training loss: 1.6271... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2240... Training loss: 1.6442... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2241... Training loss: 1.7412... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2242... Training loss: 1.6234... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2243... Training loss: 1.6284... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2244... Training loss: 1.6538... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2245... Training loss: 1.7160... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2246... Training loss: 1.7091... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2247... Training loss: 1.6667... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2248... Training loss: 1.6750... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2249... Training loss: 1.6744... 0.0782 sec/batch
Epoch: 3/20... Training Step: 2250... Training loss: 1.6311... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2251... Training loss: 1.5948... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2252... Training loss: 1.6371... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2253... Training loss: 1.6174... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2254... Training loss: 1.6056... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2255... Training loss: 1.6388... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2256... Training loss: 1.6220... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2257... Training loss: 1.5564... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2258... Training loss: 1.5637... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2259... Training loss: 1.6195... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2260... Training loss: 1.6458... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2261... Training loss: 1.6369... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2262... Training loss: 1.5478... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2263... Training loss: 1.6046... 0.0796 sec/batch
Epoch: 3/20... Training Step: 2264... Training loss: 1.7174... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2265... Training loss: 1.6564... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2266... Training loss: 1.6455... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2267... Training loss: 1.6871... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2268... Training loss: 1.6119... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2269... Training loss: 1.5793... 0.0730 sec/batch
Epoch: 3/20... Training Step: 2270... Training loss: 1.6244... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2271... Training loss: 1.6753... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2272... Training loss: 1.6089... 0.0730 sec/batch
Epoch: 3/20... Training Step: 2273... Training loss: 1.5998... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2274... Training loss: 1.6409... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2275... Training loss: 1.6317... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2276... Training loss: 1.6290... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2277... Training loss: 1.5777... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2278... Training loss: 1.5517... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2279... Training loss: 1.6502... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2280... Training loss: 1.6236... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2281... Training loss: 1.5970... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2282... Training loss: 1.7079... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2283... Training loss: 1.6742... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2284... Training loss: 1.6943... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2285... Training loss: 1.6694... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2286... Training loss: 1.6046... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2287... Training loss: 1.6287... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2288... Training loss: 1.6849... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2289... Training loss: 1.6429... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2290... Training loss: 1.7221... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2291... Training loss: 1.6097... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2292... Training loss: 1.5584... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2293... Training loss: 1.7196... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2294... Training loss: 1.6241... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2295... Training loss: 1.7002... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2296... Training loss: 1.6750... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2297... Training loss: 1.6252... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2298... Training loss: 1.6302... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2299... Training loss: 1.6512... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2300... Training loss: 1.5445... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2301... Training loss: 1.5675... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2302... Training loss: 1.5911... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2303... Training loss: 1.5805... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2304... Training loss: 1.6622... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2305... Training loss: 1.6756... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2306... Training loss: 1.5722... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2307... Training loss: 1.5763... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2308... Training loss: 1.6246... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2309... Training loss: 1.6691... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2310... Training loss: 1.6045... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2311... Training loss: 1.6128... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2312... Training loss: 1.5946... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2313... Training loss: 1.6259... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2314... Training loss: 1.6384... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2315... Training loss: 1.6621... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2316... Training loss: 1.6110... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2317... Training loss: 1.7284... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2318... Training loss: 1.7033... 0.0780 sec/batch
Epoch: 3/20... Training Step: 2319... Training loss: 1.7040... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2320... Training loss: 1.6781... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2321... Training loss: 1.6619... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2322... Training loss: 1.5907... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2323... Training loss: 1.6547... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2324... Training loss: 1.6212... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2325... Training loss: 1.6559... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2326... Training loss: 1.6166... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2327... Training loss: 1.5666... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2328... Training loss: 1.6762... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2329... Training loss: 1.6174... 0.0775 sec/batch
Epoch: 3/20... Training Step: 2330... Training loss: 1.6756... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2331... Training loss: 1.6705... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2332... Training loss: 1.6481... 0.0778 sec/batch
Epoch: 3/20... Training Step: 2333... Training loss: 1.6897... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2334... Training loss: 1.6948... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2335... Training loss: 1.6452... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2336... Training loss: 1.5774... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2337... Training loss: 1.6347... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2338... Training loss: 1.6127... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2339... Training loss: 1.6353... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2340... Training loss: 1.6350... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2341... Training loss: 1.6225... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2342... Training loss: 1.6175... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2343... Training loss: 1.6653... 0.0778 sec/batch
Epoch: 3/20... Training Step: 2344... Training loss: 1.6164... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2345... Training loss: 1.6663... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2346... Training loss: 1.7052... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2347... Training loss: 1.6470... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2348... Training loss: 1.6170... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2349... Training loss: 1.6216... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2350... Training loss: 1.6114... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2351... Training loss: 1.6192... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2352... Training loss: 1.6118... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2353... Training loss: 1.5616... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2354... Training loss: 1.5562... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2355... Training loss: 1.6740... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2356... Training loss: 1.5681... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2357... Training loss: 1.6298... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2358... Training loss: 1.5781... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2359... Training loss: 1.6318... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2360... Training loss: 1.5719... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2361... Training loss: 1.5976... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2362... Training loss: 1.6356... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2363... Training loss: 1.6293... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2364... Training loss: 1.6628... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2365... Training loss: 1.6369... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2366... Training loss: 1.6115... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2367... Training loss: 1.6434... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2368... Training loss: 1.6542... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2369... Training loss: 1.6479... 0.0781 sec/batch
Epoch: 3/20... Training Step: 2370... Training loss: 1.6266... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2371... Training loss: 1.6764... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2372... Training loss: 1.7119... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2373... Training loss: 1.6681... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2374... Training loss: 1.6192... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2375... Training loss: 1.6585... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2376... Training loss: 1.6175... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2377... Training loss: 1.6536... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2378... Training loss: 1.5881... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2379... Training loss: 1.6526... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2380... Training loss: 1.6143... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2381... Training loss: 1.5974... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2382... Training loss: 1.6926... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2383... Training loss: 1.6383... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2384... Training loss: 1.6073... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2385... Training loss: 1.5824... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2386... Training loss: 1.6671... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2387... Training loss: 1.5995... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2388... Training loss: 1.6736... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2389... Training loss: 1.5743... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2390... Training loss: 1.6283... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2391... Training loss: 1.5776... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2392... Training loss: 1.7221... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2393... Training loss: 1.5839... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2394... Training loss: 1.5905... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2395... Training loss: 1.5480... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2396... Training loss: 1.5601... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2397... Training loss: 1.6260... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2398... Training loss: 1.5505... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2399... Training loss: 1.5758... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2400... Training loss: 1.6618... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2401... Training loss: 1.5489... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2402... Training loss: 1.5970... 0.0801 sec/batch
Epoch: 3/20... Training Step: 2403... Training loss: 1.6471... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2404... Training loss: 1.7319... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2405... Training loss: 1.5945... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2406... Training loss: 1.6123... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2407... Training loss: 1.6015... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2408... Training loss: 1.6153... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2409... Training loss: 1.6344... 0.0729 sec/batch
Epoch: 3/20... Training Step: 2410... Training loss: 1.6275... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2411... Training loss: 1.6797... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2412... Training loss: 1.7183... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2413... Training loss: 1.6656... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2414... Training loss: 1.7234... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2415... Training loss: 1.6232... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2416... Training loss: 1.6818... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2417... Training loss: 1.7005... 0.0780 sec/batch
Epoch: 3/20... Training Step: 2418... Training loss: 1.6484... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2419... Training loss: 1.6815... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2420... Training loss: 1.6177... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2421... Training loss: 1.6961... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2422... Training loss: 1.7475... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2423... Training loss: 1.6172... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2424... Training loss: 1.6655... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2425... Training loss: 1.6233... 0.0726 sec/batch
Epoch: 3/20... Training Step: 2426... Training loss: 1.6675... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2427... Training loss: 1.6430... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2428... Training loss: 1.6618... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2429... Training loss: 1.6099... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2430... Training loss: 1.7013... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2431... Training loss: 1.5949... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2432... Training loss: 1.6324... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2433... Training loss: 1.6856... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2434... Training loss: 1.7140... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2435... Training loss: 1.6343... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2436... Training loss: 1.6367... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2437... Training loss: 1.6518... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2438... Training loss: 1.6917... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2439... Training loss: 1.6262... 0.0783 sec/batch
Epoch: 3/20... Training Step: 2440... Training loss: 1.6498... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2441... Training loss: 1.5518... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2442... Training loss: 1.6155... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2443... Training loss: 1.6477... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2444... Training loss: 1.6548... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2445... Training loss: 1.6585... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2446... Training loss: 1.6150... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2447... Training loss: 1.5790... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2448... Training loss: 1.5916... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2449... Training loss: 1.5691... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2450... Training loss: 1.5933... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2451... Training loss: 1.5401... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2452... Training loss: 1.6378... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2453... Training loss: 1.5987... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2454... Training loss: 1.7132... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2455... Training loss: 1.6005... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2456... Training loss: 1.6040... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2457... Training loss: 1.5907... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2458... Training loss: 1.6194... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2459... Training loss: 1.4872... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2460... Training loss: 1.5981... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2461... Training loss: 1.6026... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2462... Training loss: 1.5697... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2463... Training loss: 1.5799... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2464... Training loss: 1.5744... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2465... Training loss: 1.5552... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2466... Training loss: 1.6331... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2467... Training loss: 1.6204... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2468... Training loss: 1.6157... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2469... Training loss: 1.5590... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2470... Training loss: 1.5918... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2471... Training loss: 1.6402... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2472... Training loss: 1.5875... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2473... Training loss: 1.5653... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2474... Training loss: 1.5484... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2475... Training loss: 1.5665... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2476... Training loss: 1.5719... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2477... Training loss: 1.6096... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2478... Training loss: 1.6100... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2479... Training loss: 1.5519... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2480... Training loss: 1.5556... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2481... Training loss: 1.6402... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2482... Training loss: 1.6071... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2483... Training loss: 1.6199... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2484... Training loss: 1.6158... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2485... Training loss: 1.6024... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2486... Training loss: 1.6491... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2487... Training loss: 1.6118... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2488... Training loss: 1.6110... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2489... Training loss: 1.6428... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2490... Training loss: 1.5484... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2491... Training loss: 1.6073... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2492... Training loss: 1.6036... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2493... Training loss: 1.5820... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2494... Training loss: 1.5593... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2495... Training loss: 1.6274... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2496... Training loss: 1.6325... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2497... Training loss: 1.6756... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2498... Training loss: 1.6716... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2499... Training loss: 1.6215... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2500... Training loss: 1.5957... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2501... Training loss: 1.6190... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2502... Training loss: 1.5981... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2503... Training loss: 1.5947... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2504... Training loss: 1.5657... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2505... Training loss: 1.6427... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2506... Training loss: 1.6004... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2507... Training loss: 1.6739... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2508... Training loss: 1.5522... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2509... Training loss: 1.6018... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2510... Training loss: 1.6146... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2511... Training loss: 1.6217... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2512... Training loss: 1.5469... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2513... Training loss: 1.5845... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2514... Training loss: 1.6857... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2515... Training loss: 1.7246... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2516... Training loss: 1.6277... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2517... Training loss: 1.6503... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2518... Training loss: 1.6607... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2519... Training loss: 1.6135... 0.0782 sec/batch
Epoch: 3/20... Training Step: 2520... Training loss: 1.5880... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2521... Training loss: 1.6083... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2522... Training loss: 1.5864... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2523... Training loss: 1.5560... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2524... Training loss: 1.5648... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2525... Training loss: 1.7423... 0.0789 sec/batch
Epoch: 3/20... Training Step: 2526... Training loss: 1.5674... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2527... Training loss: 1.5771... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2528... Training loss: 1.5850... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2529... Training loss: 1.5988... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2530... Training loss: 1.5826... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2531... Training loss: 1.6863... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2532... Training loss: 1.6640... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2533... Training loss: 1.5722... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2534... Training loss: 1.5565... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2535... Training loss: 1.6039... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2536... Training loss: 1.5452... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2537... Training loss: 1.5387... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2538... Training loss: 1.6197... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2539... Training loss: 1.6058... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2540... Training loss: 1.5520... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2541... Training loss: 1.5322... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2542... Training loss: 1.6188... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2543... Training loss: 1.6134... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2544... Training loss: 1.5112... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2545... Training loss: 1.6006... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2546... Training loss: 1.6193... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2547... Training loss: 1.5315... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2548... Training loss: 1.5668... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2549... Training loss: 1.5953... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2550... Training loss: 1.6379... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2551... Training loss: 1.6068... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2552... Training loss: 1.5623... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2553... Training loss: 1.5154... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2554... Training loss: 1.5381... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2555... Training loss: 1.5451... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2556... Training loss: 1.5557... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2557... Training loss: 1.6701... 0.0782 sec/batch
Epoch: 3/20... Training Step: 2558... Training loss: 1.6207... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2559... Training loss: 1.6118... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2560... Training loss: 1.5993... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2561... Training loss: 1.6062... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2562... Training loss: 1.7108... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2563... Training loss: 1.6447... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2564... Training loss: 1.5593... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2565... Training loss: 1.5358... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2566... Training loss: 1.5649... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2567... Training loss: 1.6694... 0.0730 sec/batch
Epoch: 3/20... Training Step: 2568... Training loss: 1.5565... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2569... Training loss: 1.5744... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2570... Training loss: 1.6218... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2571... Training loss: 1.6054... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2572... Training loss: 1.6438... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2573... Training loss: 1.5750... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2574... Training loss: 1.5923... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2575... Training loss: 1.5978... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2576... Training loss: 1.6718... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2577... Training loss: 1.5579... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2578... Training loss: 1.5659... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2579... Training loss: 1.5371... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2580... Training loss: 1.6247... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2581... Training loss: 1.4919... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2582... Training loss: 1.4662... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2583... Training loss: 1.5700... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2584... Training loss: 1.5780... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2585... Training loss: 1.6530... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2586... Training loss: 1.5744... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2587... Training loss: 1.5768... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2588... Training loss: 1.5964... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2589... Training loss: 1.5561... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2590... Training loss: 1.5863... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2591... Training loss: 1.6584... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2592... Training loss: 1.6391... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2593... Training loss: 1.6265... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2594... Training loss: 1.6312... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2595... Training loss: 1.5087... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2596... Training loss: 1.5948... 0.0731 sec/batch
Epoch: 3/20... Training Step: 2597... Training loss: 1.6593... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2598... Training loss: 1.5434... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2599... Training loss: 1.6592... 0.0791 sec/batch
Epoch: 3/20... Training Step: 2600... Training loss: 1.5919... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2601... Training loss: 1.6102... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2602... Training loss: 1.6565... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2603... Training loss: 1.5433... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2604... Training loss: 1.6056... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2605... Training loss: 1.5495... 0.0774 sec/batch
Epoch: 3/20... Training Step: 2606... Training loss: 1.5125... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2607... Training loss: 1.6365... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2608... Training loss: 1.5996... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2609... Training loss: 1.5699... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2610... Training loss: 1.5473... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2611... Training loss: 1.5763... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2612... Training loss: 1.6308... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2613... Training loss: 1.4309... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2614... Training loss: 1.5987... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2615... Training loss: 1.5696... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2616... Training loss: 1.5721... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2617... Training loss: 1.5291... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2618... Training loss: 1.5926... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2619... Training loss: 1.5985... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2620... Training loss: 1.5811... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2621... Training loss: 1.6621... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2622... Training loss: 1.5454... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2623... Training loss: 1.6363... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2624... Training loss: 1.6694... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2625... Training loss: 1.6731... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2626... Training loss: 1.6298... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2627... Training loss: 1.6048... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2628... Training loss: 1.5952... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2629... Training loss: 1.5744... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2630... Training loss: 1.6234... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2631... Training loss: 1.6263... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2632... Training loss: 1.5213... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2633... Training loss: 1.5750... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2634... Training loss: 1.5666... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2635... Training loss: 1.5887... 0.0775 sec/batch
Epoch: 3/20... Training Step: 2636... Training loss: 1.6339... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2637... Training loss: 1.5950... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2638... Training loss: 1.6592... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2639... Training loss: 1.6621... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2640... Training loss: 1.6773... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2641... Training loss: 1.5438... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2642... Training loss: 1.5775... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2643... Training loss: 1.5942... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2644... Training loss: 1.6559... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2645... Training loss: 1.6536... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2646... Training loss: 1.6748... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2647... Training loss: 1.6119... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2648... Training loss: 1.7118... 0.0771 sec/batch
Epoch: 3/20... Training Step: 2649... Training loss: 1.7131... 0.0777 sec/batch
Epoch: 3/20... Training Step: 2650... Training loss: 1.5532... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2651... Training loss: 1.6687... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2652... Training loss: 1.6035... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2653... Training loss: 1.5917... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2654... Training loss: 1.5658... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2655... Training loss: 1.6195... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2656... Training loss: 1.6056... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2657... Training loss: 1.5674... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2658... Training loss: 1.6073... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2659... Training loss: 1.5458... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2660... Training loss: 1.5789... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2661... Training loss: 1.7131... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2662... Training loss: 1.5521... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2663... Training loss: 1.5462... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2664... Training loss: 1.6213... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2665... Training loss: 1.6161... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2666... Training loss: 1.5813... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2667... Training loss: 1.6078... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2668... Training loss: 1.5473... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2669... Training loss: 1.6433... 0.0782 sec/batch
Epoch: 3/20... Training Step: 2670... Training loss: 1.6385... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2671... Training loss: 1.5612... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2672... Training loss: 1.5996... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2673... Training loss: 1.5544... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2674... Training loss: 1.6250... 0.0774 sec/batch
Epoch: 3/20... Training Step: 2675... Training loss: 1.6149... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2676... Training loss: 1.6034... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2677... Training loss: 1.5418... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2678... Training loss: 1.5488... 0.0783 sec/batch
Epoch: 3/20... Training Step: 2679... Training loss: 1.6107... 0.0779 sec/batch
Epoch: 3/20... Training Step: 2680... Training loss: 1.5696... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2681... Training loss: 1.6600... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2682... Training loss: 1.6118... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2683... Training loss: 1.5737... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2684... Training loss: 1.6279... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2685... Training loss: 1.5915... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2686... Training loss: 1.5781... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2687... Training loss: 1.6164... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2688... Training loss: 1.5205... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2689... Training loss: 1.5046... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2690... Training loss: 1.5411... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2691... Training loss: 1.6167... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2692... Training loss: 1.6182... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2693... Training loss: 1.6400... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2694... Training loss: 1.5770... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2695... Training loss: 1.5753... 0.0758 sec/batch
Epoch: 3/20... Training Step: 2696... Training loss: 1.6768... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2697... Training loss: 1.5330... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2698... Training loss: 1.5715... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2699... Training loss: 1.5891... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2700... Training loss: 1.6257... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2701... Training loss: 1.6416... 0.0728 sec/batch
Epoch: 3/20... Training Step: 2702... Training loss: 1.6021... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2703... Training loss: 1.6443... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2704... Training loss: 1.7008... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2705... Training loss: 1.6165... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2706... Training loss: 1.5816... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2707... Training loss: 1.5865... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2708... Training loss: 1.6354... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2709... Training loss: 1.6408... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2710... Training loss: 1.7036... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2711... Training loss: 1.5683... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2712... Training loss: 1.6404... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2713... Training loss: 1.6285... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2714... Training loss: 1.5741... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2715... Training loss: 1.5625... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2716... Training loss: 1.5993... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2717... Training loss: 1.5403... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2718... Training loss: 1.6495... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2719... Training loss: 1.5508... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2720... Training loss: 1.6136... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2721... Training loss: 1.5368... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2722... Training loss: 1.6340... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2723... Training loss: 1.5832... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2724... Training loss: 1.5846... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2725... Training loss: 1.5635... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2726... Training loss: 1.5656... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2727... Training loss: 1.5273... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2728... Training loss: 1.6671... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2729... Training loss: 1.6217... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2730... Training loss: 1.5649... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2731... Training loss: 1.6137... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2732... Training loss: 1.6123... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2733... Training loss: 1.5739... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2734... Training loss: 1.5369... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2735... Training loss: 1.6959... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2736... Training loss: 1.5431... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2737... Training loss: 1.6093... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2738... Training loss: 1.5938... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2739... Training loss: 1.5303... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2740... Training loss: 1.5923... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2741... Training loss: 1.5864... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2742... Training loss: 1.6658... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2743... Training loss: 1.6423... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2744... Training loss: 1.6840... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2745... Training loss: 1.6318... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2746... Training loss: 1.5207... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2747... Training loss: 1.6095... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2748... Training loss: 1.5867... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2749... Training loss: 1.6033... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2750... Training loss: 1.6016... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2751... Training loss: 1.5838... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2752... Training loss: 1.5580... 0.0771 sec/batch
Epoch: 3/20... Training Step: 2753... Training loss: 1.5528... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2754... Training loss: 1.5907... 0.0775 sec/batch
Epoch: 3/20... Training Step: 2755... Training loss: 1.5834... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2756... Training loss: 1.6637... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2757... Training loss: 1.6224... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2758... Training loss: 1.6541... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2759... Training loss: 1.5999... 0.0733 sec/batch
Epoch: 3/20... Training Step: 2760... Training loss: 1.5121... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2761... Training loss: 1.5469... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2762... Training loss: 1.6081... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2763... Training loss: 1.5944... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2764... Training loss: 1.5548... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2765... Training loss: 1.5757... 0.0779 sec/batch
Epoch: 3/20... Training Step: 2766... Training loss: 1.5793... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2767... Training loss: 1.6352... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2768... Training loss: 1.6108... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2769... Training loss: 1.5999... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2770... Training loss: 1.5772... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2771... Training loss: 1.5837... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2772... Training loss: 1.6051... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2773... Training loss: 1.6069... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2774... Training loss: 1.5916... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2775... Training loss: 1.5695... 0.0777 sec/batch
Epoch: 3/20... Training Step: 2776... Training loss: 1.5570... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2777... Training loss: 1.5782... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2778... Training loss: 1.5494... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2779... Training loss: 1.4982... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2780... Training loss: 1.5909... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2781... Training loss: 1.5561... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2782... Training loss: 1.5995... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2783... Training loss: 1.5775... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2784... Training loss: 1.5070... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2785... Training loss: 1.6346... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2786... Training loss: 1.6321... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2787... Training loss: 1.6383... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2788... Training loss: 1.5958... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2789... Training loss: 1.5257... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2790... Training loss: 1.6003... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2791... Training loss: 1.5497... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2792... Training loss: 1.6070... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2793... Training loss: 1.6578... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2794... Training loss: 1.6611... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2795... Training loss: 1.6215... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2796... Training loss: 1.6483... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2797... Training loss: 1.6089... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2798... Training loss: 1.6344... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2799... Training loss: 1.6428... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2800... Training loss: 1.6954... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2801... Training loss: 1.6811... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2802... Training loss: 1.6960... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2803... Training loss: 1.6308... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2804... Training loss: 1.6180... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2805... Training loss: 1.6232... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2806... Training loss: 1.4939... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2807... Training loss: 1.6125... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2808... Training loss: 1.5994... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2809... Training loss: 1.5711... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2810... Training loss: 1.6060... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2811... Training loss: 1.6151... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2812... Training loss: 1.6660... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2813... Training loss: 1.6179... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2814... Training loss: 1.5599... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2815... Training loss: 1.6940... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2816... Training loss: 1.7026... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2817... Training loss: 1.7112... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2818... Training loss: 1.6143... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2819... Training loss: 1.5003... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2820... Training loss: 1.5839... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2821... Training loss: 1.6549... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2822... Training loss: 1.6123... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2823... Training loss: 1.6734... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2824... Training loss: 1.5665... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2825... Training loss: 1.6679... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2826... Training loss: 1.7343... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2827... Training loss: 1.6418... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2828... Training loss: 1.6712... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2829... Training loss: 1.6776... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2830... Training loss: 1.5577... 0.0763 sec/batch
Epoch: 3/20... Training Step: 2831... Training loss: 1.6007... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2832... Training loss: 1.5982... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2833... Training loss: 1.5959... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2834... Training loss: 1.6179... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2835... Training loss: 1.5839... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2836... Training loss: 1.6723... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2837... Training loss: 1.6501... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2838... Training loss: 1.6111... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2839... Training loss: 1.6078... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2840... Training loss: 1.5825... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2841... Training loss: 1.6020... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2842... Training loss: 1.6020... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2843... Training loss: 1.6526... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2844... Training loss: 1.6525... 0.0768 sec/batch
Epoch: 3/20... Training Step: 2845... Training loss: 1.6104... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2846... Training loss: 1.5532... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2847... Training loss: 1.5701... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2848... Training loss: 1.5913... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2849... Training loss: 1.5412... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2850... Training loss: 1.5920... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2851... Training loss: 1.5895... 0.0760 sec/batch
Epoch: 3/20... Training Step: 2852... Training loss: 1.7050... 0.0776 sec/batch
Epoch: 3/20... Training Step: 2853... Training loss: 1.6065... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2854... Training loss: 1.6690... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2855... Training loss: 1.6311... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2856... Training loss: 1.6156... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2857... Training loss: 1.6515... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2858... Training loss: 1.6057... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2859... Training loss: 1.6026... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2860... Training loss: 1.6396... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2861... Training loss: 1.6131... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2862... Training loss: 1.6665... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2863... Training loss: 1.5856... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2864... Training loss: 1.5758... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2865... Training loss: 1.6004... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2866... Training loss: 1.5336... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2867... Training loss: 1.6124... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2868... Training loss: 1.5977... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2869... Training loss: 1.6189... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2870... Training loss: 1.6689... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2871... Training loss: 1.6006... 0.0794 sec/batch
Epoch: 3/20... Training Step: 2872... Training loss: 1.6917... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2873... Training loss: 1.6326... 0.0775 sec/batch
Epoch: 3/20... Training Step: 2874... Training loss: 1.6836... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2875... Training loss: 1.6463... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2876... Training loss: 1.5189... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2877... Training loss: 1.5770... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2878... Training loss: 1.5244... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2879... Training loss: 1.6625... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2880... Training loss: 1.6069... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2881... Training loss: 1.5849... 0.0745 sec/batch
Epoch: 3/20... Training Step: 2882... Training loss: 1.6335... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2883... Training loss: 1.5802... 0.0787 sec/batch
Epoch: 3/20... Training Step: 2884... Training loss: 1.5647... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2885... Training loss: 1.5725... 0.0779 sec/batch
Epoch: 3/20... Training Step: 2886... Training loss: 1.5271... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2887... Training loss: 1.5274... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2888... Training loss: 1.5215... 0.0748 sec/batch
Epoch: 3/20... Training Step: 2889... Training loss: 1.5477... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2890... Training loss: 1.5881... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2891... Training loss: 1.5284... 0.0755 sec/batch
Epoch: 3/20... Training Step: 2892... Training loss: 1.6279... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2893... Training loss: 1.5436... 0.0769 sec/batch
Epoch: 3/20... Training Step: 2894... Training loss: 1.4466... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2895... Training loss: 1.5235... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2896... Training loss: 1.5820... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2897... Training loss: 1.6530... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2898... Training loss: 1.6037... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2899... Training loss: 1.5759... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2900... Training loss: 1.5601... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2901... Training loss: 1.5959... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2902... Training loss: 1.6302... 0.0762 sec/batch
Epoch: 3/20... Training Step: 2903... Training loss: 1.5726... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2904... Training loss: 1.6213... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2905... Training loss: 1.6929... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2906... Training loss: 1.6418... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2907... Training loss: 1.6196... 0.0773 sec/batch
Epoch: 3/20... Training Step: 2908... Training loss: 1.5990... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2909... Training loss: 1.5940... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2910... Training loss: 1.6300... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2911... Training loss: 1.5964... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2912... Training loss: 1.6092... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2913... Training loss: 1.6938... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2914... Training loss: 1.6509... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2915... Training loss: 1.6038... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2916... Training loss: 1.6889... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2917... Training loss: 1.5665... 0.0782 sec/batch
Epoch: 3/20... Training Step: 2918... Training loss: 1.5985... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2919... Training loss: 1.5835... 0.0754 sec/batch
Epoch: 3/20... Training Step: 2920... Training loss: 1.6109... 0.0765 sec/batch
Epoch: 3/20... Training Step: 2921... Training loss: 1.5727... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2922... Training loss: 1.5942... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2923... Training loss: 1.5634... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2924... Training loss: 1.6584... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2925... Training loss: 1.6016... 0.0759 sec/batch
Epoch: 3/20... Training Step: 2926... Training loss: 1.6055... 0.0766 sec/batch
Epoch: 3/20... Training Step: 2927... Training loss: 1.6500... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2928... Training loss: 1.6093... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2929... Training loss: 1.6632... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2930... Training loss: 1.5921... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2931... Training loss: 1.5762... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2932... Training loss: 1.5350... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2933... Training loss: 1.5900... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2934... Training loss: 1.4762... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2935... Training loss: 1.6093... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2936... Training loss: 1.5311... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2937... Training loss: 1.5821... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2938... Training loss: 1.6065... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2939... Training loss: 1.6579... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2940... Training loss: 1.6327... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2941... Training loss: 1.5900... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2942... Training loss: 1.6226... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2943... Training loss: 1.5479... 0.0761 sec/batch
Epoch: 3/20... Training Step: 2944... Training loss: 1.5559... 0.0743 sec/batch
Epoch: 3/20... Training Step: 2945... Training loss: 1.6084... 0.0756 sec/batch
Epoch: 3/20... Training Step: 2946... Training loss: 1.6005... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2947... Training loss: 1.6253... 0.0764 sec/batch
Epoch: 3/20... Training Step: 2948... Training loss: 1.5663... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2949... Training loss: 1.6954... 0.0732 sec/batch
Epoch: 3/20... Training Step: 2950... Training loss: 1.6975... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2951... Training loss: 1.6323... 0.0780 sec/batch
Epoch: 3/20... Training Step: 2952... Training loss: 1.6608... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2953... Training loss: 1.6062... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2954... Training loss: 1.7868... 0.0753 sec/batch
Epoch: 3/20... Training Step: 2955... Training loss: 1.5526... 0.0772 sec/batch
Epoch: 3/20... Training Step: 2956... Training loss: 1.7114... 0.0744 sec/batch
Epoch: 3/20... Training Step: 2957... Training loss: 1.6960... 0.0770 sec/batch
Epoch: 3/20... Training Step: 2958... Training loss: 1.5703... 0.0738 sec/batch
Epoch: 3/20... Training Step: 2959... Training loss: 1.6372... 0.0741 sec/batch
Epoch: 3/20... Training Step: 2960... Training loss: 1.6767... 0.0734 sec/batch
Epoch: 3/20... Training Step: 2961... Training loss: 1.6585... 0.0740 sec/batch
Epoch: 3/20... Training Step: 2962... Training loss: 1.6120... 0.0751 sec/batch
Epoch: 3/20... Training Step: 2963... Training loss: 1.6451... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2964... Training loss: 1.6055... 0.0767 sec/batch
Epoch: 3/20... Training Step: 2965... Training loss: 1.5262... 0.0749 sec/batch
Epoch: 3/20... Training Step: 2966... Training loss: 1.7127... 0.0752 sec/batch
Epoch: 3/20... Training Step: 2967... Training loss: 1.6601... 0.0742 sec/batch
Epoch: 3/20... Training Step: 2968... Training loss: 1.5985... 0.0750 sec/batch
Epoch: 3/20... Training Step: 2969... Training loss: 1.5972... 0.0735 sec/batch
Epoch: 3/20... Training Step: 2970... Training loss: 1.5549... 0.0739 sec/batch
Epoch: 3/20... Training Step: 2971... Training loss: 1.5693... 0.0736 sec/batch
Epoch: 3/20... Training Step: 2972... Training loss: 1.5720... 0.0747 sec/batch
Epoch: 3/20... Training Step: 2973... Training loss: 1.5933... 0.0746 sec/batch
Epoch: 3/20... Training Step: 2974... Training loss: 1.5219... 0.0757 sec/batch
Epoch: 3/20... Training Step: 2975... Training loss: 1.5750... 0.0737 sec/batch
Epoch: 3/20... Training Step: 2976... Training loss: 1.6343... 0.0776 sec/batch
Epoch: 4/20... Training Step: 2977... Training loss: 1.6737... 0.0756 sec/batch
Epoch: 4/20... Training Step: 2978... Training loss: 1.5539... 0.0792 sec/batch
Epoch: 4/20... Training Step: 2979... Training loss: 1.6517... 0.0771 sec/batch
Epoch: 4/20... Training Step: 2980... Training loss: 1.5959... 0.0768 sec/batch
Epoch: 4/20... Training Step: 2981... Training loss: 1.4850... 0.0778 sec/batch
Epoch: 4/20... Training Step: 2982... Training loss: 1.4689... 0.0753 sec/batch
Epoch: 4/20... Training Step: 2983... Training loss: 1.5424... 0.0748 sec/batch
Epoch: 4/20... Training Step: 2984... Training loss: 1.5902... 0.0779 sec/batch
Epoch: 4/20... Training Step: 2985... Training loss: 1.5760... 0.0738 sec/batch
Epoch: 4/20... Training Step: 2986... Training loss: 1.5214... 0.0770 sec/batch
Epoch: 4/20... Training Step: 2987... Training loss: 1.6417... 0.0739 sec/batch
Epoch: 4/20... Training Step: 2988... Training loss: 1.5423... 0.0767 sec/batch
Epoch: 4/20... Training Step: 2989... Training loss: 1.5649... 0.0738 sec/batch
Epoch: 4/20... Training Step: 2990... Training loss: 1.5793... 0.0751 sec/batch
Epoch: 4/20... Training Step: 2991... Training loss: 1.5989... 0.0750 sec/batch
Epoch: 4/20... Training Step: 2992... Training loss: 1.5321... 0.0753 sec/batch
Epoch: 4/20... Training Step: 2993... Training loss: 1.5749... 0.0750 sec/batch
Epoch: 4/20... Training Step: 2994... Training loss: 1.5582... 0.0745 sec/batch
Epoch: 4/20... Training Step: 2995... Training loss: 1.5798... 0.0765 sec/batch
Epoch: 4/20... Training Step: 2996... Training loss: 1.5364... 0.0795 sec/batch
Epoch: 4/20... Training Step: 2997... Training loss: 1.5818... 0.0760 sec/batch
Epoch: 4/20... Training Step: 2998... Training loss: 1.5454... 0.0738 sec/batch
Epoch: 4/20... Training Step: 2999... Training loss: 1.5548... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3000... Training loss: 1.5076... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3001... Training loss: 1.5904... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3002... Training loss: 1.5207... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3003... Training loss: 1.5573... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3004... Training loss: 1.4940... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3005... Training loss: 1.5681... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3006... Training loss: 1.6133... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3007... Training loss: 1.6108... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3008... Training loss: 1.5993... 0.0729 sec/batch
Epoch: 4/20... Training Step: 3009... Training loss: 1.5805... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3010... Training loss: 1.6171... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3011... Training loss: 1.6688... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3012... Training loss: 1.6544... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3013... Training loss: 1.5450... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3014... Training loss: 1.6479... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3015... Training loss: 1.5806... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3016... Training loss: 1.5424... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3017... Training loss: 1.5650... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3018... Training loss: 1.5441... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3019... Training loss: 1.5365... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3020... Training loss: 1.5527... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3021... Training loss: 1.6189... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3022... Training loss: 1.5347... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3023... Training loss: 1.6289... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3024... Training loss: 1.5289... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3025... Training loss: 1.6125... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3026... Training loss: 1.5890... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3027... Training loss: 1.4928... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3028... Training loss: 1.6125... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3029... Training loss: 1.5086... 0.0733 sec/batch
Epoch: 4/20... Training Step: 3030... Training loss: 1.5423... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3031... Training loss: 1.5430... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3032... Training loss: 1.5375... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3033... Training loss: 1.5758... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3034... Training loss: 1.6245... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3035... Training loss: 1.5575... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3036... Training loss: 1.5785... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3037... Training loss: 1.4685... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3038... Training loss: 1.5033... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3039... Training loss: 1.5771... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3040... Training loss: 1.4859... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3041... Training loss: 1.5835... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3042... Training loss: 1.5363... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3043... Training loss: 1.5045... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3044... Training loss: 1.5264... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3045... Training loss: 1.6037... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3046... Training loss: 1.5839... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3047... Training loss: 1.6026... 0.0787 sec/batch
Epoch: 4/20... Training Step: 3048... Training loss: 1.6150... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3049... Training loss: 1.5622... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3050... Training loss: 1.4928... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3051... Training loss: 1.5122... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3052... Training loss: 1.4961... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3053... Training loss: 1.5374... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3054... Training loss: 1.5341... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3055... Training loss: 1.5853... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3056... Training loss: 1.6131... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3057... Training loss: 1.5855... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3058... Training loss: 1.5336... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3059... Training loss: 1.5837... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3060... Training loss: 1.5992... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3061... Training loss: 1.5306... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3062... Training loss: 1.6309... 0.0791 sec/batch
Epoch: 4/20... Training Step: 3063... Training loss: 1.5430... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3064... Training loss: 1.5872... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3065... Training loss: 1.5924... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3066... Training loss: 1.6138... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3067... Training loss: 1.6466... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3068... Training loss: 1.5562... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3069... Training loss: 1.5115... 0.0789 sec/batch
Epoch: 4/20... Training Step: 3070... Training loss: 1.4455... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3071... Training loss: 1.4981... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3072... Training loss: 1.5950... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3073... Training loss: 1.5673... 0.0823 sec/batch
Epoch: 4/20... Training Step: 3074... Training loss: 1.4915... 0.0901 sec/batch
Epoch: 4/20... Training Step: 3075... Training loss: 1.5220... 0.0841 sec/batch
Epoch: 4/20... Training Step: 3076... Training loss: 1.5373... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3077... Training loss: 1.5020... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3078... Training loss: 1.5335... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3079... Training loss: 1.5605... 0.0801 sec/batch
Epoch: 4/20... Training Step: 3080... Training loss: 1.6091... 0.0826 sec/batch
Epoch: 4/20... Training Step: 3081... Training loss: 1.5749... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3082... Training loss: 1.5665... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3083... Training loss: 1.5100... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3084... Training loss: 1.6108... 0.0793 sec/batch
Epoch: 4/20... Training Step: 3085... Training loss: 1.5565... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3086... Training loss: 1.5964... 0.0792 sec/batch
Epoch: 4/20... Training Step: 3087... Training loss: 1.5642... 0.0782 sec/batch
Epoch: 4/20... Training Step: 3088... Training loss: 1.5276... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3089... Training loss: 1.6622... 0.0986 sec/batch
Epoch: 4/20... Training Step: 3090... Training loss: 1.5911... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3091... Training loss: 1.5199... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3092... Training loss: 1.6773... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3093... Training loss: 1.5986... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3094... Training loss: 1.6785... 0.0782 sec/batch
Epoch: 4/20... Training Step: 3095... Training loss: 1.5631... 0.0783 sec/batch
Epoch: 4/20... Training Step: 3096... Training loss: 1.5579... 0.0811 sec/batch
Epoch: 4/20... Training Step: 3097... Training loss: 1.6631... 0.0809 sec/batch
Epoch: 4/20... Training Step: 3098... Training loss: 1.5415... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3099... Training loss: 1.5777... 0.0905 sec/batch
Epoch: 4/20... Training Step: 3100... Training loss: 1.5531... 0.0806 sec/batch
Epoch: 4/20... Training Step: 3101... Training loss: 1.6223... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3102... Training loss: 1.6375... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3103... Training loss: 1.6074... 0.0788 sec/batch
Epoch: 4/20... Training Step: 3104... Training loss: 1.5904... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3105... Training loss: 1.5471... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3106... Training loss: 1.5620... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3107... Training loss: 1.5405... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3108... Training loss: 1.6220... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3109... Training loss: 1.5953... 0.0795 sec/batch
Epoch: 4/20... Training Step: 3110... Training loss: 1.5548... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3111... Training loss: 1.5788... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3112... Training loss: 1.5983... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3113... Training loss: 1.6138... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3114... Training loss: 1.6703... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3115... Training loss: 1.6542... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3116... Training loss: 1.5694... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3117... Training loss: 1.5656... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3118... Training loss: 1.5215... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3119... Training loss: 1.5339... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3120... Training loss: 1.5731... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3121... Training loss: 1.4645... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3122... Training loss: 1.6308... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3123... Training loss: 1.5424... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3124... Training loss: 1.4847... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3125... Training loss: 1.5399... 0.0837 sec/batch
Epoch: 4/20... Training Step: 3126... Training loss: 1.5290... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3127... Training loss: 1.4369... 0.0839 sec/batch
Epoch: 4/20... Training Step: 3128... Training loss: 1.5551... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3129... Training loss: 1.5158... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3130... Training loss: 1.5499... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3131... Training loss: 1.5120... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3132... Training loss: 1.4911... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3133... Training loss: 1.5417... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3134... Training loss: 1.5414... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3135... Training loss: 1.5564... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3136... Training loss: 1.5750... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3137... Training loss: 1.5157... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3138... Training loss: 1.5062... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3139... Training loss: 1.5439... 0.0797 sec/batch
Epoch: 4/20... Training Step: 3140... Training loss: 1.6095... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3141... Training loss: 1.5873... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3142... Training loss: 1.6644... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3143... Training loss: 1.5872... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3144... Training loss: 1.5637... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3145... Training loss: 1.6028... 0.0786 sec/batch
Epoch: 4/20... Training Step: 3146... Training loss: 1.5483... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3147... Training loss: 1.6078... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3148... Training loss: 1.5072... 0.0781 sec/batch
Epoch: 4/20... Training Step: 3149... Training loss: 1.5547... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3150... Training loss: 1.6220... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3151... Training loss: 1.5450... 0.0815 sec/batch
Epoch: 4/20... Training Step: 3152... Training loss: 1.5902... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3153... Training loss: 1.6039... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3154... Training loss: 1.6677... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3155... Training loss: 1.6287... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3156... Training loss: 1.5356... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3157... Training loss: 1.6070... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3158... Training loss: 1.6542... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3159... Training loss: 1.5316... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3160... Training loss: 1.5809... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3161... Training loss: 1.5901... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3162... Training loss: 1.5967... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3163... Training loss: 1.6076... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3164... Training loss: 1.5562... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3165... Training loss: 1.5382... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3166... Training loss: 1.6128... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3167... Training loss: 1.6467... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3168... Training loss: 1.6202... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3169... Training loss: 1.6168... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3170... Training loss: 1.6767... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3171... Training loss: 1.5524... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3172... Training loss: 1.5142... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3173... Training loss: 1.6164... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3174... Training loss: 1.5396... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3175... Training loss: 1.6123... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3176... Training loss: 1.5693... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3177... Training loss: 1.5581... 0.0781 sec/batch
Epoch: 4/20... Training Step: 3178... Training loss: 1.5513... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3179... Training loss: 1.5623... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3180... Training loss: 1.5501... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3181... Training loss: 1.5690... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3182... Training loss: 1.6324... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3183... Training loss: 1.5648... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3184... Training loss: 1.5317... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3185... Training loss: 1.6190... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3186... Training loss: 1.5671... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3187... Training loss: 1.4900... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3188... Training loss: 1.6430... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3189... Training loss: 1.5768... 0.0791 sec/batch
Epoch: 4/20... Training Step: 3190... Training loss: 1.5895... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3191... Training loss: 1.5386... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3192... Training loss: 1.5773... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3193... Training loss: 1.5389... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3194... Training loss: 1.5892... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3195... Training loss: 1.5851... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3196... Training loss: 1.5352... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3197... Training loss: 1.5084... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3198... Training loss: 1.6322... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3199... Training loss: 1.5981... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3200... Training loss: 1.5334... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3201... Training loss: 1.6056... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3202... Training loss: 1.5768... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3203... Training loss: 1.5624... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3204... Training loss: 1.5684... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3205... Training loss: 1.4660... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3206... Training loss: 1.5183... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3207... Training loss: 1.5483... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3208... Training loss: 1.5785... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3209... Training loss: 1.5375... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3210... Training loss: 1.6415... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3211... Training loss: 1.4988... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3212... Training loss: 1.5583... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3213... Training loss: 1.5158... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3214... Training loss: 1.5964... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3215... Training loss: 1.5976... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3216... Training loss: 1.5450... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3217... Training loss: 1.5570... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3218... Training loss: 1.5591... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3219... Training loss: 1.5858... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3220... Training loss: 1.6206... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3221... Training loss: 1.5398... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3222... Training loss: 1.6025... 0.0790 sec/batch
Epoch: 4/20... Training Step: 3223... Training loss: 1.5229... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3224... Training loss: 1.4847... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3225... Training loss: 1.5424... 0.0805 sec/batch
Epoch: 4/20... Training Step: 3226... Training loss: 1.5836... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3227... Training loss: 1.5459... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3228... Training loss: 1.5396... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3229... Training loss: 1.5374... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3230... Training loss: 1.5444... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3231... Training loss: 1.5106... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3232... Training loss: 1.5548... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3233... Training loss: 1.6501... 0.0733 sec/batch
Epoch: 4/20... Training Step: 3234... Training loss: 1.5375... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3235... Training loss: 1.5344... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3236... Training loss: 1.5583... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3237... Training loss: 1.6221... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3238... Training loss: 1.6204... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3239... Training loss: 1.5832... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3240... Training loss: 1.5816... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3241... Training loss: 1.5641... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3242... Training loss: 1.5170... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3243... Training loss: 1.5105... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3244... Training loss: 1.5431... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3245... Training loss: 1.4823... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3246... Training loss: 1.4978... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3247... Training loss: 1.5369... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3248... Training loss: 1.5378... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3249... Training loss: 1.4509... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3250... Training loss: 1.4583... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3251... Training loss: 1.5177... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3252... Training loss: 1.5046... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3253... Training loss: 1.5502... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3254... Training loss: 1.4361... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3255... Training loss: 1.5135... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3256... Training loss: 1.6188... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3257... Training loss: 1.5534... 0.0728 sec/batch
Epoch: 4/20... Training Step: 3258... Training loss: 1.5495... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3259... Training loss: 1.5641... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3260... Training loss: 1.5076... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3261... Training loss: 1.4702... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3262... Training loss: 1.5116... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3263... Training loss: 1.5657... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3264... Training loss: 1.5202... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3265... Training loss: 1.5299... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3266... Training loss: 1.5345... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3267... Training loss: 1.5260... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3268... Training loss: 1.5371... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3269... Training loss: 1.4817... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3270... Training loss: 1.4902... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3271... Training loss: 1.5576... 0.0782 sec/batch
Epoch: 4/20... Training Step: 3272... Training loss: 1.5557... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3273... Training loss: 1.4945... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3274... Training loss: 1.5879... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3275... Training loss: 1.5614... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3276... Training loss: 1.5796... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3277... Training loss: 1.5912... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3278... Training loss: 1.5164... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3279... Training loss: 1.5113... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3280... Training loss: 1.6008... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3281... Training loss: 1.5528... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3282... Training loss: 1.5994... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3283... Training loss: 1.5008... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3284... Training loss: 1.4581... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3285... Training loss: 1.6344... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3286... Training loss: 1.5392... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3287... Training loss: 1.5973... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3288... Training loss: 1.5787... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3289... Training loss: 1.5167... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3290... Training loss: 1.5364... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3291... Training loss: 1.5575... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3292... Training loss: 1.4953... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3293... Training loss: 1.4809... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3294... Training loss: 1.4838... 0.0830 sec/batch
Epoch: 4/20... Training Step: 3295... Training loss: 1.4975... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3296... Training loss: 1.5641... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3297... Training loss: 1.5891... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3298... Training loss: 1.4652... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3299... Training loss: 1.4894... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3300... Training loss: 1.5046... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3301... Training loss: 1.5786... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3302... Training loss: 1.4998... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3303... Training loss: 1.5355... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3304... Training loss: 1.5000... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3305... Training loss: 1.5282... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3306... Training loss: 1.5565... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3307... Training loss: 1.5867... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3308... Training loss: 1.4984... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3309... Training loss: 1.6304... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3310... Training loss: 1.6113... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3311... Training loss: 1.6166... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3312... Training loss: 1.5771... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3313... Training loss: 1.5589... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3314... Training loss: 1.5298... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3315... Training loss: 1.5196... 0.0845 sec/batch
Epoch: 4/20... Training Step: 3316... Training loss: 1.5138... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3317... Training loss: 1.5565... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3318... Training loss: 1.5155... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3319... Training loss: 1.4764... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3320... Training loss: 1.5787... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3321... Training loss: 1.5093... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3322... Training loss: 1.5754... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3323... Training loss: 1.5850... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3324... Training loss: 1.5576... 0.0787 sec/batch
Epoch: 4/20... Training Step: 3325... Training loss: 1.6022... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3326... Training loss: 1.6131... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3327... Training loss: 1.5643... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3328... Training loss: 1.5025... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3329... Training loss: 1.5362... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3330... Training loss: 1.5257... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3331... Training loss: 1.5739... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3332... Training loss: 1.5489... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3333... Training loss: 1.4950... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3334... Training loss: 1.5537... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3335... Training loss: 1.5444... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3336... Training loss: 1.5466... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3337... Training loss: 1.5690... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3338... Training loss: 1.6235... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3339... Training loss: 1.5386... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3340... Training loss: 1.5411... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3341... Training loss: 1.5284... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3342... Training loss: 1.5229... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3343... Training loss: 1.5319... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3344... Training loss: 1.5089... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3345... Training loss: 1.4801... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3346... Training loss: 1.4802... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3347... Training loss: 1.6080... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3348... Training loss: 1.5054... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3349... Training loss: 1.5349... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3350... Training loss: 1.4714... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3351... Training loss: 1.5475... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3352... Training loss: 1.5159... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3353... Training loss: 1.5139... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3354... Training loss: 1.5336... 0.0813 sec/batch
Epoch: 4/20... Training Step: 3355... Training loss: 1.5160... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3356... Training loss: 1.5566... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3357... Training loss: 1.5268... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3358... Training loss: 1.5152... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3359... Training loss: 1.5714... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3360... Training loss: 1.5695... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3361... Training loss: 1.5672... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3362... Training loss: 1.5197... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3363... Training loss: 1.5801... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3364... Training loss: 1.6049... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3365... Training loss: 1.5866... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3366... Training loss: 1.4973... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3367... Training loss: 1.5639... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3368... Training loss: 1.5342... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3369... Training loss: 1.5777... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3370... Training loss: 1.4944... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3371... Training loss: 1.5702... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3372... Training loss: 1.5301... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3373... Training loss: 1.5180... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3374... Training loss: 1.5994... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3375... Training loss: 1.5401... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3376... Training loss: 1.5327... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3377... Training loss: 1.4856... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3378... Training loss: 1.5764... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3379... Training loss: 1.5097... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3380... Training loss: 1.5781... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3381... Training loss: 1.5039... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3382... Training loss: 1.5452... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3383... Training loss: 1.4956... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3384... Training loss: 1.6350... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3385... Training loss: 1.4988... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3386... Training loss: 1.4844... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3387... Training loss: 1.4571... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3388... Training loss: 1.4667... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3389... Training loss: 1.5282... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3390... Training loss: 1.4624... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3391... Training loss: 1.5144... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3392... Training loss: 1.5700... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3393... Training loss: 1.4483... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3394... Training loss: 1.4850... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3395... Training loss: 1.5623... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3396... Training loss: 1.6301... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3397... Training loss: 1.5069... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3398... Training loss: 1.5356... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3399... Training loss: 1.4984... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3400... Training loss: 1.5515... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3401... Training loss: 1.5586... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3402... Training loss: 1.5593... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3403... Training loss: 1.6080... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3404... Training loss: 1.6137... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3405... Training loss: 1.5728... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3406... Training loss: 1.6346... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3407... Training loss: 1.5395... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3408... Training loss: 1.5919... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3409... Training loss: 1.6057... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3410... Training loss: 1.5802... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3411... Training loss: 1.5904... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3412... Training loss: 1.5299... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3413... Training loss: 1.5996... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3414... Training loss: 1.6658... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3415... Training loss: 1.5526... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3416... Training loss: 1.5881... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3417... Training loss: 1.5772... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3418... Training loss: 1.5748... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3419... Training loss: 1.5537... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3420... Training loss: 1.5521... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3421... Training loss: 1.5101... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3422... Training loss: 1.6083... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3423... Training loss: 1.5251... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3424... Training loss: 1.5569... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3425... Training loss: 1.6004... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3426... Training loss: 1.6605... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3427... Training loss: 1.5350... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3428... Training loss: 1.5571... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3429... Training loss: 1.5691... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3430... Training loss: 1.5697... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3431... Training loss: 1.5589... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3432... Training loss: 1.5422... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3433... Training loss: 1.4753... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3434... Training loss: 1.5170... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3435... Training loss: 1.5636... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3436... Training loss: 1.5876... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3437... Training loss: 1.5793... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3438... Training loss: 1.5378... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3439... Training loss: 1.4868... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3440... Training loss: 1.5372... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3441... Training loss: 1.4954... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3442... Training loss: 1.5076... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3443... Training loss: 1.4515... 0.0791 sec/batch
Epoch: 4/20... Training Step: 3444... Training loss: 1.5311... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3445... Training loss: 1.5162... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3446... Training loss: 1.6127... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3447... Training loss: 1.5241... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3448... Training loss: 1.5220... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3449... Training loss: 1.5332... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3450... Training loss: 1.5111... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3451... Training loss: 1.4072... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3452... Training loss: 1.4924... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3453... Training loss: 1.5120... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3454... Training loss: 1.5003... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3455... Training loss: 1.4678... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3456... Training loss: 1.4833... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3457... Training loss: 1.4677... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3458... Training loss: 1.5547... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3459... Training loss: 1.5472... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3460... Training loss: 1.5509... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3461... Training loss: 1.4830... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3462... Training loss: 1.4953... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3463... Training loss: 1.5940... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3464... Training loss: 1.4912... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3465... Training loss: 1.5185... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3466... Training loss: 1.4888... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3467... Training loss: 1.4934... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3468... Training loss: 1.4722... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3469... Training loss: 1.5140... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3470... Training loss: 1.5064... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3471... Training loss: 1.4654... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3472... Training loss: 1.4721... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3473... Training loss: 1.5243... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3474... Training loss: 1.5102... 0.0803 sec/batch
Epoch: 4/20... Training Step: 3475... Training loss: 1.5325... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3476... Training loss: 1.5633... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3477... Training loss: 1.5075... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3478... Training loss: 1.5543... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3479... Training loss: 1.5160... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3480... Training loss: 1.5226... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3481... Training loss: 1.5437... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3482... Training loss: 1.4495... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3483... Training loss: 1.5037... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3484... Training loss: 1.5037... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3485... Training loss: 1.4759... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3486... Training loss: 1.4774... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3487... Training loss: 1.5530... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3488... Training loss: 1.5515... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3489... Training loss: 1.6035... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3490... Training loss: 1.5824... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3491... Training loss: 1.5213... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3492... Training loss: 1.5302... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3493... Training loss: 1.5341... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3494... Training loss: 1.5273... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3495... Training loss: 1.5154... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3496... Training loss: 1.4939... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3497... Training loss: 1.5482... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3498... Training loss: 1.4799... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3499... Training loss: 1.5806... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3500... Training loss: 1.5210... 0.0804 sec/batch
Epoch: 4/20... Training Step: 3501... Training loss: 1.5069... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3502... Training loss: 1.5420... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3503... Training loss: 1.5385... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3504... Training loss: 1.4672... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3505... Training loss: 1.5205... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3506... Training loss: 1.5549... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3507... Training loss: 1.6310... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3508... Training loss: 1.5544... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3509... Training loss: 1.5871... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3510... Training loss: 1.5673... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3511... Training loss: 1.5429... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3512... Training loss: 1.4925... 0.0863 sec/batch
Epoch: 4/20... Training Step: 3513... Training loss: 1.4823... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3514... Training loss: 1.5147... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3515... Training loss: 1.4980... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3516... Training loss: 1.4881... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3517... Training loss: 1.6300... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3518... Training loss: 1.4864... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3519... Training loss: 1.4842... 0.0784 sec/batch
Epoch: 4/20... Training Step: 3520... Training loss: 1.5038... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3521... Training loss: 1.4781... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3522... Training loss: 1.4936... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3523... Training loss: 1.5628... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3524... Training loss: 1.5853... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3525... Training loss: 1.5009... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3526... Training loss: 1.4882... 0.0789 sec/batch
Epoch: 4/20... Training Step: 3527... Training loss: 1.5436... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3528... Training loss: 1.4616... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3529... Training loss: 1.4671... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3530... Training loss: 1.5320... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3531... Training loss: 1.5369... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3532... Training loss: 1.4681... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3533... Training loss: 1.4646... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3534... Training loss: 1.5243... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3535... Training loss: 1.5346... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3536... Training loss: 1.4402... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3537... Training loss: 1.5004... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3538... Training loss: 1.5228... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3539... Training loss: 1.4563... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3540... Training loss: 1.4621... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3541... Training loss: 1.5113... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3542... Training loss: 1.5468... 0.0786 sec/batch
Epoch: 4/20... Training Step: 3543... Training loss: 1.4854... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3544... Training loss: 1.4560... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3545... Training loss: 1.4212... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3546... Training loss: 1.4489... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3547... Training loss: 1.4779... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3548... Training loss: 1.4935... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3549... Training loss: 1.5716... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3550... Training loss: 1.5573... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3551... Training loss: 1.5379... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3552... Training loss: 1.5198... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3553... Training loss: 1.5194... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3554... Training loss: 1.6396... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3555... Training loss: 1.5609... 0.0729 sec/batch
Epoch: 4/20... Training Step: 3556... Training loss: 1.4713... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3557... Training loss: 1.4408... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3558... Training loss: 1.4770... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3559... Training loss: 1.5915... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3560... Training loss: 1.4639... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3561... Training loss: 1.4921... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3562... Training loss: 1.5366... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3563... Training loss: 1.5129... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3564... Training loss: 1.5507... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3565... Training loss: 1.5100... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3566... Training loss: 1.5137... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3567... Training loss: 1.5202... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3568... Training loss: 1.5708... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3569... Training loss: 1.4891... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3570... Training loss: 1.4825... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3571... Training loss: 1.4610... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3572... Training loss: 1.5462... 0.0786 sec/batch
Epoch: 4/20... Training Step: 3573... Training loss: 1.4097... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3574... Training loss: 1.4174... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3575... Training loss: 1.4718... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3576... Training loss: 1.4890... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3577... Training loss: 1.5628... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3578... Training loss: 1.4823... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3579... Training loss: 1.5225... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3580... Training loss: 1.5342... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3581... Training loss: 1.4546... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3582... Training loss: 1.5045... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3583... Training loss: 1.5614... 0.0778 sec/batch
Epoch: 4/20... Training Step: 3584... Training loss: 1.5453... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3585... Training loss: 1.5531... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3586... Training loss: 1.5442... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3587... Training loss: 1.4158... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3588... Training loss: 1.5256... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3589... Training loss: 1.5828... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3590... Training loss: 1.4653... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3591... Training loss: 1.5825... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3592... Training loss: 1.5106... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3593... Training loss: 1.5303... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3594... Training loss: 1.5734... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3595... Training loss: 1.4565... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3596... Training loss: 1.5270... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3597... Training loss: 1.4746... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3598... Training loss: 1.4108... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3599... Training loss: 1.5343... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3600... Training loss: 1.5320... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3601... Training loss: 1.4754... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3602... Training loss: 1.4865... 0.0732 sec/batch
Epoch: 4/20... Training Step: 3603... Training loss: 1.4953... 0.0730 sec/batch
Epoch: 4/20... Training Step: 3604... Training loss: 1.5180... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3605... Training loss: 1.3621... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3606... Training loss: 1.5245... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3607... Training loss: 1.5172... 0.0794 sec/batch
Epoch: 4/20... Training Step: 3608... Training loss: 1.4878... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3609... Training loss: 1.4495... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3610... Training loss: 1.5115... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3611... Training loss: 1.5157... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3612... Training loss: 1.4816... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3613... Training loss: 1.5837... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3614... Training loss: 1.4822... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3615... Training loss: 1.5348... 0.0742 sec/batch
Epoch: 4/20... Training Step: 3616... Training loss: 1.5811... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3617... Training loss: 1.6208... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3618... Training loss: 1.5570... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3619... Training loss: 1.4998... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3620... Training loss: 1.4951... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3621... Training loss: 1.4800... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3622... Training loss: 1.5493... 0.0744 sec/batch
Epoch: 4/20... Training Step: 3623... Training loss: 1.5516... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3624... Training loss: 1.4181... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3625... Training loss: 1.4724... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3626... Training loss: 1.4956... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3627... Training loss: 1.5177... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3628... Training loss: 1.5715... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3629... Training loss: 1.5034... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3630... Training loss: 1.5671... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3631... Training loss: 1.5875... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3632... Training loss: 1.5709... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3633... Training loss: 1.4436... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3634... Training loss: 1.4980... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3635... Training loss: 1.5130... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3636... Training loss: 1.5836... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3637... Training loss: 1.5843... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3638... Training loss: 1.5991... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3639... Training loss: 1.5348... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3640... Training loss: 1.6645... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3641... Training loss: 1.6164... 0.0726 sec/batch
Epoch: 4/20... Training Step: 3642... Training loss: 1.4802... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3643... Training loss: 1.6261... 0.0783 sec/batch
Epoch: 4/20... Training Step: 3644... Training loss: 1.5574... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3645... Training loss: 1.5163... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3646... Training loss: 1.4692... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3647... Training loss: 1.5524... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3648... Training loss: 1.5470... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3649... Training loss: 1.5075... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3650... Training loss: 1.5281... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3651... Training loss: 1.4872... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3652... Training loss: 1.5102... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3653... Training loss: 1.6305... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3654... Training loss: 1.4756... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3655... Training loss: 1.4765... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3656... Training loss: 1.5424... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3657... Training loss: 1.5472... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3658... Training loss: 1.5027... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3659... Training loss: 1.5358... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3660... Training loss: 1.4996... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3661... Training loss: 1.5822... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3662... Training loss: 1.5592... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3663... Training loss: 1.5028... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3664... Training loss: 1.4899... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3665... Training loss: 1.4972... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3666... Training loss: 1.5661... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3667... Training loss: 1.5062... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3668... Training loss: 1.5336... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3669... Training loss: 1.4591... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3670... Training loss: 1.4677... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3671... Training loss: 1.5403... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3672... Training loss: 1.5140... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3673... Training loss: 1.5745... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3674... Training loss: 1.5348... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3675... Training loss: 1.4880... 0.0800 sec/batch
Epoch: 4/20... Training Step: 3676... Training loss: 1.5381... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3677... Training loss: 1.5460... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3678... Training loss: 1.4952... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3679... Training loss: 1.5241... 0.0782 sec/batch
Epoch: 4/20... Training Step: 3680... Training loss: 1.4471... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3681... Training loss: 1.4220... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3682... Training loss: 1.4683... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3683... Training loss: 1.5370... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3684... Training loss: 1.5510... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3685... Training loss: 1.5445... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3686... Training loss: 1.5139... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3687... Training loss: 1.5045... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3688... Training loss: 1.6101... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3689... Training loss: 1.4694... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3690... Training loss: 1.4926... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3691... Training loss: 1.5353... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3692... Training loss: 1.5365... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3693... Training loss: 1.5757... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3694... Training loss: 1.5290... 0.0778 sec/batch
Epoch: 4/20... Training Step: 3695... Training loss: 1.5658... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3696... Training loss: 1.6351... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3697... Training loss: 1.5311... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3698... Training loss: 1.5260... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3699... Training loss: 1.4968... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3700... Training loss: 1.5410... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3701... Training loss: 1.5577... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3702... Training loss: 1.6182... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3703... Training loss: 1.5015... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3704... Training loss: 1.5656... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3705... Training loss: 1.5607... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3706... Training loss: 1.4989... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3707... Training loss: 1.5053... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3708... Training loss: 1.5565... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3709... Training loss: 1.4667... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3710... Training loss: 1.5721... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3711... Training loss: 1.4760... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3712... Training loss: 1.5109... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3713... Training loss: 1.4615... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3714... Training loss: 1.5821... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3715... Training loss: 1.4885... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3716... Training loss: 1.5281... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3717... Training loss: 1.4930... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3718... Training loss: 1.5036... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3719... Training loss: 1.4978... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3720... Training loss: 1.5755... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3721... Training loss: 1.5467... 0.0784 sec/batch
Epoch: 4/20... Training Step: 3722... Training loss: 1.5115... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3723... Training loss: 1.5397... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3724... Training loss: 1.5350... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3725... Training loss: 1.4570... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3726... Training loss: 1.4282... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3727... Training loss: 1.6091... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3728... Training loss: 1.4664... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3729... Training loss: 1.5243... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3730... Training loss: 1.5105... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3731... Training loss: 1.4405... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3732... Training loss: 1.5004... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3733... Training loss: 1.5283... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3734... Training loss: 1.5851... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3735... Training loss: 1.5871... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3736... Training loss: 1.5937... 0.0783 sec/batch
Epoch: 4/20... Training Step: 3737... Training loss: 1.5625... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3738... Training loss: 1.4651... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3739... Training loss: 1.5444... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3740... Training loss: 1.4903... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3741... Training loss: 1.5544... 0.0727 sec/batch
Epoch: 4/20... Training Step: 3742... Training loss: 1.5439... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3743... Training loss: 1.5097... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3744... Training loss: 1.4896... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3745... Training loss: 1.4997... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3746... Training loss: 1.5322... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3747... Training loss: 1.5335... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3748... Training loss: 1.6112... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3749... Training loss: 1.5654... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3750... Training loss: 1.5550... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3751... Training loss: 1.5039... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3752... Training loss: 1.4459... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3753... Training loss: 1.4667... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3754... Training loss: 1.5212... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3755... Training loss: 1.4987... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3756... Training loss: 1.4687... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3757... Training loss: 1.5111... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3758... Training loss: 1.5033... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3759... Training loss: 1.5467... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3760... Training loss: 1.5320... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3761... Training loss: 1.5179... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3762... Training loss: 1.5125... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3763... Training loss: 1.5189... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3764... Training loss: 1.5198... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3765... Training loss: 1.5290... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3766... Training loss: 1.5111... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3767... Training loss: 1.5123... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3768... Training loss: 1.4773... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3769... Training loss: 1.4910... 0.0786 sec/batch
Epoch: 4/20... Training Step: 3770... Training loss: 1.4785... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3771... Training loss: 1.4343... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3772... Training loss: 1.5137... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3773... Training loss: 1.5029... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3774... Training loss: 1.5068... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3775... Training loss: 1.5180... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3776... Training loss: 1.4599... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3777... Training loss: 1.5533... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3778... Training loss: 1.5576... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3779... Training loss: 1.5873... 0.0788 sec/batch
Epoch: 4/20... Training Step: 3780... Training loss: 1.5275... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3781... Training loss: 1.4844... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3782... Training loss: 1.5088... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3783... Training loss: 1.4882... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3784... Training loss: 1.5391... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3785... Training loss: 1.5852... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3786... Training loss: 1.5820... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3787... Training loss: 1.5343... 0.0883 sec/batch
Epoch: 4/20... Training Step: 3788... Training loss: 1.5865... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3789... Training loss: 1.5201... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3790... Training loss: 1.5821... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3791... Training loss: 1.5750... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3792... Training loss: 1.5992... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3793... Training loss: 1.6029... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3794... Training loss: 1.6454... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3795... Training loss: 1.5715... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3796... Training loss: 1.5336... 0.0855 sec/batch
Epoch: 4/20... Training Step: 3797... Training loss: 1.5095... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3798... Training loss: 1.4415... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3799... Training loss: 1.5385... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3800... Training loss: 1.5255... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3801... Training loss: 1.4945... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3802... Training loss: 1.5525... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3803... Training loss: 1.5412... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3804... Training loss: 1.5792... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3805... Training loss: 1.5437... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3806... Training loss: 1.4968... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3807... Training loss: 1.6160... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3808... Training loss: 1.5959... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3809... Training loss: 1.6199... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3810... Training loss: 1.5360... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3811... Training loss: 1.4690... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3812... Training loss: 1.5102... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3813... Training loss: 1.5398... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3814... Training loss: 1.5818... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3815... Training loss: 1.5964... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3816... Training loss: 1.5077... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3817... Training loss: 1.5799... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3818... Training loss: 1.6789... 0.0772 sec/batch
Epoch: 4/20... Training Step: 3819... Training loss: 1.5632... 0.0759 sec/batch
Epoch: 4/20... Training Step: 3820... Training loss: 1.6043... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3821... Training loss: 1.6224... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3822... Training loss: 1.4637... 0.0836 sec/batch
Epoch: 4/20... Training Step: 3823... Training loss: 1.5330... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3824... Training loss: 1.5331... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3825... Training loss: 1.5227... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3826... Training loss: 1.5365... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3827... Training loss: 1.5047... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3828... Training loss: 1.6063... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3829... Training loss: 1.5997... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3830... Training loss: 1.5526... 0.0758 sec/batch
Epoch: 4/20... Training Step: 3831... Training loss: 1.5276... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3832... Training loss: 1.5189... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3833... Training loss: 1.5262... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3834... Training loss: 1.5162... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3835... Training loss: 1.5706... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3836... Training loss: 1.5560... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3837... Training loss: 1.5490... 0.0761 sec/batch
Epoch: 4/20... Training Step: 3838... Training loss: 1.4914... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3839... Training loss: 1.4909... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3840... Training loss: 1.5039... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3841... Training loss: 1.4683... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3842... Training loss: 1.5188... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3843... Training loss: 1.5179... 0.0783 sec/batch
Epoch: 4/20... Training Step: 3844... Training loss: 1.6183... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3845... Training loss: 1.5194... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3846... Training loss: 1.6098... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3847... Training loss: 1.5622... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3848... Training loss: 1.5334... 0.0732 sec/batch
Epoch: 4/20... Training Step: 3849... Training loss: 1.5747... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3850... Training loss: 1.5366... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3851... Training loss: 1.5391... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3852... Training loss: 1.5687... 0.0775 sec/batch
Epoch: 4/20... Training Step: 3853... Training loss: 1.5455... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3854... Training loss: 1.5767... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3855... Training loss: 1.5461... 0.0791 sec/batch
Epoch: 4/20... Training Step: 3856... Training loss: 1.4934... 0.0768 sec/batch
Epoch: 4/20... Training Step: 3857... Training loss: 1.5361... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3858... Training loss: 1.4539... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3859... Training loss: 1.5435... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3860... Training loss: 1.5224... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3861... Training loss: 1.5345... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3862... Training loss: 1.6126... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3863... Training loss: 1.5334... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3864... Training loss: 1.6100... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3865... Training loss: 1.5501... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3866... Training loss: 1.5993... 0.0771 sec/batch
Epoch: 4/20... Training Step: 3867... Training loss: 1.5905... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3868... Training loss: 1.4755... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3869... Training loss: 1.5131... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3870... Training loss: 1.4472... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3871... Training loss: 1.5875... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3872... Training loss: 1.5141... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3873... Training loss: 1.5272... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3874... Training loss: 1.5497... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3875... Training loss: 1.4971... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3876... Training loss: 1.4690... 0.0797 sec/batch
Epoch: 4/20... Training Step: 3877... Training loss: 1.5026... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3878... Training loss: 1.4837... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3879... Training loss: 1.4968... 0.0782 sec/batch
Epoch: 4/20... Training Step: 3880... Training loss: 1.4586... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3881... Training loss: 1.4617... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3882... Training loss: 1.5254... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3883... Training loss: 1.4452... 0.0740 sec/batch
Epoch: 4/20... Training Step: 3884... Training loss: 1.5292... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3885... Training loss: 1.4684... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3886... Training loss: 1.4064... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3887... Training loss: 1.4545... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3888... Training loss: 1.5219... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3889... Training loss: 1.6100... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3890... Training loss: 1.5199... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3891... Training loss: 1.5106... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3892... Training loss: 1.5091... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3893... Training loss: 1.4927... 0.0785 sec/batch
Epoch: 4/20... Training Step: 3894... Training loss: 1.5598... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3895... Training loss: 1.5301... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3896... Training loss: 1.5487... 0.0757 sec/batch
Epoch: 4/20... Training Step: 3897... Training loss: 1.6579... 0.0783 sec/batch
Epoch: 4/20... Training Step: 3898... Training loss: 1.5732... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3899... Training loss: 1.5275... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3900... Training loss: 1.5109... 0.0767 sec/batch
Epoch: 4/20... Training Step: 3901... Training loss: 1.5573... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3902... Training loss: 1.5803... 0.0766 sec/batch
Epoch: 4/20... Training Step: 3903... Training loss: 1.5125... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3904... Training loss: 1.5367... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3905... Training loss: 1.6006... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3906... Training loss: 1.5715... 0.0773 sec/batch
Epoch: 4/20... Training Step: 3907... Training loss: 1.5399... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3908... Training loss: 1.6191... 0.0737 sec/batch
Epoch: 4/20... Training Step: 3909... Training loss: 1.5249... 0.0779 sec/batch
Epoch: 4/20... Training Step: 3910... Training loss: 1.5325... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3911... Training loss: 1.5500... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3912... Training loss: 1.5696... 0.0745 sec/batch
Epoch: 4/20... Training Step: 3913... Training loss: 1.4975... 0.0799 sec/batch
Epoch: 4/20... Training Step: 3914... Training loss: 1.5323... 0.0760 sec/batch
Epoch: 4/20... Training Step: 3915... Training loss: 1.5198... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3916... Training loss: 1.6547... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3917... Training loss: 1.5574... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3918... Training loss: 1.5666... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3919... Training loss: 1.5840... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3920... Training loss: 1.5451... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3921... Training loss: 1.5881... 0.0770 sec/batch
Epoch: 4/20... Training Step: 3922... Training loss: 1.5230... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3923... Training loss: 1.4897... 0.0780 sec/batch
Epoch: 4/20... Training Step: 3924... Training loss: 1.4788... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3925... Training loss: 1.5218... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3926... Training loss: 1.4260... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3927... Training loss: 1.5253... 0.0784 sec/batch
Epoch: 4/20... Training Step: 3928... Training loss: 1.4885... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3929... Training loss: 1.5212... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3930... Training loss: 1.5457... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3931... Training loss: 1.5632... 0.0769 sec/batch
Epoch: 4/20... Training Step: 3932... Training loss: 1.5634... 0.0764 sec/batch
Epoch: 4/20... Training Step: 3933... Training loss: 1.5075... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3934... Training loss: 1.5714... 0.0754 sec/batch
Epoch: 4/20... Training Step: 3935... Training loss: 1.4700... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3936... Training loss: 1.4829... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3937... Training loss: 1.5315... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3938... Training loss: 1.5311... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3939... Training loss: 1.5610... 0.0794 sec/batch
Epoch: 4/20... Training Step: 3940... Training loss: 1.4939... 0.0753 sec/batch
Epoch: 4/20... Training Step: 3941... Training loss: 1.5831... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3942... Training loss: 1.6398... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3943... Training loss: 1.5714... 0.0746 sec/batch
Epoch: 4/20... Training Step: 3944... Training loss: 1.6037... 0.0747 sec/batch
Epoch: 4/20... Training Step: 3945... Training loss: 1.5733... 0.0756 sec/batch
Epoch: 4/20... Training Step: 3946... Training loss: 1.7375... 0.0736 sec/batch
Epoch: 4/20... Training Step: 3947... Training loss: 1.4764... 0.0749 sec/batch
Epoch: 4/20... Training Step: 3948... Training loss: 1.6310... 0.0776 sec/batch
Epoch: 4/20... Training Step: 3949... Training loss: 1.6223... 0.0748 sec/batch
Epoch: 4/20... Training Step: 3950... Training loss: 1.5022... 0.0731 sec/batch
Epoch: 4/20... Training Step: 3951... Training loss: 1.6045... 0.0774 sec/batch
Epoch: 4/20... Training Step: 3952... Training loss: 1.5926... 0.0750 sec/batch
Epoch: 4/20... Training Step: 3953... Training loss: 1.5928... 0.0735 sec/batch
Epoch: 4/20... Training Step: 3954... Training loss: 1.5718... 0.0777 sec/batch
Epoch: 4/20... Training Step: 3955... Training loss: 1.5660... 0.0765 sec/batch
Epoch: 4/20... Training Step: 3956... Training loss: 1.5358... 0.0763 sec/batch
Epoch: 4/20... Training Step: 3957... Training loss: 1.4448... 0.0755 sec/batch
Epoch: 4/20... Training Step: 3958... Training loss: 1.6342... 0.0739 sec/batch
Epoch: 4/20... Training Step: 3959... Training loss: 1.5721... 0.0751 sec/batch
Epoch: 4/20... Training Step: 3960... Training loss: 1.5355... 0.0762 sec/batch
Epoch: 4/20... Training Step: 3961... Training loss: 1.5403... 0.0734 sec/batch
Epoch: 4/20... Training Step: 3962... Training loss: 1.4743... 0.0741 sec/batch
Epoch: 4/20... Training Step: 3963... Training loss: 1.4898... 0.0790 sec/batch
Epoch: 4/20... Training Step: 3964... Training loss: 1.5122... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3965... Training loss: 1.5201... 0.0738 sec/batch
Epoch: 4/20... Training Step: 3966... Training loss: 1.4280... 0.0752 sec/batch
Epoch: 4/20... Training Step: 3967... Training loss: 1.5128... 0.0743 sec/batch
Epoch: 4/20... Training Step: 3968... Training loss: 1.5737... 0.0740 sec/batch
Epoch: 5/20... Training Step: 3969... Training loss: 1.6181... 0.0777 sec/batch
Epoch: 5/20... Training Step: 3970... Training loss: 1.4817... 0.0756 sec/batch
Epoch: 5/20... Training Step: 3971... Training loss: 1.5810... 0.0755 sec/batch
Epoch: 5/20... Training Step: 3972... Training loss: 1.5184... 0.0776 sec/batch
Epoch: 5/20... Training Step: 3973... Training loss: 1.4397... 0.0755 sec/batch
Epoch: 5/20... Training Step: 3974... Training loss: 1.4090... 0.0745 sec/batch
Epoch: 5/20... Training Step: 3975... Training loss: 1.4974... 0.0775 sec/batch
Epoch: 5/20... Training Step: 3976... Training loss: 1.5345... 0.0738 sec/batch
Epoch: 5/20... Training Step: 3977... Training loss: 1.5513... 0.0772 sec/batch
Epoch: 5/20... Training Step: 3978... Training loss: 1.4218... 0.0748 sec/batch
Epoch: 5/20... Training Step: 3979... Training loss: 1.5747... 0.0767 sec/batch
Epoch: 5/20... Training Step: 3980... Training loss: 1.4416... 0.0768 sec/batch
Epoch: 5/20... Training Step: 3981... Training loss: 1.4934... 0.0775 sec/batch
Epoch: 5/20... Training Step: 3982... Training loss: 1.5159... 0.0735 sec/batch
Epoch: 5/20... Training Step: 3983... Training loss: 1.5433... 0.0733 sec/batch
Epoch: 5/20... Training Step: 3984... Training loss: 1.4631... 0.0769 sec/batch
Epoch: 5/20... Training Step: 3985... Training loss: 1.5132... 0.0749 sec/batch
Epoch: 5/20... Training Step: 3986... Training loss: 1.4900... 0.0747 sec/batch
Epoch: 5/20... Training Step: 3987... Training loss: 1.5064... 0.0756 sec/batch
Epoch: 5/20... Training Step: 3988... Training loss: 1.4669... 0.0739 sec/batch
Epoch: 5/20... Training Step: 3989... Training loss: 1.5053... 0.0744 sec/batch
Epoch: 5/20... Training Step: 3990... Training loss: 1.4665... 0.0791 sec/batch
Epoch: 5/20... Training Step: 3991... Training loss: 1.4778... 0.0754 sec/batch
Epoch: 5/20... Training Step: 3992... Training loss: 1.4484... 0.0746 sec/batch
Epoch: 5/20... Training Step: 3993... Training loss: 1.5392... 0.0759 sec/batch
Epoch: 5/20... Training Step: 3994... Training loss: 1.4236... 0.0738 sec/batch
Epoch: 5/20... Training Step: 3995... Training loss: 1.4943... 0.0742 sec/batch
Epoch: 5/20... Training Step: 3996... Training loss: 1.4194... 0.0762 sec/batch
Epoch: 5/20... Training Step: 3997... Training loss: 1.5027... 0.0748 sec/batch
Epoch: 5/20... Training Step: 3998... Training loss: 1.5843... 0.0750 sec/batch
Epoch: 5/20... Training Step: 3999... Training loss: 1.5531... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4000... Training loss: 1.5410... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4001... Training loss: 1.4899... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4002... Training loss: 1.5526... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4003... Training loss: 1.6064... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4004... Training loss: 1.5874... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4005... Training loss: 1.5070... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4006... Training loss: 1.5783... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4007... Training loss: 1.4987... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4008... Training loss: 1.4976... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4009... Training loss: 1.4981... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4010... Training loss: 1.4826... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4011... Training loss: 1.4716... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4012... Training loss: 1.4933... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4013... Training loss: 1.5259... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4014... Training loss: 1.4819... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4015... Training loss: 1.5665... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4016... Training loss: 1.4646... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4017... Training loss: 1.5278... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4018... Training loss: 1.5201... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4019... Training loss: 1.3778... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4020... Training loss: 1.5594... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4021... Training loss: 1.4454... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4022... Training loss: 1.4487... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4023... Training loss: 1.4747... 0.0789 sec/batch
Epoch: 5/20... Training Step: 4024... Training loss: 1.4767... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4025... Training loss: 1.5166... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4026... Training loss: 1.5487... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4027... Training loss: 1.4989... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4028... Training loss: 1.5238... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4029... Training loss: 1.4163... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4030... Training loss: 1.4403... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4031... Training loss: 1.5022... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4032... Training loss: 1.4284... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4033... Training loss: 1.5193... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4034... Training loss: 1.4874... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4035... Training loss: 1.4546... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4036... Training loss: 1.4716... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4037... Training loss: 1.5151... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4038... Training loss: 1.4995... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4039... Training loss: 1.5525... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4040... Training loss: 1.5640... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4041... Training loss: 1.4945... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4042... Training loss: 1.4642... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4043... Training loss: 1.4596... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4044... Training loss: 1.4262... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4045... Training loss: 1.4721... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4046... Training loss: 1.4920... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4047... Training loss: 1.5157... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4048... Training loss: 1.5689... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4049... Training loss: 1.5158... 0.0791 sec/batch
Epoch: 5/20... Training Step: 4050... Training loss: 1.4689... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4051... Training loss: 1.5113... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4052... Training loss: 1.5278... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4053... Training loss: 1.4632... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4054... Training loss: 1.5709... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4055... Training loss: 1.4784... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4056... Training loss: 1.5169... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4057... Training loss: 1.5346... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4058... Training loss: 1.5602... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4059... Training loss: 1.5777... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4060... Training loss: 1.4985... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4061... Training loss: 1.4491... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4062... Training loss: 1.3988... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4063... Training loss: 1.4210... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4064... Training loss: 1.5476... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4065... Training loss: 1.4639... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4066... Training loss: 1.4268... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4067... Training loss: 1.4768... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4068... Training loss: 1.4589... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4069... Training loss: 1.4519... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4070... Training loss: 1.4615... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4071... Training loss: 1.5057... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4072... Training loss: 1.5342... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4073... Training loss: 1.5196... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4074... Training loss: 1.5208... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4075... Training loss: 1.4389... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4076... Training loss: 1.5444... 0.0802 sec/batch
Epoch: 5/20... Training Step: 4077... Training loss: 1.4895... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4078... Training loss: 1.5422... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4079... Training loss: 1.5049... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4080... Training loss: 1.4767... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4081... Training loss: 1.5695... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4082... Training loss: 1.5378... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4083... Training loss: 1.4363... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4084... Training loss: 1.6197... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4085... Training loss: 1.5280... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4086... Training loss: 1.6248... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4087... Training loss: 1.5009... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4088... Training loss: 1.5083... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4089... Training loss: 1.6109... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4090... Training loss: 1.4811... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4091... Training loss: 1.5455... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4092... Training loss: 1.5079... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4093... Training loss: 1.5727... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4094... Training loss: 1.5855... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4095... Training loss: 1.5743... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4096... Training loss: 1.5495... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4097... Training loss: 1.4465... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4098... Training loss: 1.5136... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4099... Training loss: 1.4690... 0.0803 sec/batch
Epoch: 5/20... Training Step: 4100... Training loss: 1.5846... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4101... Training loss: 1.5147... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4102... Training loss: 1.4809... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4103... Training loss: 1.5443... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4104... Training loss: 1.5389... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4105... Training loss: 1.5458... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4106... Training loss: 1.5969... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4107... Training loss: 1.5637... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4108... Training loss: 1.4887... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4109... Training loss: 1.5169... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4110... Training loss: 1.4629... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4111... Training loss: 1.4791... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4112... Training loss: 1.5110... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4113... Training loss: 1.3887... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4114... Training loss: 1.5552... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4115... Training loss: 1.4909... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4116... Training loss: 1.4250... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4117... Training loss: 1.4600... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4118... Training loss: 1.4626... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4119... Training loss: 1.3649... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4120... Training loss: 1.4908... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4121... Training loss: 1.4536... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4122... Training loss: 1.5105... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4123... Training loss: 1.4622... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4124... Training loss: 1.4310... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4125... Training loss: 1.4791... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4126... Training loss: 1.4918... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4127... Training loss: 1.5188... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4128... Training loss: 1.5139... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4129... Training loss: 1.4719... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4130... Training loss: 1.4564... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4131... Training loss: 1.4824... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4132... Training loss: 1.5610... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4133... Training loss: 1.5336... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4134... Training loss: 1.6087... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4135... Training loss: 1.5336... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4136... Training loss: 1.5162... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4137... Training loss: 1.5380... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4138... Training loss: 1.4831... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4139... Training loss: 1.5368... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4140... Training loss: 1.4650... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4141... Training loss: 1.4806... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4142... Training loss: 1.5335... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4143... Training loss: 1.4820... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4144... Training loss: 1.5196... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4145... Training loss: 1.5238... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4146... Training loss: 1.6023... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4147... Training loss: 1.5386... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4148... Training loss: 1.4540... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4149... Training loss: 1.5173... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4150... Training loss: 1.5953... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4151... Training loss: 1.4858... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4152... Training loss: 1.5330... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4153... Training loss: 1.5463... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4154... Training loss: 1.5386... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4155... Training loss: 1.5317... 0.0791 sec/batch
Epoch: 5/20... Training Step: 4156... Training loss: 1.4957... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4157... Training loss: 1.5023... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4158... Training loss: 1.5566... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4159... Training loss: 1.5992... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4160... Training loss: 1.5555... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4161... Training loss: 1.5542... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4162... Training loss: 1.6125... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4163... Training loss: 1.4932... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4164... Training loss: 1.4559... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4165... Training loss: 1.5710... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4166... Training loss: 1.5085... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4167... Training loss: 1.5468... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4168... Training loss: 1.5067... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4169... Training loss: 1.5027... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4170... Training loss: 1.5195... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4171... Training loss: 1.5094... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4172... Training loss: 1.4812... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4173... Training loss: 1.5100... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4174... Training loss: 1.5743... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4175... Training loss: 1.5003... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4176... Training loss: 1.4810... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4177... Training loss: 1.5251... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4178... Training loss: 1.5175... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4179... Training loss: 1.4625... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4180... Training loss: 1.5736... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4181... Training loss: 1.4894... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4182... Training loss: 1.5264... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4183... Training loss: 1.4824... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4184... Training loss: 1.5040... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4185... Training loss: 1.4617... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4186... Training loss: 1.5209... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4187... Training loss: 1.5087... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4188... Training loss: 1.4767... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4189... Training loss: 1.4665... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4190... Training loss: 1.5854... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4191... Training loss: 1.5076... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4192... Training loss: 1.4865... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4193... Training loss: 1.5375... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4194... Training loss: 1.4803... 0.0800 sec/batch
Epoch: 5/20... Training Step: 4195... Training loss: 1.5185... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4196... Training loss: 1.5202... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4197... Training loss: 1.4202... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4198... Training loss: 1.4495... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4199... Training loss: 1.4851... 0.0785 sec/batch
Epoch: 5/20... Training Step: 4200... Training loss: 1.4868... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4201... Training loss: 1.4845... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4202... Training loss: 1.5866... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4203... Training loss: 1.4679... 0.0804 sec/batch
Epoch: 5/20... Training Step: 4204... Training loss: 1.4978... 0.0844 sec/batch
Epoch: 5/20... Training Step: 4205... Training loss: 1.4629... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4206... Training loss: 1.5369... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4207... Training loss: 1.5452... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4208... Training loss: 1.4655... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4209... Training loss: 1.4940... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4210... Training loss: 1.4934... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4211... Training loss: 1.5425... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4212... Training loss: 1.5630... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4213... Training loss: 1.5072... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4214... Training loss: 1.5356... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4215... Training loss: 1.4496... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4216... Training loss: 1.4191... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4217... Training loss: 1.4816... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4218... Training loss: 1.5069... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4219... Training loss: 1.4808... 0.0809 sec/batch
Epoch: 5/20... Training Step: 4220... Training loss: 1.4869... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4221... Training loss: 1.4890... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4222... Training loss: 1.4757... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4223... Training loss: 1.4439... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4224... Training loss: 1.4645... 0.0808 sec/batch
Epoch: 5/20... Training Step: 4225... Training loss: 1.6299... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4226... Training loss: 1.4707... 0.0776 sec/batch
Epoch: 5/20... Training Step: 4227... Training loss: 1.4863... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4228... Training loss: 1.4837... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4229... Training loss: 1.5441... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4230... Training loss: 1.5403... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4231... Training loss: 1.5376... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4232... Training loss: 1.5155... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4233... Training loss: 1.5307... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4234... Training loss: 1.4431... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4235... Training loss: 1.4533... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4236... Training loss: 1.5066... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4237... Training loss: 1.4333... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4238... Training loss: 1.4221... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4239... Training loss: 1.4869... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4240... Training loss: 1.4638... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4241... Training loss: 1.3884... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4242... Training loss: 1.4137... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4243... Training loss: 1.4750... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4244... Training loss: 1.4664... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4245... Training loss: 1.4864... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4246... Training loss: 1.3924... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4247... Training loss: 1.4503... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4248... Training loss: 1.5599... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4249... Training loss: 1.4813... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4250... Training loss: 1.4890... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4251... Training loss: 1.5048... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4252... Training loss: 1.4506... 0.0829 sec/batch
Epoch: 5/20... Training Step: 4253... Training loss: 1.4043... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4254... Training loss: 1.4479... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4255... Training loss: 1.5237... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4256... Training loss: 1.4486... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4257... Training loss: 1.4615... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4258... Training loss: 1.4773... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4259... Training loss: 1.4579... 0.0798 sec/batch
Epoch: 5/20... Training Step: 4260... Training loss: 1.4775... 0.0791 sec/batch
Epoch: 5/20... Training Step: 4261... Training loss: 1.4104... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4262... Training loss: 1.4226... 0.0799 sec/batch
Epoch: 5/20... Training Step: 4263... Training loss: 1.5063... 0.0808 sec/batch
Epoch: 5/20... Training Step: 4264... Training loss: 1.4940... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4265... Training loss: 1.4316... 0.0816 sec/batch
Epoch: 5/20... Training Step: 4266... Training loss: 1.5625... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4267... Training loss: 1.4946... 0.0811 sec/batch
Epoch: 5/20... Training Step: 4268... Training loss: 1.5113... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4269... Training loss: 1.5428... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4270... Training loss: 1.4569... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4271... Training loss: 1.4398... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4272... Training loss: 1.5040... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4273... Training loss: 1.4874... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4274... Training loss: 1.5603... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4275... Training loss: 1.4408... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4276... Training loss: 1.3987... 0.0798 sec/batch
Epoch: 5/20... Training Step: 4277... Training loss: 1.5663... 0.0812 sec/batch
Epoch: 5/20... Training Step: 4278... Training loss: 1.4812... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4279... Training loss: 1.5462... 0.0811 sec/batch
Epoch: 5/20... Training Step: 4280... Training loss: 1.5024... 0.0804 sec/batch
Epoch: 5/20... Training Step: 4281... Training loss: 1.4886... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4282... Training loss: 1.4803... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4283... Training loss: 1.5105... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4284... Training loss: 1.4279... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4285... Training loss: 1.4344... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4286... Training loss: 1.4446... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4287... Training loss: 1.4402... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4288... Training loss: 1.5251... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4289... Training loss: 1.5051... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4290... Training loss: 1.4112... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4291... Training loss: 1.4420... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4292... Training loss: 1.4620... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4293... Training loss: 1.5047... 0.0805 sec/batch
Epoch: 5/20... Training Step: 4294... Training loss: 1.4677... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4295... Training loss: 1.4624... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4296... Training loss: 1.4518... 0.0824 sec/batch
Epoch: 5/20... Training Step: 4297... Training loss: 1.4849... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4298... Training loss: 1.5071... 0.0814 sec/batch
Epoch: 5/20... Training Step: 4299... Training loss: 1.5097... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4300... Training loss: 1.4465... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4301... Training loss: 1.5627... 0.0816 sec/batch
Epoch: 5/20... Training Step: 4302... Training loss: 1.5619... 0.0805 sec/batch
Epoch: 5/20... Training Step: 4303... Training loss: 1.5344... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4304... Training loss: 1.5067... 0.0798 sec/batch
Epoch: 5/20... Training Step: 4305... Training loss: 1.4782... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4306... Training loss: 1.4618... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4307... Training loss: 1.5138... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4308... Training loss: 1.4714... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4309... Training loss: 1.4936... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4310... Training loss: 1.4508... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4311... Training loss: 1.4514... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4312... Training loss: 1.5182... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4313... Training loss: 1.4633... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4314... Training loss: 1.5331... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4315... Training loss: 1.5274... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4316... Training loss: 1.5144... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4317... Training loss: 1.5395... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4318... Training loss: 1.5703... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4319... Training loss: 1.5312... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4320... Training loss: 1.4458... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4321... Training loss: 1.4721... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4322... Training loss: 1.4631... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4323... Training loss: 1.5062... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4324... Training loss: 1.4961... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4325... Training loss: 1.4750... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4326... Training loss: 1.4983... 0.0810 sec/batch
Epoch: 5/20... Training Step: 4327... Training loss: 1.4893... 0.0807 sec/batch
Epoch: 5/20... Training Step: 4328... Training loss: 1.4664... 0.0785 sec/batch
Epoch: 5/20... Training Step: 4329... Training loss: 1.5174... 0.0967 sec/batch
Epoch: 5/20... Training Step: 4330... Training loss: 1.5828... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4331... Training loss: 1.4750... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4332... Training loss: 1.4721... 0.0804 sec/batch
Epoch: 5/20... Training Step: 4333... Training loss: 1.4670... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4334... Training loss: 1.4623... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4335... Training loss: 1.4658... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4336... Training loss: 1.4191... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4337... Training loss: 1.4365... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4338... Training loss: 1.4098... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4339... Training loss: 1.5413... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4340... Training loss: 1.4110... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4341... Training loss: 1.4896... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4342... Training loss: 1.4083... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4343... Training loss: 1.4997... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4344... Training loss: 1.4430... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4345... Training loss: 1.4477... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4346... Training loss: 1.4806... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4347... Training loss: 1.4704... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4348... Training loss: 1.4863... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4349... Training loss: 1.4614... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4350... Training loss: 1.4474... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4351... Training loss: 1.5085... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4352... Training loss: 1.4884... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4353... Training loss: 1.5271... 0.0810 sec/batch
Epoch: 5/20... Training Step: 4354... Training loss: 1.4779... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4355... Training loss: 1.5230... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4356... Training loss: 1.5481... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4357... Training loss: 1.5196... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4358... Training loss: 1.4418... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4359... Training loss: 1.4990... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4360... Training loss: 1.4790... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4361... Training loss: 1.5124... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4362... Training loss: 1.4506... 0.0804 sec/batch
Epoch: 5/20... Training Step: 4363... Training loss: 1.4901... 0.0799 sec/batch
Epoch: 5/20... Training Step: 4364... Training loss: 1.4718... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4365... Training loss: 1.4342... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4366... Training loss: 1.5434... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4367... Training loss: 1.5068... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4368... Training loss: 1.4446... 0.0785 sec/batch
Epoch: 5/20... Training Step: 4369... Training loss: 1.4303... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4370... Training loss: 1.5090... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4371... Training loss: 1.4460... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4372... Training loss: 1.4933... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4373... Training loss: 1.4703... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4374... Training loss: 1.4905... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4375... Training loss: 1.4424... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4376... Training loss: 1.5345... 0.0832 sec/batch
Epoch: 5/20... Training Step: 4377... Training loss: 1.4323... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4378... Training loss: 1.4449... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4379... Training loss: 1.3880... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4380... Training loss: 1.4240... 0.0789 sec/batch
Epoch: 5/20... Training Step: 4381... Training loss: 1.4618... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4382... Training loss: 1.4133... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4383... Training loss: 1.4233... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4384... Training loss: 1.5205... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4385... Training loss: 1.3855... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4386... Training loss: 1.4385... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4387... Training loss: 1.5111... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4388... Training loss: 1.5887... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4389... Training loss: 1.4421... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4390... Training loss: 1.4767... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4391... Training loss: 1.4734... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4392... Training loss: 1.4892... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4393... Training loss: 1.5037... 0.0802 sec/batch
Epoch: 5/20... Training Step: 4394... Training loss: 1.4877... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4395... Training loss: 1.5746... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4396... Training loss: 1.5839... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4397... Training loss: 1.5340... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4398... Training loss: 1.5945... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4399... Training loss: 1.4895... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4400... Training loss: 1.5165... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4401... Training loss: 1.5535... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4402... Training loss: 1.5092... 0.0785 sec/batch
Epoch: 5/20... Training Step: 4403... Training loss: 1.5365... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4404... Training loss: 1.4526... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4405... Training loss: 1.5232... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4406... Training loss: 1.6028... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4407... Training loss: 1.4851... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4408... Training loss: 1.5458... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4409... Training loss: 1.4920... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4410... Training loss: 1.5142... 0.0800 sec/batch
Epoch: 5/20... Training Step: 4411... Training loss: 1.5123... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4412... Training loss: 1.5080... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4413... Training loss: 1.4571... 0.0791 sec/batch
Epoch: 5/20... Training Step: 4414... Training loss: 1.5611... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4415... Training loss: 1.4762... 0.0856 sec/batch
Epoch: 5/20... Training Step: 4416... Training loss: 1.5073... 0.0855 sec/batch
Epoch: 5/20... Training Step: 4417... Training loss: 1.5818... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4418... Training loss: 1.5661... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4419... Training loss: 1.4634... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4420... Training loss: 1.4955... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4421... Training loss: 1.5099... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4422... Training loss: 1.5262... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4423... Training loss: 1.5158... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4424... Training loss: 1.4787... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4425... Training loss: 1.4301... 0.0776 sec/batch
Epoch: 5/20... Training Step: 4426... Training loss: 1.4849... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4427... Training loss: 1.4753... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4428... Training loss: 1.5050... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4429... Training loss: 1.5323... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4430... Training loss: 1.4741... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4431... Training loss: 1.4363... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4432... Training loss: 1.4887... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4433... Training loss: 1.4508... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4434... Training loss: 1.4562... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4435... Training loss: 1.4051... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4436... Training loss: 1.4973... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4437... Training loss: 1.4833... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4438... Training loss: 1.5579... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4439... Training loss: 1.4970... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4440... Training loss: 1.4828... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4441... Training loss: 1.4701... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4442... Training loss: 1.4877... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4443... Training loss: 1.3361... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4444... Training loss: 1.4626... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4445... Training loss: 1.4660... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4446... Training loss: 1.4750... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4447... Training loss: 1.4492... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4448... Training loss: 1.4403... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4449... Training loss: 1.4294... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4450... Training loss: 1.4979... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4451... Training loss: 1.4789... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4452... Training loss: 1.4806... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4453... Training loss: 1.4028... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4454... Training loss: 1.4417... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4455... Training loss: 1.5046... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4456... Training loss: 1.4465... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4457... Training loss: 1.4591... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4458... Training loss: 1.4309... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4459... Training loss: 1.4359... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4460... Training loss: 1.4212... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4461... Training loss: 1.4614... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4462... Training loss: 1.4650... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4463... Training loss: 1.4286... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4464... Training loss: 1.4155... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4465... Training loss: 1.4722... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4466... Training loss: 1.4699... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4467... Training loss: 1.4582... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4468... Training loss: 1.4882... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4469... Training loss: 1.4466... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4470... Training loss: 1.5176... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4471... Training loss: 1.4526... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4472... Training loss: 1.4535... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4473... Training loss: 1.4867... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4474... Training loss: 1.4069... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4475... Training loss: 1.4471... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4476... Training loss: 1.4292... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4477... Training loss: 1.4317... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4478... Training loss: 1.4257... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4479... Training loss: 1.4840... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4480... Training loss: 1.4903... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4481... Training loss: 1.5459... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4482... Training loss: 1.5237... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4483... Training loss: 1.4725... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4484... Training loss: 1.4886... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4485... Training loss: 1.4950... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4486... Training loss: 1.4566... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4487... Training loss: 1.4675... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4488... Training loss: 1.4166... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4489... Training loss: 1.4738... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4490... Training loss: 1.4326... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4491... Training loss: 1.5509... 0.0793 sec/batch
Epoch: 5/20... Training Step: 4492... Training loss: 1.4441... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4493... Training loss: 1.4766... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4494... Training loss: 1.4857... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4495... Training loss: 1.4776... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4496... Training loss: 1.4397... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4497... Training loss: 1.4587... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4498... Training loss: 1.5392... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4499... Training loss: 1.5947... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4500... Training loss: 1.4947... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4501... Training loss: 1.5269... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4502... Training loss: 1.5236... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4503... Training loss: 1.4893... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4504... Training loss: 1.4402... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4505... Training loss: 1.4499... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4506... Training loss: 1.4742... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4507... Training loss: 1.4583... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4508... Training loss: 1.4643... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4509... Training loss: 1.5727... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4510... Training loss: 1.4409... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4511... Training loss: 1.4506... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4512... Training loss: 1.4492... 0.0797 sec/batch
Epoch: 5/20... Training Step: 4513... Training loss: 1.4370... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4514... Training loss: 1.4153... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4515... Training loss: 1.5409... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4516... Training loss: 1.5031... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4517... Training loss: 1.4325... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4518... Training loss: 1.4690... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4519... Training loss: 1.4478... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4520... Training loss: 1.4182... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4521... Training loss: 1.4015... 0.0820 sec/batch
Epoch: 5/20... Training Step: 4522... Training loss: 1.4494... 0.0808 sec/batch
Epoch: 5/20... Training Step: 4523... Training loss: 1.4767... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4524... Training loss: 1.4150... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4525... Training loss: 1.4402... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4526... Training loss: 1.4494... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4527... Training loss: 1.4725... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4528... Training loss: 1.3994... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4529... Training loss: 1.4482... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4530... Training loss: 1.4662... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4531... Training loss: 1.4278... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4532... Training loss: 1.4338... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4533... Training loss: 1.4694... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4534... Training loss: 1.4952... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4535... Training loss: 1.4109... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4536... Training loss: 1.4012... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4537... Training loss: 1.3797... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4538... Training loss: 1.3841... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4539... Training loss: 1.4268... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4540... Training loss: 1.4257... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4541... Training loss: 1.5259... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4542... Training loss: 1.4898... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4543... Training loss: 1.4849... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4544... Training loss: 1.4715... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4545... Training loss: 1.4640... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4546... Training loss: 1.6028... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4547... Training loss: 1.5032... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4548... Training loss: 1.4379... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4549... Training loss: 1.4085... 0.0776 sec/batch
Epoch: 5/20... Training Step: 4550... Training loss: 1.4352... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4551... Training loss: 1.5215... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4552... Training loss: 1.3930... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4553... Training loss: 1.4417... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4554... Training loss: 1.4848... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4555... Training loss: 1.4756... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4556... Training loss: 1.5016... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4557... Training loss: 1.4359... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4558... Training loss: 1.4580... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4559... Training loss: 1.4682... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4560... Training loss: 1.5127... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4561... Training loss: 1.4351... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4562... Training loss: 1.4397... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4563... Training loss: 1.4040... 0.0793 sec/batch
Epoch: 5/20... Training Step: 4564... Training loss: 1.5055... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4565... Training loss: 1.3500... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4566... Training loss: 1.3727... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4567... Training loss: 1.4389... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4568... Training loss: 1.4379... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4569... Training loss: 1.5101... 0.0830 sec/batch
Epoch: 5/20... Training Step: 4570... Training loss: 1.4414... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4571... Training loss: 1.4512... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4572... Training loss: 1.4954... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4573... Training loss: 1.4264... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4574... Training loss: 1.4398... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4575... Training loss: 1.4915... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4576... Training loss: 1.4833... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4577... Training loss: 1.4834... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4578... Training loss: 1.5158... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4579... Training loss: 1.3879... 0.0704 sec/batch
Epoch: 5/20... Training Step: 4580... Training loss: 1.4914... 0.0810 sec/batch
Epoch: 5/20... Training Step: 4581... Training loss: 1.5187... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4582... Training loss: 1.4311... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4583... Training loss: 1.5116... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4584... Training loss: 1.4654... 0.0873 sec/batch
Epoch: 5/20... Training Step: 4585... Training loss: 1.4793... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4586... Training loss: 1.5167... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4587... Training loss: 1.3988... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4588... Training loss: 1.4492... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4589... Training loss: 1.4353... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4590... Training loss: 1.3598... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4591... Training loss: 1.5037... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4592... Training loss: 1.4690... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4593... Training loss: 1.4036... 0.0781 sec/batch
Epoch: 5/20... Training Step: 4594... Training loss: 1.4085... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4595... Training loss: 1.4290... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4596... Training loss: 1.4860... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4597... Training loss: 1.2724... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4598... Training loss: 1.4683... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4599... Training loss: 1.4459... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4600... Training loss: 1.4348... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4601... Training loss: 1.4189... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4602... Training loss: 1.4596... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4603... Training loss: 1.4714... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4604... Training loss: 1.4296... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4605... Training loss: 1.5202... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4606... Training loss: 1.4565... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4607... Training loss: 1.4749... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4608... Training loss: 1.5432... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4609... Training loss: 1.5714... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4610... Training loss: 1.5047... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4611... Training loss: 1.4628... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4612... Training loss: 1.4611... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4613... Training loss: 1.4252... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4614... Training loss: 1.4891... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4615... Training loss: 1.4819... 0.0799 sec/batch
Epoch: 5/20... Training Step: 4616... Training loss: 1.3823... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4617... Training loss: 1.4207... 0.0801 sec/batch
Epoch: 5/20... Training Step: 4618... Training loss: 1.4534... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4619... Training loss: 1.4529... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4620... Training loss: 1.5218... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4621... Training loss: 1.4642... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4622... Training loss: 1.5282... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4623... Training loss: 1.5532... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4624... Training loss: 1.5097... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4625... Training loss: 1.4000... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4626... Training loss: 1.4446... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4627... Training loss: 1.4560... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4628... Training loss: 1.5318... 0.0831 sec/batch
Epoch: 5/20... Training Step: 4629... Training loss: 1.5269... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4630... Training loss: 1.5643... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4631... Training loss: 1.4750... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4632... Training loss: 1.5937... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4633... Training loss: 1.5742... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4634... Training loss: 1.4197... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4635... Training loss: 1.5723... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4636... Training loss: 1.5153... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4637... Training loss: 1.4795... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4638... Training loss: 1.4434... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4639... Training loss: 1.5133... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4640... Training loss: 1.4928... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4641... Training loss: 1.4523... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4642... Training loss: 1.4828... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4643... Training loss: 1.4111... 0.0809 sec/batch
Epoch: 5/20... Training Step: 4644... Training loss: 1.4458... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4645... Training loss: 1.5688... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4646... Training loss: 1.4397... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4647... Training loss: 1.4209... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4648... Training loss: 1.4983... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4649... Training loss: 1.5243... 0.0809 sec/batch
Epoch: 5/20... Training Step: 4650... Training loss: 1.4802... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4651... Training loss: 1.4849... 0.0797 sec/batch
Epoch: 5/20... Training Step: 4652... Training loss: 1.4212... 0.0797 sec/batch
Epoch: 5/20... Training Step: 4653... Training loss: 1.5371... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4654... Training loss: 1.5101... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4655... Training loss: 1.4387... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4656... Training loss: 1.4697... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4657... Training loss: 1.4391... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4658... Training loss: 1.5132... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4659... Training loss: 1.4629... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4660... Training loss: 1.4639... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4661... Training loss: 1.4402... 0.0807 sec/batch
Epoch: 5/20... Training Step: 4662... Training loss: 1.4170... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4663... Training loss: 1.4950... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4664... Training loss: 1.4788... 0.0849 sec/batch
Epoch: 5/20... Training Step: 4665... Training loss: 1.5429... 0.0828 sec/batch
Epoch: 5/20... Training Step: 4666... Training loss: 1.5021... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4667... Training loss: 1.4699... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4668... Training loss: 1.4875... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4669... Training loss: 1.4673... 0.0797 sec/batch
Epoch: 5/20... Training Step: 4670... Training loss: 1.4543... 0.0789 sec/batch
Epoch: 5/20... Training Step: 4671... Training loss: 1.4502... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4672... Training loss: 1.3946... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4673... Training loss: 1.3891... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4674... Training loss: 1.4305... 0.0833 sec/batch
Epoch: 5/20... Training Step: 4675... Training loss: 1.5025... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4676... Training loss: 1.4955... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4677... Training loss: 1.4985... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4678... Training loss: 1.4390... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4679... Training loss: 1.4573... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4680... Training loss: 1.5592... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4681... Training loss: 1.4333... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4682... Training loss: 1.4498... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4683... Training loss: 1.4702... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4684... Training loss: 1.5109... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4685... Training loss: 1.5188... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4686... Training loss: 1.4851... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4687... Training loss: 1.4991... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4688... Training loss: 1.5778... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4689... Training loss: 1.4933... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4690... Training loss: 1.4608... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4691... Training loss: 1.4512... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4692... Training loss: 1.5057... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4693... Training loss: 1.5188... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4694... Training loss: 1.5550... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4695... Training loss: 1.4384... 0.0816 sec/batch
Epoch: 5/20... Training Step: 4696... Training loss: 1.5200... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4697... Training loss: 1.5026... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4698... Training loss: 1.4517... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4699... Training loss: 1.4583... 0.0811 sec/batch
Epoch: 5/20... Training Step: 4700... Training loss: 1.4891... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4701... Training loss: 1.4294... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4702... Training loss: 1.4970... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4703... Training loss: 1.4444... 0.0772 sec/batch
Epoch: 5/20... Training Step: 4704... Training loss: 1.4460... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4705... Training loss: 1.4035... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4706... Training loss: 1.5472... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4707... Training loss: 1.4611... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4708... Training loss: 1.4948... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4709... Training loss: 1.4236... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4710... Training loss: 1.4329... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4711... Training loss: 1.4546... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4712... Training loss: 1.5412... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4713... Training loss: 1.4826... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4714... Training loss: 1.4378... 0.0919 sec/batch
Epoch: 5/20... Training Step: 4715... Training loss: 1.4869... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4716... Training loss: 1.4854... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4717... Training loss: 1.4087... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4718... Training loss: 1.4021... 0.0789 sec/batch
Epoch: 5/20... Training Step: 4719... Training loss: 1.5698... 0.0785 sec/batch
Epoch: 5/20... Training Step: 4720... Training loss: 1.4418... 0.0774 sec/batch
Epoch: 5/20... Training Step: 4721... Training loss: 1.4908... 0.0825 sec/batch
Epoch: 5/20... Training Step: 4722... Training loss: 1.4803... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4723... Training loss: 1.3905... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4724... Training loss: 1.4420... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4725... Training loss: 1.4664... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4726... Training loss: 1.5307... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4727... Training loss: 1.5308... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4728... Training loss: 1.5672... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4729... Training loss: 1.5197... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4730... Training loss: 1.3950... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4731... Training loss: 1.4706... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4732... Training loss: 1.4688... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4733... Training loss: 1.4887... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4734... Training loss: 1.5147... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4735... Training loss: 1.4775... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4736... Training loss: 1.4517... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4737... Training loss: 1.4381... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4738... Training loss: 1.4846... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4739... Training loss: 1.4751... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4740... Training loss: 1.5416... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4741... Training loss: 1.5244... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4742... Training loss: 1.5131... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4743... Training loss: 1.4743... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4744... Training loss: 1.3657... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4745... Training loss: 1.4405... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4746... Training loss: 1.4782... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4747... Training loss: 1.4231... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4748... Training loss: 1.4262... 0.0794 sec/batch
Epoch: 5/20... Training Step: 4749... Training loss: 1.4463... 0.0734 sec/batch
Epoch: 5/20... Training Step: 4750... Training loss: 1.4680... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4751... Training loss: 1.4939... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4752... Training loss: 1.4889... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4753... Training loss: 1.4784... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4754... Training loss: 1.4635... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4755... Training loss: 1.4916... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4756... Training loss: 1.4874... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4757... Training loss: 1.4837... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4758... Training loss: 1.4630... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4759... Training loss: 1.4623... 0.0790 sec/batch
Epoch: 5/20... Training Step: 4760... Training loss: 1.4262... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4761... Training loss: 1.4648... 0.0802 sec/batch
Epoch: 5/20... Training Step: 4762... Training loss: 1.4197... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4763... Training loss: 1.3764... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4764... Training loss: 1.4495... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4765... Training loss: 1.4490... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4766... Training loss: 1.4593... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4767... Training loss: 1.4532... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4768... Training loss: 1.4031... 0.0793 sec/batch
Epoch: 5/20... Training Step: 4769... Training loss: 1.5008... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4770... Training loss: 1.5224... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4771... Training loss: 1.5254... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4772... Training loss: 1.4731... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4773... Training loss: 1.4182... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4774... Training loss: 1.4625... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4775... Training loss: 1.4723... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4776... Training loss: 1.5009... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4777... Training loss: 1.5209... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4778... Training loss: 1.5447... 0.0776 sec/batch
Epoch: 5/20... Training Step: 4779... Training loss: 1.4706... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4780... Training loss: 1.5280... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4781... Training loss: 1.4728... 0.0783 sec/batch
Epoch: 5/20... Training Step: 4782... Training loss: 1.5130... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4783... Training loss: 1.5256... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4784... Training loss: 1.5669... 0.0776 sec/batch
Epoch: 5/20... Training Step: 4785... Training loss: 1.5285... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4786... Training loss: 1.6091... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4787... Training loss: 1.5147... 0.0788 sec/batch
Epoch: 5/20... Training Step: 4788... Training loss: 1.4620... 0.0732 sec/batch
Epoch: 5/20... Training Step: 4789... Training loss: 1.4831... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4790... Training loss: 1.4013... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4791... Training loss: 1.4705... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4792... Training loss: 1.4710... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4793... Training loss: 1.4370... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4794... Training loss: 1.4970... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4795... Training loss: 1.4943... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4796... Training loss: 1.5390... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4797... Training loss: 1.5163... 0.0731 sec/batch
Epoch: 5/20... Training Step: 4798... Training loss: 1.4359... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4799... Training loss: 1.5477... 0.0822 sec/batch
Epoch: 5/20... Training Step: 4800... Training loss: 1.5519... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4801... Training loss: 1.5423... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4802... Training loss: 1.4884... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4803... Training loss: 1.4118... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4804... Training loss: 1.4526... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4805... Training loss: 1.5103... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4806... Training loss: 1.5056... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4807... Training loss: 1.5322... 0.0786 sec/batch
Epoch: 5/20... Training Step: 4808... Training loss: 1.4655... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4809... Training loss: 1.5057... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4810... Training loss: 1.6100... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4811... Training loss: 1.5360... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4812... Training loss: 1.5381... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4813... Training loss: 1.5608... 0.0793 sec/batch
Epoch: 5/20... Training Step: 4814... Training loss: 1.4315... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4815... Training loss: 1.4777... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4816... Training loss: 1.4626... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4817... Training loss: 1.4895... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4818... Training loss: 1.4893... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4819... Training loss: 1.4628... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4820... Training loss: 1.5655... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4821... Training loss: 1.5462... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4822... Training loss: 1.4989... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4823... Training loss: 1.4799... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4824... Training loss: 1.4774... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4825... Training loss: 1.4800... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4826... Training loss: 1.4633... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4827... Training loss: 1.5049... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4828... Training loss: 1.5055... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4829... Training loss: 1.4894... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4830... Training loss: 1.4464... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4831... Training loss: 1.4317... 0.0780 sec/batch
Epoch: 5/20... Training Step: 4832... Training loss: 1.4526... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4833... Training loss: 1.4210... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4834... Training loss: 1.4801... 0.0789 sec/batch
Epoch: 5/20... Training Step: 4835... Training loss: 1.4719... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4836... Training loss: 1.5601... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4837... Training loss: 1.4799... 0.0764 sec/batch
Epoch: 5/20... Training Step: 4838... Training loss: 1.5582... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4839... Training loss: 1.5406... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4840... Training loss: 1.5121... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4841... Training loss: 1.5334... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4842... Training loss: 1.4903... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4843... Training loss: 1.5120... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4844... Training loss: 1.4875... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4845... Training loss: 1.4930... 0.0746 sec/batch
Epoch: 5/20... Training Step: 4846... Training loss: 1.5453... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4847... Training loss: 1.4938... 0.0750 sec/batch
Epoch: 5/20... Training Step: 4848... Training loss: 1.4417... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4849... Training loss: 1.4525... 0.0779 sec/batch
Epoch: 5/20... Training Step: 4850... Training loss: 1.4140... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4851... Training loss: 1.5043... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4852... Training loss: 1.4723... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4853... Training loss: 1.4742... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4854... Training loss: 1.5577... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4855... Training loss: 1.4751... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4856... Training loss: 1.5536... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4857... Training loss: 1.5115... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4858... Training loss: 1.5451... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4859... Training loss: 1.5394... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4860... Training loss: 1.4243... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4861... Training loss: 1.4630... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4862... Training loss: 1.4138... 0.0735 sec/batch
Epoch: 5/20... Training Step: 4863... Training loss: 1.5387... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4864... Training loss: 1.4587... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4865... Training loss: 1.4909... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4866... Training loss: 1.5189... 0.0749 sec/batch
Epoch: 5/20... Training Step: 4867... Training loss: 1.4654... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4868... Training loss: 1.4248... 0.0736 sec/batch
Epoch: 5/20... Training Step: 4869... Training loss: 1.4237... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4870... Training loss: 1.4365... 0.0766 sec/batch
Epoch: 5/20... Training Step: 4871... Training loss: 1.4505... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4872... Training loss: 1.4083... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4873... Training loss: 1.4044... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4874... Training loss: 1.4928... 0.0855 sec/batch
Epoch: 5/20... Training Step: 4875... Training loss: 1.3876... 0.0733 sec/batch
Epoch: 5/20... Training Step: 4876... Training loss: 1.5085... 0.0775 sec/batch
Epoch: 5/20... Training Step: 4877... Training loss: 1.4232... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4878... Training loss: 1.3610... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4879... Training loss: 1.4075... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4880... Training loss: 1.4500... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4881... Training loss: 1.5795... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4882... Training loss: 1.4774... 0.0840 sec/batch
Epoch: 5/20... Training Step: 4883... Training loss: 1.4634... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4884... Training loss: 1.4545... 0.0765 sec/batch
Epoch: 5/20... Training Step: 4885... Training loss: 1.4926... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4886... Training loss: 1.5124... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4887... Training loss: 1.4562... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4888... Training loss: 1.5067... 0.0837 sec/batch
Epoch: 5/20... Training Step: 4889... Training loss: 1.5793... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4890... Training loss: 1.5121... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4891... Training loss: 1.4855... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4892... Training loss: 1.4575... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4893... Training loss: 1.5108... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4894... Training loss: 1.5312... 0.0770 sec/batch
Epoch: 5/20... Training Step: 4895... Training loss: 1.4711... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4896... Training loss: 1.4916... 0.0771 sec/batch
Epoch: 5/20... Training Step: 4897... Training loss: 1.5680... 0.0763 sec/batch
Epoch: 5/20... Training Step: 4898... Training loss: 1.5310... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4899... Training loss: 1.4893... 0.0756 sec/batch
Epoch: 5/20... Training Step: 4900... Training loss: 1.5761... 0.0838 sec/batch
Epoch: 5/20... Training Step: 4901... Training loss: 1.4594... 0.0741 sec/batch
Epoch: 5/20... Training Step: 4902... Training loss: 1.4834... 0.0752 sec/batch
Epoch: 5/20... Training Step: 4903... Training loss: 1.4972... 0.0751 sec/batch
Epoch: 5/20... Training Step: 4904... Training loss: 1.5035... 0.0759 sec/batch
Epoch: 5/20... Training Step: 4905... Training loss: 1.4406... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4906... Training loss: 1.4877... 0.0757 sec/batch
Epoch: 5/20... Training Step: 4907... Training loss: 1.4667... 0.0742 sec/batch
Epoch: 5/20... Training Step: 4908... Training loss: 1.5702... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4909... Training loss: 1.4821... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4910... Training loss: 1.5090... 0.0761 sec/batch
Epoch: 5/20... Training Step: 4911... Training loss: 1.5420... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4912... Training loss: 1.5067... 0.0753 sec/batch
Epoch: 5/20... Training Step: 4913... Training loss: 1.5347... 0.0737 sec/batch
Epoch: 5/20... Training Step: 4914... Training loss: 1.4959... 0.0743 sec/batch
Epoch: 5/20... Training Step: 4915... Training loss: 1.4599... 0.0791 sec/batch
Epoch: 5/20... Training Step: 4916... Training loss: 1.4384... 0.0748 sec/batch
Epoch: 5/20... Training Step: 4917... Training loss: 1.4822... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4918... Training loss: 1.3688... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4919... Training loss: 1.4658... 0.0769 sec/batch
Epoch: 5/20... Training Step: 4920... Training loss: 1.4334... 0.0745 sec/batch
Epoch: 5/20... Training Step: 4921... Training loss: 1.4704... 0.0799 sec/batch
Epoch: 5/20... Training Step: 4922... Training loss: 1.4905... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4923... Training loss: 1.5031... 0.0739 sec/batch
Epoch: 5/20... Training Step: 4924... Training loss: 1.5032... 0.0762 sec/batch
Epoch: 5/20... Training Step: 4925... Training loss: 1.4647... 0.0754 sec/batch
Epoch: 5/20... Training Step: 4926... Training loss: 1.5253... 0.0755 sec/batch
Epoch: 5/20... Training Step: 4927... Training loss: 1.4242... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4928... Training loss: 1.4470... 0.0744 sec/batch
Epoch: 5/20... Training Step: 4929... Training loss: 1.4932... 0.0793 sec/batch
Epoch: 5/20... Training Step: 4930... Training loss: 1.4715... 0.0918 sec/batch
Epoch: 5/20... Training Step: 4931... Training loss: 1.5078... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4932... Training loss: 1.4368... 0.0809 sec/batch
Epoch: 5/20... Training Step: 4933... Training loss: 1.5506... 0.0767 sec/batch
Epoch: 5/20... Training Step: 4934... Training loss: 1.5662... 0.0784 sec/batch
Epoch: 5/20... Training Step: 4935... Training loss: 1.5422... 0.0787 sec/batch
Epoch: 5/20... Training Step: 4936... Training loss: 1.5149... 0.0782 sec/batch
Epoch: 5/20... Training Step: 4937... Training loss: 1.5128... 0.0819 sec/batch
Epoch: 5/20... Training Step: 4938... Training loss: 1.6672... 0.0807 sec/batch
Epoch: 5/20... Training Step: 4939... Training loss: 1.4213... 0.0795 sec/batch
Epoch: 5/20... Training Step: 4940... Training loss: 1.5732... 0.0768 sec/batch
Epoch: 5/20... Training Step: 4941... Training loss: 1.5863... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4942... Training loss: 1.4605... 0.0802 sec/batch
Epoch: 5/20... Training Step: 4943... Training loss: 1.5321... 0.0801 sec/batch
Epoch: 5/20... Training Step: 4944... Training loss: 1.5491... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4945... Training loss: 1.5384... 0.0909 sec/batch
Epoch: 5/20... Training Step: 4946... Training loss: 1.5222... 0.0811 sec/batch
Epoch: 5/20... Training Step: 4947... Training loss: 1.5289... 0.0806 sec/batch
Epoch: 5/20... Training Step: 4948... Training loss: 1.4407... 0.0747 sec/batch
Epoch: 5/20... Training Step: 4949... Training loss: 1.3943... 0.0738 sec/batch
Epoch: 5/20... Training Step: 4950... Training loss: 1.6002... 0.0740 sec/batch
Epoch: 5/20... Training Step: 4951... Training loss: 1.5169... 0.0778 sec/batch
Epoch: 5/20... Training Step: 4952... Training loss: 1.4745... 0.0796 sec/batch
Epoch: 5/20... Training Step: 4953... Training loss: 1.4910... 0.0773 sec/batch
Epoch: 5/20... Training Step: 4954... Training loss: 1.4303... 0.0792 sec/batch
Epoch: 5/20... Training Step: 4955... Training loss: 1.4414... 0.0800 sec/batch
Epoch: 5/20... Training Step: 4956... Training loss: 1.4720... 0.0818 sec/batch
Epoch: 5/20... Training Step: 4957... Training loss: 1.4737... 0.0760 sec/batch
Epoch: 5/20... Training Step: 4958... Training loss: 1.3937... 0.0777 sec/batch
Epoch: 5/20... Training Step: 4959... Training loss: 1.4620... 0.0758 sec/batch
Epoch: 5/20... Training Step: 4960... Training loss: 1.5127... 0.0767 sec/batch
Epoch: 6/20... Training Step: 4961... Training loss: 1.5548... 0.0755 sec/batch
Epoch: 6/20... Training Step: 4962... Training loss: 1.4409... 0.0747 sec/batch
Epoch: 6/20... Training Step: 4963... Training loss: 1.5512... 0.0758 sec/batch
Epoch: 6/20... Training Step: 4964... Training loss: 1.4947... 0.0800 sec/batch
Epoch: 6/20... Training Step: 4965... Training loss: 1.3959... 0.0767 sec/batch
Epoch: 6/20... Training Step: 4966... Training loss: 1.3474... 0.0757 sec/batch
Epoch: 6/20... Training Step: 4967... Training loss: 1.4496... 0.0778 sec/batch
Epoch: 6/20... Training Step: 4968... Training loss: 1.4769... 0.0775 sec/batch
Epoch: 6/20... Training Step: 4969... Training loss: 1.4935... 0.0754 sec/batch
Epoch: 6/20... Training Step: 4970... Training loss: 1.3982... 0.0733 sec/batch
Epoch: 6/20... Training Step: 4971... Training loss: 1.5153... 0.0750 sec/batch
Epoch: 6/20... Training Step: 4972... Training loss: 1.4133... 0.0749 sec/batch
Epoch: 6/20... Training Step: 4973... Training loss: 1.4650... 0.0827 sec/batch
Epoch: 6/20... Training Step: 4974... Training loss: 1.4817... 0.0755 sec/batch
Epoch: 6/20... Training Step: 4975... Training loss: 1.4939... 0.0775 sec/batch
Epoch: 6/20... Training Step: 4976... Training loss: 1.4127... 0.0735 sec/batch
Epoch: 6/20... Training Step: 4977... Training loss: 1.4505... 0.0761 sec/batch
Epoch: 6/20... Training Step: 4978... Training loss: 1.4703... 0.0771 sec/batch
Epoch: 6/20... Training Step: 4979... Training loss: 1.4576... 0.0760 sec/batch
Epoch: 6/20... Training Step: 4980... Training loss: 1.4351... 0.0765 sec/batch
Epoch: 6/20... Training Step: 4981... Training loss: 1.4576... 0.0776 sec/batch
Epoch: 6/20... Training Step: 4982... Training loss: 1.4136... 0.0780 sec/batch
Epoch: 6/20... Training Step: 4983... Training loss: 1.4172... 0.0743 sec/batch
Epoch: 6/20... Training Step: 4984... Training loss: 1.4092... 0.0772 sec/batch
Epoch: 6/20... Training Step: 4985... Training loss: 1.4953... 0.0776 sec/batch
Epoch: 6/20... Training Step: 4986... Training loss: 1.3981... 0.0747 sec/batch
Epoch: 6/20... Training Step: 4987... Training loss: 1.4495... 0.0754 sec/batch
Epoch: 6/20... Training Step: 4988... Training loss: 1.3862... 0.0769 sec/batch
Epoch: 6/20... Training Step: 4989... Training loss: 1.4969... 0.0776 sec/batch
Epoch: 6/20... Training Step: 4990... Training loss: 1.5268... 0.0749 sec/batch
Epoch: 6/20... Training Step: 4991... Training loss: 1.5024... 0.0738 sec/batch
Epoch: 6/20... Training Step: 4992... Training loss: 1.4885... 0.0748 sec/batch
Epoch: 6/20... Training Step: 4993... Training loss: 1.4473... 0.0772 sec/batch
Epoch: 6/20... Training Step: 4994... Training loss: 1.5104... 0.0768 sec/batch
Epoch: 6/20... Training Step: 4995... Training loss: 1.5556... 0.0786 sec/batch
Epoch: 6/20... Training Step: 4996... Training loss: 1.5228... 0.0744 sec/batch
Epoch: 6/20... Training Step: 4997... Training loss: 1.4549... 0.0746 sec/batch
Epoch: 6/20... Training Step: 4998... Training loss: 1.5300... 0.0755 sec/batch
Epoch: 6/20... Training Step: 4999... Training loss: 1.4595... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5000... Training loss: 1.4607... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5001... Training loss: 1.4538... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5002... Training loss: 1.4259... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5003... Training loss: 1.4287... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5004... Training loss: 1.4345... 0.0794 sec/batch
Epoch: 6/20... Training Step: 5005... Training loss: 1.4896... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5006... Training loss: 1.4240... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5007... Training loss: 1.5220... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5008... Training loss: 1.4134... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5009... Training loss: 1.4989... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5010... Training loss: 1.4979... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5011... Training loss: 1.3986... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5012... Training loss: 1.5045... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5013... Training loss: 1.3972... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5014... Training loss: 1.4236... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5015... Training loss: 1.4521... 0.0786 sec/batch
Epoch: 6/20... Training Step: 5016... Training loss: 1.4097... 0.0784 sec/batch
Epoch: 6/20... Training Step: 5017... Training loss: 1.4679... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5018... Training loss: 1.5043... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5019... Training loss: 1.4594... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5020... Training loss: 1.4714... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5021... Training loss: 1.3768... 0.0785 sec/batch
Epoch: 6/20... Training Step: 5022... Training loss: 1.3989... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5023... Training loss: 1.4326... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5024... Training loss: 1.3929... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5025... Training loss: 1.4757... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5026... Training loss: 1.4614... 0.0780 sec/batch
Epoch: 6/20... Training Step: 5027... Training loss: 1.4308... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5028... Training loss: 1.4442... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5029... Training loss: 1.4813... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5030... Training loss: 1.4675... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5031... Training loss: 1.4881... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5032... Training loss: 1.5194... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5033... Training loss: 1.4414... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5034... Training loss: 1.4003... 0.0797 sec/batch
Epoch: 6/20... Training Step: 5035... Training loss: 1.4209... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5036... Training loss: 1.4102... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5037... Training loss: 1.4164... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5038... Training loss: 1.4257... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5039... Training loss: 1.4706... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5040... Training loss: 1.5099... 0.0788 sec/batch
Epoch: 6/20... Training Step: 5041... Training loss: 1.4657... 0.0796 sec/batch
Epoch: 6/20... Training Step: 5042... Training loss: 1.4362... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5043... Training loss: 1.4728... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5044... Training loss: 1.4871... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5045... Training loss: 1.4349... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5046... Training loss: 1.4990... 0.0812 sec/batch
Epoch: 6/20... Training Step: 5047... Training loss: 1.4511... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5048... Training loss: 1.4884... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5049... Training loss: 1.4717... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5050... Training loss: 1.5065... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5051... Training loss: 1.5316... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5052... Training loss: 1.4389... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5053... Training loss: 1.3936... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5054... Training loss: 1.3533... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5055... Training loss: 1.3775... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5056... Training loss: 1.4935... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5057... Training loss: 1.4457... 0.0823 sec/batch
Epoch: 6/20... Training Step: 5058... Training loss: 1.3776... 0.0827 sec/batch
Epoch: 6/20... Training Step: 5059... Training loss: 1.4232... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5060... Training loss: 1.4232... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5061... Training loss: 1.4159... 0.0794 sec/batch
Epoch: 6/20... Training Step: 5062... Training loss: 1.4207... 0.0780 sec/batch
Epoch: 6/20... Training Step: 5063... Training loss: 1.4507... 0.0781 sec/batch
Epoch: 6/20... Training Step: 5064... Training loss: 1.4850... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5065... Training loss: 1.4658... 0.0817 sec/batch
Epoch: 6/20... Training Step: 5066... Training loss: 1.4413... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5067... Training loss: 1.4027... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5068... Training loss: 1.5030... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5069... Training loss: 1.4206... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5070... Training loss: 1.5053... 0.0806 sec/batch
Epoch: 6/20... Training Step: 5071... Training loss: 1.4879... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5072... Training loss: 1.4373... 0.0797 sec/batch
Epoch: 6/20... Training Step: 5073... Training loss: 1.5387... 0.0784 sec/batch
Epoch: 6/20... Training Step: 5074... Training loss: 1.4777... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5075... Training loss: 1.4106... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5076... Training loss: 1.5716... 0.0826 sec/batch
Epoch: 6/20... Training Step: 5077... Training loss: 1.4884... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5078... Training loss: 1.5588... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5079... Training loss: 1.4675... 0.0792 sec/batch
Epoch: 6/20... Training Step: 5080... Training loss: 1.4721... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5081... Training loss: 1.5652... 0.0780 sec/batch
Epoch: 6/20... Training Step: 5082... Training loss: 1.4303... 0.0782 sec/batch
Epoch: 6/20... Training Step: 5083... Training loss: 1.4936... 0.0784 sec/batch
Epoch: 6/20... Training Step: 5084... Training loss: 1.4612... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5085... Training loss: 1.5400... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5086... Training loss: 1.5320... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5087... Training loss: 1.5187... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5088... Training loss: 1.4686... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5089... Training loss: 1.4074... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5090... Training loss: 1.4506... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5091... Training loss: 1.4438... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5092... Training loss: 1.5103... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5093... Training loss: 1.4880... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5094... Training loss: 1.4511... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5095... Training loss: 1.4949... 0.0803 sec/batch
Epoch: 6/20... Training Step: 5096... Training loss: 1.4900... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5097... Training loss: 1.4833... 0.0798 sec/batch
Epoch: 6/20... Training Step: 5098... Training loss: 1.5451... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5099... Training loss: 1.5314... 0.0806 sec/batch
Epoch: 6/20... Training Step: 5100... Training loss: 1.4563... 0.0897 sec/batch
Epoch: 6/20... Training Step: 5101... Training loss: 1.4858... 0.0834 sec/batch
Epoch: 6/20... Training Step: 5102... Training loss: 1.4527... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5103... Training loss: 1.4368... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5104... Training loss: 1.4559... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5105... Training loss: 1.3474... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5106... Training loss: 1.5254... 0.0785 sec/batch
Epoch: 6/20... Training Step: 5107... Training loss: 1.4614... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5108... Training loss: 1.3785... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5109... Training loss: 1.4079... 0.0792 sec/batch
Epoch: 6/20... Training Step: 5110... Training loss: 1.3988... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5111... Training loss: 1.3469... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5112... Training loss: 1.4620... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5113... Training loss: 1.3882... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5114... Training loss: 1.4395... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5115... Training loss: 1.4093... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5116... Training loss: 1.4067... 0.0781 sec/batch
Epoch: 6/20... Training Step: 5117... Training loss: 1.4460... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5118... Training loss: 1.4355... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5119... Training loss: 1.4387... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5120... Training loss: 1.4850... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5121... Training loss: 1.4220... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5122... Training loss: 1.4274... 0.0826 sec/batch
Epoch: 6/20... Training Step: 5123... Training loss: 1.4462... 0.0817 sec/batch
Epoch: 6/20... Training Step: 5124... Training loss: 1.5140... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5125... Training loss: 1.4791... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5126... Training loss: 1.5529... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5127... Training loss: 1.4941... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5128... Training loss: 1.4767... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5129... Training loss: 1.4841... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5130... Training loss: 1.4768... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5131... Training loss: 1.4823... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5132... Training loss: 1.4393... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5133... Training loss: 1.4400... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5134... Training loss: 1.5010... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5135... Training loss: 1.4307... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5136... Training loss: 1.4789... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5137... Training loss: 1.4803... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5138... Training loss: 1.5725... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5139... Training loss: 1.5068... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5140... Training loss: 1.4285... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5141... Training loss: 1.4816... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5142... Training loss: 1.5412... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5143... Training loss: 1.4396... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5144... Training loss: 1.4864... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5145... Training loss: 1.4955... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5146... Training loss: 1.5171... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5147... Training loss: 1.4702... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5148... Training loss: 1.4442... 0.0787 sec/batch
Epoch: 6/20... Training Step: 5149... Training loss: 1.4562... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5150... Training loss: 1.5133... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5151... Training loss: 1.5323... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5152... Training loss: 1.5087... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5153... Training loss: 1.5124... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5154... Training loss: 1.5754... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5155... Training loss: 1.4653... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5156... Training loss: 1.4319... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5157... Training loss: 1.5113... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5158... Training loss: 1.4537... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5159... Training loss: 1.4953... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5160... Training loss: 1.4733... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5161... Training loss: 1.4423... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5162... Training loss: 1.4615... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5163... Training loss: 1.4838... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5164... Training loss: 1.4470... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5165... Training loss: 1.4451... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5166... Training loss: 1.5418... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5167... Training loss: 1.4541... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5168... Training loss: 1.4389... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5169... Training loss: 1.4756... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5170... Training loss: 1.4787... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5171... Training loss: 1.4115... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5172... Training loss: 1.5440... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5173... Training loss: 1.4447... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5174... Training loss: 1.4766... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5175... Training loss: 1.4384... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5176... Training loss: 1.4653... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5177... Training loss: 1.4258... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5178... Training loss: 1.4926... 0.0809 sec/batch
Epoch: 6/20... Training Step: 5179... Training loss: 1.4744... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5180... Training loss: 1.4242... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5181... Training loss: 1.4034... 0.0784 sec/batch
Epoch: 6/20... Training Step: 5182... Training loss: 1.5511... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5183... Training loss: 1.4808... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5184... Training loss: 1.4522... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5185... Training loss: 1.5030... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5186... Training loss: 1.4273... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5187... Training loss: 1.4668... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5188... Training loss: 1.4681... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5189... Training loss: 1.3865... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5190... Training loss: 1.4328... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5191... Training loss: 1.4484... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5192... Training loss: 1.4539... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5193... Training loss: 1.4386... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5194... Training loss: 1.5496... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5195... Training loss: 1.4144... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5196... Training loss: 1.4460... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5197... Training loss: 1.4352... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5198... Training loss: 1.4928... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5199... Training loss: 1.5164... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5200... Training loss: 1.4440... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5201... Training loss: 1.4440... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5202... Training loss: 1.4553... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5203... Training loss: 1.4867... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5204... Training loss: 1.4934... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5205... Training loss: 1.4237... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5206... Training loss: 1.5001... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5207... Training loss: 1.4173... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5208... Training loss: 1.3886... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5209... Training loss: 1.4506... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5210... Training loss: 1.4557... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5211... Training loss: 1.4409... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5212... Training loss: 1.4411... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5213... Training loss: 1.4223... 0.0819 sec/batch
Epoch: 6/20... Training Step: 5214... Training loss: 1.4478... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5215... Training loss: 1.4246... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5216... Training loss: 1.4333... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5217... Training loss: 1.5582... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5218... Training loss: 1.4349... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5219... Training loss: 1.4439... 0.0805 sec/batch
Epoch: 6/20... Training Step: 5220... Training loss: 1.4584... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5221... Training loss: 1.5041... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5222... Training loss: 1.5137... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5223... Training loss: 1.4827... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5224... Training loss: 1.4692... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5225... Training loss: 1.4969... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5226... Training loss: 1.3770... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5227... Training loss: 1.4230... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5228... Training loss: 1.4449... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5229... Training loss: 1.4113... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5230... Training loss: 1.3936... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5231... Training loss: 1.4489... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5232... Training loss: 1.4325... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5233... Training loss: 1.3477... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5234... Training loss: 1.3795... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5235... Training loss: 1.4144... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5236... Training loss: 1.4336... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5237... Training loss: 1.4648... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5238... Training loss: 1.3680... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5239... Training loss: 1.3985... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5240... Training loss: 1.5102... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5241... Training loss: 1.4380... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5242... Training loss: 1.4310... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5243... Training loss: 1.4666... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5244... Training loss: 1.4222... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5245... Training loss: 1.3534... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5246... Training loss: 1.3900... 0.0823 sec/batch
Epoch: 6/20... Training Step: 5247... Training loss: 1.4930... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5248... Training loss: 1.4308... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5249... Training loss: 1.4275... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5250... Training loss: 1.4162... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5251... Training loss: 1.4346... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5252... Training loss: 1.4535... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5253... Training loss: 1.3599... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5254... Training loss: 1.3886... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5255... Training loss: 1.4641... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5256... Training loss: 1.4628... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5257... Training loss: 1.3922... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5258... Training loss: 1.5132... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5259... Training loss: 1.4400... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5260... Training loss: 1.4958... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5261... Training loss: 1.4989... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5262... Training loss: 1.4195... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5263... Training loss: 1.4049... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5264... Training loss: 1.4809... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5265... Training loss: 1.4452... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5266... Training loss: 1.5108... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5267... Training loss: 1.4135... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5268... Training loss: 1.3600... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5269... Training loss: 1.5375... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5270... Training loss: 1.4426... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5271... Training loss: 1.5088... 0.0782 sec/batch
Epoch: 6/20... Training Step: 5272... Training loss: 1.4439... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5273... Training loss: 1.4351... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5274... Training loss: 1.4679... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5275... Training loss: 1.4636... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5276... Training loss: 1.3947... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5277... Training loss: 1.3854... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5278... Training loss: 1.3700... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5279... Training loss: 1.4210... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5280... Training loss: 1.4612... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5281... Training loss: 1.4529... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5282... Training loss: 1.3653... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5283... Training loss: 1.4052... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5284... Training loss: 1.4189... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5285... Training loss: 1.5125... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5286... Training loss: 1.4158... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5287... Training loss: 1.4326... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5288... Training loss: 1.4157... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5289... Training loss: 1.4309... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5290... Training loss: 1.4461... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5291... Training loss: 1.5031... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5292... Training loss: 1.3941... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5293... Training loss: 1.5148... 0.0795 sec/batch
Epoch: 6/20... Training Step: 5294... Training loss: 1.5164... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5295... Training loss: 1.4896... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5296... Training loss: 1.4726... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5297... Training loss: 1.4218... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5298... Training loss: 1.4303... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5299... Training loss: 1.4524... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5300... Training loss: 1.4262... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5301... Training loss: 1.4413... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5302... Training loss: 1.4212... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5303... Training loss: 1.3972... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5304... Training loss: 1.4647... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5305... Training loss: 1.4388... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5306... Training loss: 1.4585... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5307... Training loss: 1.4755... 0.0786 sec/batch
Epoch: 6/20... Training Step: 5308... Training loss: 1.4824... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5309... Training loss: 1.5120... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5310... Training loss: 1.5248... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5311... Training loss: 1.4674... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5312... Training loss: 1.4208... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5313... Training loss: 1.4554... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5314... Training loss: 1.4493... 0.0731 sec/batch
Epoch: 6/20... Training Step: 5315... Training loss: 1.4565... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5316... Training loss: 1.4529... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5317... Training loss: 1.4166... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5318... Training loss: 1.4627... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5319... Training loss: 1.4494... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5320... Training loss: 1.4353... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5321... Training loss: 1.4720... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5322... Training loss: 1.5184... 0.0727 sec/batch
Epoch: 6/20... Training Step: 5323... Training loss: 1.4304... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5324... Training loss: 1.4586... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5325... Training loss: 1.4461... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5326... Training loss: 1.4169... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5327... Training loss: 1.4042... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5328... Training loss: 1.3758... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5329... Training loss: 1.3803... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5330... Training loss: 1.3822... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5331... Training loss: 1.4968... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5332... Training loss: 1.3762... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5333... Training loss: 1.4453... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5334... Training loss: 1.3646... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5335... Training loss: 1.4491... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5336... Training loss: 1.4123... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5337... Training loss: 1.4230... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5338... Training loss: 1.4404... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5339... Training loss: 1.3976... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5340... Training loss: 1.4524... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5341... Training loss: 1.4357... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5342... Training loss: 1.4509... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5343... Training loss: 1.4729... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5344... Training loss: 1.4713... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5345... Training loss: 1.4846... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5346... Training loss: 1.4297... 0.0801 sec/batch
Epoch: 6/20... Training Step: 5347... Training loss: 1.4857... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5348... Training loss: 1.4993... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5349... Training loss: 1.4796... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5350... Training loss: 1.3908... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5351... Training loss: 1.4557... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5352... Training loss: 1.4117... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5353... Training loss: 1.4842... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5354... Training loss: 1.4062... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5355... Training loss: 1.4489... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5356... Training loss: 1.4115... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5357... Training loss: 1.4163... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5358... Training loss: 1.5099... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5359... Training loss: 1.4799... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5360... Training loss: 1.4242... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5361... Training loss: 1.3911... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5362... Training loss: 1.4727... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5363... Training loss: 1.3837... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5364... Training loss: 1.4973... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5365... Training loss: 1.4219... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5366... Training loss: 1.4181... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5367... Training loss: 1.3804... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5368... Training loss: 1.5371... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5369... Training loss: 1.3976... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5370... Training loss: 1.3882... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5371... Training loss: 1.3603... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5372... Training loss: 1.3919... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5373... Training loss: 1.4324... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5374... Training loss: 1.3799... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5375... Training loss: 1.4043... 0.0794 sec/batch
Epoch: 6/20... Training Step: 5376... Training loss: 1.4744... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5377... Training loss: 1.3741... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5378... Training loss: 1.4052... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5379... Training loss: 1.4825... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5380... Training loss: 1.5205... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5381... Training loss: 1.3999... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5382... Training loss: 1.4359... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5383... Training loss: 1.4290... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5384... Training loss: 1.4432... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5385... Training loss: 1.4431... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5386... Training loss: 1.4428... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5387... Training loss: 1.5139... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5388... Training loss: 1.5320... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5389... Training loss: 1.4926... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5390... Training loss: 1.5287... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5391... Training loss: 1.4421... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5392... Training loss: 1.4582... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5393... Training loss: 1.4876... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5394... Training loss: 1.4836... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5395... Training loss: 1.5195... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5396... Training loss: 1.4175... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5397... Training loss: 1.4897... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5398... Training loss: 1.5531... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5399... Training loss: 1.4631... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5400... Training loss: 1.5020... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5401... Training loss: 1.4558... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5402... Training loss: 1.4900... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5403... Training loss: 1.4559... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5404... Training loss: 1.4588... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5405... Training loss: 1.4279... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5406... Training loss: 1.5381... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5407... Training loss: 1.4466... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5408... Training loss: 1.4475... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5409... Training loss: 1.5115... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5410... Training loss: 1.5306... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5411... Training loss: 1.4599... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5412... Training loss: 1.4553... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5413... Training loss: 1.4840... 0.0781 sec/batch
Epoch: 6/20... Training Step: 5414... Training loss: 1.4765... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5415... Training loss: 1.4785... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5416... Training loss: 1.4437... 0.0790 sec/batch
Epoch: 6/20... Training Step: 5417... Training loss: 1.4001... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5418... Training loss: 1.4523... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5419... Training loss: 1.4418... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5420... Training loss: 1.4630... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5421... Training loss: 1.4819... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5422... Training loss: 1.4428... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5423... Training loss: 1.3981... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5424... Training loss: 1.4660... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5425... Training loss: 1.3976... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5426... Training loss: 1.4116... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5427... Training loss: 1.3643... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5428... Training loss: 1.4303... 0.0798 sec/batch
Epoch: 6/20... Training Step: 5429... Training loss: 1.4348... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5430... Training loss: 1.5152... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5431... Training loss: 1.4557... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5432... Training loss: 1.4448... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5433... Training loss: 1.4239... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5434... Training loss: 1.4327... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5435... Training loss: 1.3306... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5436... Training loss: 1.4275... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5437... Training loss: 1.4362... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5438... Training loss: 1.4380... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5439... Training loss: 1.4050... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5440... Training loss: 1.3919... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5441... Training loss: 1.3829... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5442... Training loss: 1.4735... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5443... Training loss: 1.4532... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5444... Training loss: 1.4557... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5445... Training loss: 1.3986... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5446... Training loss: 1.4028... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5447... Training loss: 1.4735... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5448... Training loss: 1.4246... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5449... Training loss: 1.4000... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5450... Training loss: 1.3631... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5451... Training loss: 1.4054... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5452... Training loss: 1.3983... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5453... Training loss: 1.4380... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5454... Training loss: 1.3888... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5455... Training loss: 1.3994... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5456... Training loss: 1.3794... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5457... Training loss: 1.4504... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5458... Training loss: 1.3953... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5459... Training loss: 1.4379... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5460... Training loss: 1.4618... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5461... Training loss: 1.4100... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5462... Training loss: 1.4375... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5463... Training loss: 1.4325... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5464... Training loss: 1.4306... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5465... Training loss: 1.4532... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5466... Training loss: 1.3627... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5467... Training loss: 1.4073... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5468... Training loss: 1.4089... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5469... Training loss: 1.3811... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5470... Training loss: 1.3857... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5471... Training loss: 1.4638... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5472... Training loss: 1.4415... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5473... Training loss: 1.5043... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5474... Training loss: 1.4872... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5475... Training loss: 1.4233... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5476... Training loss: 1.4435... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5477... Training loss: 1.4497... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5478... Training loss: 1.4167... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5479... Training loss: 1.4514... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5480... Training loss: 1.4017... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5481... Training loss: 1.4409... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5482... Training loss: 1.3933... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5483... Training loss: 1.5185... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5484... Training loss: 1.4013... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5485... Training loss: 1.4434... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5486... Training loss: 1.4282... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5487... Training loss: 1.4380... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5488... Training loss: 1.3997... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5489... Training loss: 1.4206... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5490... Training loss: 1.4946... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5491... Training loss: 1.5235... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5492... Training loss: 1.4708... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5493... Training loss: 1.4908... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5494... Training loss: 1.4889... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5495... Training loss: 1.4470... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5496... Training loss: 1.4201... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5497... Training loss: 1.4009... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5498... Training loss: 1.4133... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5499... Training loss: 1.4118... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5500... Training loss: 1.3996... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5501... Training loss: 1.5324... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5502... Training loss: 1.4071... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5503... Training loss: 1.4056... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5504... Training loss: 1.4089... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5505... Training loss: 1.3771... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5506... Training loss: 1.4035... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5507... Training loss: 1.5049... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5508... Training loss: 1.4593... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5509... Training loss: 1.3862... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5510... Training loss: 1.3978... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5511... Training loss: 1.4303... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5512... Training loss: 1.3765... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5513... Training loss: 1.3542... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5514... Training loss: 1.4290... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5515... Training loss: 1.4403... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5516... Training loss: 1.3750... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5517... Training loss: 1.3869... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5518... Training loss: 1.4143... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5519... Training loss: 1.4338... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5520... Training loss: 1.3462... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5521... Training loss: 1.4129... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5522... Training loss: 1.4676... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5523... Training loss: 1.3835... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5524... Training loss: 1.4095... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5525... Training loss: 1.4378... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5526... Training loss: 1.4489... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5527... Training loss: 1.3953... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5528... Training loss: 1.3624... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5529... Training loss: 1.3475... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5530... Training loss: 1.3591... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5531... Training loss: 1.3491... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5532... Training loss: 1.3993... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5533... Training loss: 1.4745... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5534... Training loss: 1.4707... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5535... Training loss: 1.4526... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5536... Training loss: 1.4291... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5537... Training loss: 1.4339... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5538... Training loss: 1.5721... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5539... Training loss: 1.4725... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5540... Training loss: 1.3861... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5541... Training loss: 1.3741... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5542... Training loss: 1.3912... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5543... Training loss: 1.4996... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5544... Training loss: 1.3751... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5545... Training loss: 1.4115... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5546... Training loss: 1.4372... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5547... Training loss: 1.4418... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5548... Training loss: 1.4780... 0.0807 sec/batch
Epoch: 6/20... Training Step: 5549... Training loss: 1.4229... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5550... Training loss: 1.4254... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5551... Training loss: 1.4246... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5552... Training loss: 1.4658... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5553... Training loss: 1.3967... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5554... Training loss: 1.3832... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5555... Training loss: 1.3304... 0.0728 sec/batch
Epoch: 6/20... Training Step: 5556... Training loss: 1.4586... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5557... Training loss: 1.2976... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5558... Training loss: 1.3344... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5559... Training loss: 1.3885... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5560... Training loss: 1.3939... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5561... Training loss: 1.4922... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5562... Training loss: 1.3941... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5563... Training loss: 1.3861... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5564... Training loss: 1.4315... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5565... Training loss: 1.3749... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5566... Training loss: 1.4061... 0.0822 sec/batch
Epoch: 6/20... Training Step: 5567... Training loss: 1.4738... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5568... Training loss: 1.4557... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5569... Training loss: 1.4657... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5570... Training loss: 1.4829... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5571... Training loss: 1.3370... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5572... Training loss: 1.4525... 0.0829 sec/batch
Epoch: 6/20... Training Step: 5573... Training loss: 1.4724... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5574... Training loss: 1.3882... 0.0780 sec/batch
Epoch: 6/20... Training Step: 5575... Training loss: 1.4720... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5576... Training loss: 1.4081... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5577... Training loss: 1.4425... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5578... Training loss: 1.4844... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5579... Training loss: 1.3780... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5580... Training loss: 1.4114... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5581... Training loss: 1.3764... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5582... Training loss: 1.3341... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5583... Training loss: 1.4522... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5584... Training loss: 1.4381... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5585... Training loss: 1.3698... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5586... Training loss: 1.3905... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5587... Training loss: 1.4200... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5588... Training loss: 1.4264... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5589... Training loss: 1.2633... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5590... Training loss: 1.4137... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5591... Training loss: 1.4075... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5592... Training loss: 1.4001... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5593... Training loss: 1.3633... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5594... Training loss: 1.4032... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5595... Training loss: 1.4377... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5596... Training loss: 1.4235... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5597... Training loss: 1.4631... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5598... Training loss: 1.4028... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5599... Training loss: 1.4495... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5600... Training loss: 1.4936... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5601... Training loss: 1.5197... 0.0730 sec/batch
Epoch: 6/20... Training Step: 5602... Training loss: 1.4717... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5603... Training loss: 1.4347... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5604... Training loss: 1.4195... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5605... Training loss: 1.3874... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5606... Training loss: 1.4619... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5607... Training loss: 1.4653... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5608... Training loss: 1.3552... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5609... Training loss: 1.3778... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5610... Training loss: 1.4257... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5611... Training loss: 1.4322... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5612... Training loss: 1.4694... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5613... Training loss: 1.4309... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5614... Training loss: 1.4801... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5615... Training loss: 1.4834... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5616... Training loss: 1.4777... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5617... Training loss: 1.3605... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5618... Training loss: 1.4057... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5619... Training loss: 1.4202... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5620... Training loss: 1.5049... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5621... Training loss: 1.4814... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5622... Training loss: 1.5227... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5623... Training loss: 1.4290... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5624... Training loss: 1.5606... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5625... Training loss: 1.5406... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5626... Training loss: 1.3839... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5627... Training loss: 1.5410... 0.0784 sec/batch
Epoch: 6/20... Training Step: 5628... Training loss: 1.5041... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5629... Training loss: 1.4281... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5630... Training loss: 1.4040... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5631... Training loss: 1.4665... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5632... Training loss: 1.4962... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5633... Training loss: 1.4184... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5634... Training loss: 1.4456... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5635... Training loss: 1.3663... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5636... Training loss: 1.3938... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5637... Training loss: 1.5434... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5638... Training loss: 1.4107... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5639... Training loss: 1.3976... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5640... Training loss: 1.4585... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5641... Training loss: 1.4878... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5642... Training loss: 1.4301... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5643... Training loss: 1.4652... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5644... Training loss: 1.4013... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5645... Training loss: 1.5067... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5646... Training loss: 1.4982... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5647... Training loss: 1.4242... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5648... Training loss: 1.4350... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5649... Training loss: 1.3955... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5650... Training loss: 1.4771... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5651... Training loss: 1.4245... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5652... Training loss: 1.4549... 0.0813 sec/batch
Epoch: 6/20... Training Step: 5653... Training loss: 1.4073... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5654... Training loss: 1.3772... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5655... Training loss: 1.4132... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5656... Training loss: 1.4424... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5657... Training loss: 1.5137... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5658... Training loss: 1.4756... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5659... Training loss: 1.4202... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5660... Training loss: 1.4361... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5661... Training loss: 1.4626... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5662... Training loss: 1.4243... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5663... Training loss: 1.4320... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5664... Training loss: 1.3515... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5665... Training loss: 1.3549... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5666... Training loss: 1.3716... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5667... Training loss: 1.4561... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5668... Training loss: 1.4749... 0.0782 sec/batch
Epoch: 6/20... Training Step: 5669... Training loss: 1.4691... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5670... Training loss: 1.4107... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5671... Training loss: 1.4462... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5672... Training loss: 1.5159... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5673... Training loss: 1.3963... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5674... Training loss: 1.4365... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5675... Training loss: 1.4664... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5676... Training loss: 1.4688... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5677... Training loss: 1.4893... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5678... Training loss: 1.4530... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5679... Training loss: 1.4711... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5680... Training loss: 1.5459... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5681... Training loss: 1.4595... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5682... Training loss: 1.4261... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5683... Training loss: 1.4275... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5684... Training loss: 1.4732... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5685... Training loss: 1.4955... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5686... Training loss: 1.5065... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5687... Training loss: 1.4120... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5688... Training loss: 1.4724... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5689... Training loss: 1.4642... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5690... Training loss: 1.3901... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5691... Training loss: 1.4048... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5692... Training loss: 1.4450... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5693... Training loss: 1.3878... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5694... Training loss: 1.4914... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5695... Training loss: 1.4172... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5696... Training loss: 1.4145... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5697... Training loss: 1.3884... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5698... Training loss: 1.5136... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5699... Training loss: 1.4330... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5700... Training loss: 1.4345... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5701... Training loss: 1.3888... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5702... Training loss: 1.4046... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5703... Training loss: 1.3960... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5704... Training loss: 1.5054... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5705... Training loss: 1.4582... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5706... Training loss: 1.4240... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5707... Training loss: 1.4234... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5708... Training loss: 1.4382... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5709... Training loss: 1.3723... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5710... Training loss: 1.3872... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5711... Training loss: 1.5421... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5712... Training loss: 1.3873... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5713... Training loss: 1.4638... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5714... Training loss: 1.4441... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5715... Training loss: 1.3661... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5716... Training loss: 1.4389... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5717... Training loss: 1.4497... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5718... Training loss: 1.5064... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5719... Training loss: 1.4765... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5720... Training loss: 1.5040... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5721... Training loss: 1.4549... 0.0778 sec/batch
Epoch: 6/20... Training Step: 5722... Training loss: 1.3727... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5723... Training loss: 1.4467... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5724... Training loss: 1.4124... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5725... Training loss: 1.4674... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5726... Training loss: 1.4759... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5727... Training loss: 1.4490... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5728... Training loss: 1.4040... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5729... Training loss: 1.3975... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5730... Training loss: 1.4556... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5731... Training loss: 1.4555... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5732... Training loss: 1.5047... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5733... Training loss: 1.4678... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5734... Training loss: 1.4680... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5735... Training loss: 1.4309... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5736... Training loss: 1.3452... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5737... Training loss: 1.4038... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5738... Training loss: 1.4225... 0.0735 sec/batch
Epoch: 6/20... Training Step: 5739... Training loss: 1.4199... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5740... Training loss: 1.4042... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5741... Training loss: 1.4409... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5742... Training loss: 1.4286... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5743... Training loss: 1.4710... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5744... Training loss: 1.4562... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5745... Training loss: 1.4568... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5746... Training loss: 1.4236... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5747... Training loss: 1.4205... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5748... Training loss: 1.4477... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5749... Training loss: 1.4360... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5750... Training loss: 1.4688... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5751... Training loss: 1.4347... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5752... Training loss: 1.4015... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5753... Training loss: 1.4383... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5754... Training loss: 1.3894... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5755... Training loss: 1.3300... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5756... Training loss: 1.4294... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5757... Training loss: 1.4031... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5758... Training loss: 1.4354... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5759... Training loss: 1.4069... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5760... Training loss: 1.4175... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5761... Training loss: 1.4802... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5762... Training loss: 1.4772... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5763... Training loss: 1.4944... 0.0743 sec/batch
Epoch: 6/20... Training Step: 5764... Training loss: 1.4479... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5765... Training loss: 1.3826... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5766... Training loss: 1.4273... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5767... Training loss: 1.4173... 0.0732 sec/batch
Epoch: 6/20... Training Step: 5768... Training loss: 1.4678... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5769... Training loss: 1.4805... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5770... Training loss: 1.5018... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5771... Training loss: 1.4629... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5772... Training loss: 1.4878... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5773... Training loss: 1.4430... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5774... Training loss: 1.5024... 0.0795 sec/batch
Epoch: 6/20... Training Step: 5775... Training loss: 1.4807... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5776... Training loss: 1.4827... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5777... Training loss: 1.5191... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5778... Training loss: 1.5507... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5779... Training loss: 1.4949... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5780... Training loss: 1.4269... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5781... Training loss: 1.4458... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5782... Training loss: 1.3755... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5783... Training loss: 1.4440... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5784... Training loss: 1.4243... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5785... Training loss: 1.4184... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5786... Training loss: 1.4406... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5787... Training loss: 1.4393... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5788... Training loss: 1.5101... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5789... Training loss: 1.4620... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5790... Training loss: 1.4100... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5791... Training loss: 1.5014... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5792... Training loss: 1.4686... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5793... Training loss: 1.5359... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5794... Training loss: 1.4332... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5795... Training loss: 1.3671... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5796... Training loss: 1.4300... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5797... Training loss: 1.4789... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5798... Training loss: 1.4820... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5799... Training loss: 1.4982... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5800... Training loss: 1.4226... 0.0736 sec/batch
Epoch: 6/20... Training Step: 5801... Training loss: 1.4636... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5802... Training loss: 1.5958... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5803... Training loss: 1.4897... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5804... Training loss: 1.5310... 0.0823 sec/batch
Epoch: 6/20... Training Step: 5805... Training loss: 1.5175... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5806... Training loss: 1.3786... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5807... Training loss: 1.4615... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5808... Training loss: 1.4599... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5809... Training loss: 1.4180... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5810... Training loss: 1.4612... 0.0761 sec/batch
Epoch: 6/20... Training Step: 5811... Training loss: 1.4104... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5812... Training loss: 1.5409... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5813... Training loss: 1.5248... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5814... Training loss: 1.4551... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5815... Training loss: 1.4386... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5816... Training loss: 1.4557... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5817... Training loss: 1.4399... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5818... Training loss: 1.4293... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5819... Training loss: 1.4735... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5820... Training loss: 1.4859... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5821... Training loss: 1.4663... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5822... Training loss: 1.4002... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5823... Training loss: 1.4012... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5824... Training loss: 1.4093... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5825... Training loss: 1.3854... 0.0770 sec/batch
Epoch: 6/20... Training Step: 5826... Training loss: 1.4361... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5827... Training loss: 1.4582... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5828... Training loss: 1.5120... 0.0759 sec/batch
Epoch: 6/20... Training Step: 5829... Training loss: 1.4520... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5830... Training loss: 1.5200... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5831... Training loss: 1.4686... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5832... Training loss: 1.4685... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5833... Training loss: 1.4725... 0.0746 sec/batch
Epoch: 6/20... Training Step: 5834... Training loss: 1.4756... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5835... Training loss: 1.4510... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5836... Training loss: 1.4788... 0.0772 sec/batch
Epoch: 6/20... Training Step: 5837... Training loss: 1.4651... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5838... Training loss: 1.5044... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5839... Training loss: 1.4556... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5840... Training loss: 1.4175... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5841... Training loss: 1.4357... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5842... Training loss: 1.3556... 0.0760 sec/batch
Epoch: 6/20... Training Step: 5843... Training loss: 1.4821... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5844... Training loss: 1.4284... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5845... Training loss: 1.4612... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5846... Training loss: 1.5162... 0.0794 sec/batch
Epoch: 6/20... Training Step: 5847... Training loss: 1.4296... 0.0791 sec/batch
Epoch: 6/20... Training Step: 5848... Training loss: 1.4798... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5849... Training loss: 1.4784... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5850... Training loss: 1.5155... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5851... Training loss: 1.4899... 0.0802 sec/batch
Epoch: 6/20... Training Step: 5852... Training loss: 1.4122... 0.0840 sec/batch
Epoch: 6/20... Training Step: 5853... Training loss: 1.4334... 0.0798 sec/batch
Epoch: 6/20... Training Step: 5854... Training loss: 1.3667... 0.0811 sec/batch
Epoch: 6/20... Training Step: 5855... Training loss: 1.4918... 0.0818 sec/batch
Epoch: 6/20... Training Step: 5856... Training loss: 1.4059... 0.0794 sec/batch
Epoch: 6/20... Training Step: 5857... Training loss: 1.4708... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5858... Training loss: 1.4618... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5859... Training loss: 1.4439... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5860... Training loss: 1.3843... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5861... Training loss: 1.4265... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5862... Training loss: 1.4036... 0.0773 sec/batch
Epoch: 6/20... Training Step: 5863... Training loss: 1.3927... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5864... Training loss: 1.3751... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5865... Training loss: 1.3642... 0.0777 sec/batch
Epoch: 6/20... Training Step: 5866... Training loss: 1.4423... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5867... Training loss: 1.3693... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5868... Training loss: 1.4512... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5869... Training loss: 1.4095... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5870... Training loss: 1.3193... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5871... Training loss: 1.3768... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5872... Training loss: 1.4256... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5873... Training loss: 1.5329... 0.0769 sec/batch
Epoch: 6/20... Training Step: 5874... Training loss: 1.4347... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5875... Training loss: 1.4392... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5876... Training loss: 1.4303... 0.0775 sec/batch
Epoch: 6/20... Training Step: 5877... Training loss: 1.4155... 0.0787 sec/batch
Epoch: 6/20... Training Step: 5878... Training loss: 1.4926... 0.0779 sec/batch
Epoch: 6/20... Training Step: 5879... Training loss: 1.4240... 0.0812 sec/batch
Epoch: 6/20... Training Step: 5880... Training loss: 1.4704... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5881... Training loss: 1.5601... 0.0797 sec/batch
Epoch: 6/20... Training Step: 5882... Training loss: 1.4934... 0.0805 sec/batch
Epoch: 6/20... Training Step: 5883... Training loss: 1.4680... 0.0798 sec/batch
Epoch: 6/20... Training Step: 5884... Training loss: 1.4115... 0.0783 sec/batch
Epoch: 6/20... Training Step: 5885... Training loss: 1.4578... 0.0789 sec/batch
Epoch: 6/20... Training Step: 5886... Training loss: 1.4921... 0.0741 sec/batch
Epoch: 6/20... Training Step: 5887... Training loss: 1.4440... 0.0757 sec/batch
Epoch: 6/20... Training Step: 5888... Training loss: 1.4420... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5889... Training loss: 1.5287... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5890... Training loss: 1.4806... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5891... Training loss: 1.4602... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5892... Training loss: 1.5355... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5893... Training loss: 1.4315... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5894... Training loss: 1.4616... 0.0776 sec/batch
Epoch: 6/20... Training Step: 5895... Training loss: 1.4613... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5896... Training loss: 1.4911... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5897... Training loss: 1.4281... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5898... Training loss: 1.4478... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5899... Training loss: 1.4280... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5900... Training loss: 1.5149... 0.0750 sec/batch
Epoch: 6/20... Training Step: 5901... Training loss: 1.4636... 0.0734 sec/batch
Epoch: 6/20... Training Step: 5902... Training loss: 1.4692... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5903... Training loss: 1.4987... 0.0852 sec/batch
Epoch: 6/20... Training Step: 5904... Training loss: 1.4908... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5905... Training loss: 1.4956... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5906... Training loss: 1.4323... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5907... Training loss: 1.4240... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5908... Training loss: 1.4150... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5909... Training loss: 1.4586... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5910... Training loss: 1.3470... 0.0740 sec/batch
Epoch: 6/20... Training Step: 5911... Training loss: 1.4420... 0.0751 sec/batch
Epoch: 6/20... Training Step: 5912... Training loss: 1.3773... 0.0788 sec/batch
Epoch: 6/20... Training Step: 5913... Training loss: 1.4318... 0.0747 sec/batch
Epoch: 6/20... Training Step: 5914... Training loss: 1.4778... 0.0756 sec/batch
Epoch: 6/20... Training Step: 5915... Training loss: 1.4931... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5916... Training loss: 1.4583... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5917... Training loss: 1.4429... 0.0764 sec/batch
Epoch: 6/20... Training Step: 5918... Training loss: 1.4772... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5919... Training loss: 1.4010... 0.0780 sec/batch
Epoch: 6/20... Training Step: 5920... Training loss: 1.3984... 0.0781 sec/batch
Epoch: 6/20... Training Step: 5921... Training loss: 1.4642... 0.0754 sec/batch
Epoch: 6/20... Training Step: 5922... Training loss: 1.4247... 0.0767 sec/batch
Epoch: 6/20... Training Step: 5923... Training loss: 1.4772... 0.0786 sec/batch
Epoch: 6/20... Training Step: 5924... Training loss: 1.4202... 0.0758 sec/batch
Epoch: 6/20... Training Step: 5925... Training loss: 1.5140... 0.0748 sec/batch
Epoch: 6/20... Training Step: 5926... Training loss: 1.5220... 0.0744 sec/batch
Epoch: 6/20... Training Step: 5927... Training loss: 1.4843... 0.0771 sec/batch
Epoch: 6/20... Training Step: 5928... Training loss: 1.5066... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5929... Training loss: 1.4806... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5930... Training loss: 1.6540... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5931... Training loss: 1.4265... 0.0737 sec/batch
Epoch: 6/20... Training Step: 5932... Training loss: 1.5175... 0.0798 sec/batch
Epoch: 6/20... Training Step: 5933... Training loss: 1.5258... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5934... Training loss: 1.4379... 0.0774 sec/batch
Epoch: 6/20... Training Step: 5935... Training loss: 1.4772... 0.0766 sec/batch
Epoch: 6/20... Training Step: 5936... Training loss: 1.5108... 0.0739 sec/batch
Epoch: 6/20... Training Step: 5937... Training loss: 1.5265... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5938... Training loss: 1.4914... 0.0733 sec/batch
Epoch: 6/20... Training Step: 5939... Training loss: 1.4700... 0.0753 sec/batch
Epoch: 6/20... Training Step: 5940... Training loss: 1.4197... 0.0755 sec/batch
Epoch: 6/20... Training Step: 5941... Training loss: 1.3766... 0.0752 sec/batch
Epoch: 6/20... Training Step: 5942... Training loss: 1.5233... 0.0745 sec/batch
Epoch: 6/20... Training Step: 5943... Training loss: 1.4888... 0.0768 sec/batch
Epoch: 6/20... Training Step: 5944... Training loss: 1.4633... 0.0738 sec/batch
Epoch: 6/20... Training Step: 5945... Training loss: 1.4745... 0.0765 sec/batch
Epoch: 6/20... Training Step: 5946... Training loss: 1.4028... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5947... Training loss: 1.4112... 0.0762 sec/batch
Epoch: 6/20... Training Step: 5948... Training loss: 1.4380... 0.0763 sec/batch
Epoch: 6/20... Training Step: 5949... Training loss: 1.4275... 0.0823 sec/batch
Epoch: 6/20... Training Step: 5950... Training loss: 1.3536... 0.0749 sec/batch
Epoch: 6/20... Training Step: 5951... Training loss: 1.4235... 0.0742 sec/batch
Epoch: 6/20... Training Step: 5952... Training loss: 1.4798... 0.0752 sec/batch
Epoch: 7/20... Training Step: 5953... Training loss: 1.5162... 0.0733 sec/batch
Epoch: 7/20... Training Step: 5954... Training loss: 1.3959... 0.0756 sec/batch
Epoch: 7/20... Training Step: 5955... Training loss: 1.5060... 0.0742 sec/batch
Epoch: 7/20... Training Step: 5956... Training loss: 1.4515... 0.0752 sec/batch
Epoch: 7/20... Training Step: 5957... Training loss: 1.3592... 0.0768 sec/batch
Epoch: 7/20... Training Step: 5958... Training loss: 1.3200... 0.0768 sec/batch
Epoch: 7/20... Training Step: 5959... Training loss: 1.4184... 0.0734 sec/batch
Epoch: 7/20... Training Step: 5960... Training loss: 1.4489... 0.0753 sec/batch
Epoch: 7/20... Training Step: 5961... Training loss: 1.4682... 0.0770 sec/batch
Epoch: 7/20... Training Step: 5962... Training loss: 1.3404... 0.0752 sec/batch
Epoch: 7/20... Training Step: 5963... Training loss: 1.5195... 0.0765 sec/batch
Epoch: 7/20... Training Step: 5964... Training loss: 1.3981... 0.0756 sec/batch
Epoch: 7/20... Training Step: 5965... Training loss: 1.4239... 0.0768 sec/batch
Epoch: 7/20... Training Step: 5966... Training loss: 1.4361... 0.0810 sec/batch
Epoch: 7/20... Training Step: 5967... Training loss: 1.4550... 0.0737 sec/batch
Epoch: 7/20... Training Step: 5968... Training loss: 1.3988... 0.0766 sec/batch
Epoch: 7/20... Training Step: 5969... Training loss: 1.4036... 0.0776 sec/batch
Epoch: 7/20... Training Step: 5970... Training loss: 1.4141... 0.0750 sec/batch
Epoch: 7/20... Training Step: 5971... Training loss: 1.4145... 0.0769 sec/batch
Epoch: 7/20... Training Step: 5972... Training loss: 1.3847... 0.0745 sec/batch
Epoch: 7/20... Training Step: 5973... Training loss: 1.4299... 0.0753 sec/batch
Epoch: 7/20... Training Step: 5974... Training loss: 1.3934... 0.0736 sec/batch
Epoch: 7/20... Training Step: 5975... Training loss: 1.3941... 0.0794 sec/batch
Epoch: 7/20... Training Step: 5976... Training loss: 1.3668... 0.0746 sec/batch
Epoch: 7/20... Training Step: 5977... Training loss: 1.4407... 0.0761 sec/batch
Epoch: 7/20... Training Step: 5978... Training loss: 1.3698... 0.0769 sec/batch
Epoch: 7/20... Training Step: 5979... Training loss: 1.4089... 0.0756 sec/batch
Epoch: 7/20... Training Step: 5980... Training loss: 1.3741... 0.0767 sec/batch
Epoch: 7/20... Training Step: 5981... Training loss: 1.4427... 0.0743 sec/batch
Epoch: 7/20... Training Step: 5982... Training loss: 1.4794... 0.0742 sec/batch
Epoch: 7/20... Training Step: 5983... Training loss: 1.4516... 0.0740 sec/batch
Epoch: 7/20... Training Step: 5984... Training loss: 1.4514... 0.0749 sec/batch
Epoch: 7/20... Training Step: 5985... Training loss: 1.4426... 0.0751 sec/batch
Epoch: 7/20... Training Step: 5986... Training loss: 1.4652... 0.0734 sec/batch
Epoch: 7/20... Training Step: 5987... Training loss: 1.5079... 0.0774 sec/batch
Epoch: 7/20... Training Step: 5988... Training loss: 1.4845... 0.0746 sec/batch
Epoch: 7/20... Training Step: 5989... Training loss: 1.4073... 0.0768 sec/batch
Epoch: 7/20... Training Step: 5990... Training loss: 1.5115... 0.0761 sec/batch
Epoch: 7/20... Training Step: 5991... Training loss: 1.4310... 0.0741 sec/batch
Epoch: 7/20... Training Step: 5992... Training loss: 1.4099... 0.0767 sec/batch
Epoch: 7/20... Training Step: 5993... Training loss: 1.4341... 0.0756 sec/batch
Epoch: 7/20... Training Step: 5994... Training loss: 1.4089... 0.0741 sec/batch
Epoch: 7/20... Training Step: 5995... Training loss: 1.3959... 0.0735 sec/batch
Epoch: 7/20... Training Step: 5996... Training loss: 1.4122... 0.0761 sec/batch
Epoch: 7/20... Training Step: 5997... Training loss: 1.4558... 0.0738 sec/batch
Epoch: 7/20... Training Step: 5998... Training loss: 1.3997... 0.0775 sec/batch
Epoch: 7/20... Training Step: 5999... Training loss: 1.4647... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6000... Training loss: 1.3868... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6001... Training loss: 1.4474... 0.0736 sec/batch
Epoch: 7/20... Training Step: 6002... Training loss: 1.4286... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6003... Training loss: 1.3435... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6004... Training loss: 1.4820... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6005... Training loss: 1.3589... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6006... Training loss: 1.3836... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6007... Training loss: 1.4112... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6008... Training loss: 1.3666... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6009... Training loss: 1.4319... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6010... Training loss: 1.4789... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6011... Training loss: 1.3987... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6012... Training loss: 1.4484... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6013... Training loss: 1.3438... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6014... Training loss: 1.4056... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6015... Training loss: 1.4082... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6016... Training loss: 1.3748... 0.0839 sec/batch
Epoch: 7/20... Training Step: 6017... Training loss: 1.4475... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6018... Training loss: 1.3936... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6019... Training loss: 1.3815... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6020... Training loss: 1.4155... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6021... Training loss: 1.4439... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6022... Training loss: 1.4274... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6023... Training loss: 1.4552... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6024... Training loss: 1.4744... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6025... Training loss: 1.4332... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6026... Training loss: 1.3569... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6027... Training loss: 1.3996... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6028... Training loss: 1.3672... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6029... Training loss: 1.4069... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6030... Training loss: 1.4082... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6031... Training loss: 1.4314... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6032... Training loss: 1.4722... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6033... Training loss: 1.4390... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6034... Training loss: 1.3922... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6035... Training loss: 1.4507... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6036... Training loss: 1.4694... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6037... Training loss: 1.3855... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6038... Training loss: 1.4794... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6039... Training loss: 1.4099... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6040... Training loss: 1.4309... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6041... Training loss: 1.4679... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6042... Training loss: 1.4804... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6043... Training loss: 1.5084... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6044... Training loss: 1.3864... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6045... Training loss: 1.3630... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6046... Training loss: 1.3218... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6047... Training loss: 1.3610... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6048... Training loss: 1.4429... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6049... Training loss: 1.4064... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6050... Training loss: 1.3748... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6051... Training loss: 1.3961... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6052... Training loss: 1.4003... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6053... Training loss: 1.4119... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6054... Training loss: 1.3771... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6055... Training loss: 1.4354... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6056... Training loss: 1.4722... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6057... Training loss: 1.4416... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6058... Training loss: 1.4278... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6059... Training loss: 1.3503... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6060... Training loss: 1.4713... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6061... Training loss: 1.4001... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6062... Training loss: 1.4637... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6063... Training loss: 1.4358... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6064... Training loss: 1.4057... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6065... Training loss: 1.4898... 0.0729 sec/batch
Epoch: 7/20... Training Step: 6066... Training loss: 1.4535... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6067... Training loss: 1.3701... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6068... Training loss: 1.5457... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6069... Training loss: 1.4474... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6070... Training loss: 1.5302... 0.0816 sec/batch
Epoch: 7/20... Training Step: 6071... Training loss: 1.4350... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6072... Training loss: 1.4392... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6073... Training loss: 1.5332... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6074... Training loss: 1.4103... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6075... Training loss: 1.4536... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6076... Training loss: 1.4187... 0.0818 sec/batch
Epoch: 7/20... Training Step: 6077... Training loss: 1.5006... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6078... Training loss: 1.4838... 0.0735 sec/batch
Epoch: 7/20... Training Step: 6079... Training loss: 1.4590... 0.0830 sec/batch
Epoch: 7/20... Training Step: 6080... Training loss: 1.4522... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6081... Training loss: 1.3693... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6082... Training loss: 1.4386... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6083... Training loss: 1.4063... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6084... Training loss: 1.4661... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6085... Training loss: 1.4517... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6086... Training loss: 1.4245... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6087... Training loss: 1.4452... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6088... Training loss: 1.4557... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6089... Training loss: 1.4765... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6090... Training loss: 1.5123... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6091... Training loss: 1.4771... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6092... Training loss: 1.4295... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6093... Training loss: 1.4299... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6094... Training loss: 1.3995... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6095... Training loss: 1.4147... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6096... Training loss: 1.4088... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6097... Training loss: 1.3113... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6098... Training loss: 1.4763... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6099... Training loss: 1.4136... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6100... Training loss: 1.3528... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6101... Training loss: 1.3868... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6102... Training loss: 1.3970... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6103... Training loss: 1.3208... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6104... Training loss: 1.4189... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6105... Training loss: 1.3599... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6106... Training loss: 1.4064... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6107... Training loss: 1.3979... 0.0731 sec/batch
Epoch: 7/20... Training Step: 6108... Training loss: 1.3655... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6109... Training loss: 1.4153... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6110... Training loss: 1.4139... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6111... Training loss: 1.4216... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6112... Training loss: 1.4370... 0.0863 sec/batch
Epoch: 7/20... Training Step: 6113... Training loss: 1.4073... 0.0791 sec/batch
Epoch: 7/20... Training Step: 6114... Training loss: 1.4037... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6115... Training loss: 1.3968... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6116... Training loss: 1.4964... 0.0785 sec/batch
Epoch: 7/20... Training Step: 6117... Training loss: 1.4261... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6118... Training loss: 1.5170... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6119... Training loss: 1.4408... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6120... Training loss: 1.4401... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6121... Training loss: 1.4526... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6122... Training loss: 1.4193... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6123... Training loss: 1.4503... 0.0951 sec/batch
Epoch: 7/20... Training Step: 6124... Training loss: 1.3921... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6125... Training loss: 1.4038... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6126... Training loss: 1.4830... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6127... Training loss: 1.4084... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6128... Training loss: 1.4604... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6129... Training loss: 1.4605... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6130... Training loss: 1.5168... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6131... Training loss: 1.4778... 0.0784 sec/batch
Epoch: 7/20... Training Step: 6132... Training loss: 1.4059... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6133... Training loss: 1.4341... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6134... Training loss: 1.5091... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6135... Training loss: 1.4257... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6136... Training loss: 1.4551... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6137... Training loss: 1.4636... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6138... Training loss: 1.4610... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6139... Training loss: 1.4430... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6140... Training loss: 1.4160... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6141... Training loss: 1.4065... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6142... Training loss: 1.4573... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6143... Training loss: 1.4772... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6144... Training loss: 1.4767... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6145... Training loss: 1.4835... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6146... Training loss: 1.5385... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6147... Training loss: 1.4202... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6148... Training loss: 1.4128... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6149... Training loss: 1.5050... 0.0790 sec/batch
Epoch: 7/20... Training Step: 6150... Training loss: 1.4026... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6151... Training loss: 1.4805... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6152... Training loss: 1.4285... 0.0735 sec/batch
Epoch: 7/20... Training Step: 6153... Training loss: 1.4342... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6154... Training loss: 1.4430... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6155... Training loss: 1.4341... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6156... Training loss: 1.4146... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6157... Training loss: 1.4150... 0.0790 sec/batch
Epoch: 7/20... Training Step: 6158... Training loss: 1.4710... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6159... Training loss: 1.4133... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6160... Training loss: 1.4152... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6161... Training loss: 1.4413... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6162... Training loss: 1.4350... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6163... Training loss: 1.3832... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6164... Training loss: 1.4849... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6165... Training loss: 1.4176... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6166... Training loss: 1.4481... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6167... Training loss: 1.3895... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6168... Training loss: 1.4394... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6169... Training loss: 1.3975... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6170... Training loss: 1.4417... 0.0734 sec/batch
Epoch: 7/20... Training Step: 6171... Training loss: 1.4470... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6172... Training loss: 1.3836... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6173... Training loss: 1.3785... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6174... Training loss: 1.5297... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6175... Training loss: 1.4398... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6176... Training loss: 1.4237... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6177... Training loss: 1.4654... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6178... Training loss: 1.4108... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6179... Training loss: 1.4382... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6180... Training loss: 1.4663... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6181... Training loss: 1.3429... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6182... Training loss: 1.4004... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6183... Training loss: 1.4148... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6184... Training loss: 1.4142... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6185... Training loss: 1.4226... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6186... Training loss: 1.5066... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6187... Training loss: 1.3998... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6188... Training loss: 1.4186... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6189... Training loss: 1.3979... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6190... Training loss: 1.4551... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6191... Training loss: 1.4903... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6192... Training loss: 1.4015... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6193... Training loss: 1.4244... 0.0826 sec/batch
Epoch: 7/20... Training Step: 6194... Training loss: 1.4292... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6195... Training loss: 1.4496... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6196... Training loss: 1.4847... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6197... Training loss: 1.3945... 0.0731 sec/batch
Epoch: 7/20... Training Step: 6198... Training loss: 1.4503... 0.0785 sec/batch
Epoch: 7/20... Training Step: 6199... Training loss: 1.3772... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6200... Training loss: 1.3525... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6201... Training loss: 1.4275... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6202... Training loss: 1.4410... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6203... Training loss: 1.3959... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6204... Training loss: 1.4051... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6205... Training loss: 1.4071... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6206... Training loss: 1.4053... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6207... Training loss: 1.3539... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6208... Training loss: 1.4040... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6209... Training loss: 1.5156... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6210... Training loss: 1.4026... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6211... Training loss: 1.4100... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6212... Training loss: 1.4115... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6213... Training loss: 1.4737... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6214... Training loss: 1.4691... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6215... Training loss: 1.4491... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6216... Training loss: 1.4313... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6217... Training loss: 1.4525... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6218... Training loss: 1.3570... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6219... Training loss: 1.3993... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6220... Training loss: 1.4440... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6221... Training loss: 1.3804... 0.0799 sec/batch
Epoch: 7/20... Training Step: 6222... Training loss: 1.3519... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6223... Training loss: 1.4021... 0.0795 sec/batch
Epoch: 7/20... Training Step: 6224... Training loss: 1.3965... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6225... Training loss: 1.3321... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6226... Training loss: 1.3413... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6227... Training loss: 1.3796... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6228... Training loss: 1.4016... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6229... Training loss: 1.4278... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6230... Training loss: 1.3413... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6231... Training loss: 1.3714... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6232... Training loss: 1.4834... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6233... Training loss: 1.4005... 0.0884 sec/batch
Epoch: 7/20... Training Step: 6234... Training loss: 1.3788... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6235... Training loss: 1.4273... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6236... Training loss: 1.3791... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6237... Training loss: 1.3105... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6238... Training loss: 1.3516... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6239... Training loss: 1.4433... 0.0736 sec/batch
Epoch: 7/20... Training Step: 6240... Training loss: 1.3878... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6241... Training loss: 1.3899... 0.0809 sec/batch
Epoch: 7/20... Training Step: 6242... Training loss: 1.4004... 0.0736 sec/batch
Epoch: 7/20... Training Step: 6243... Training loss: 1.4094... 0.0844 sec/batch
Epoch: 7/20... Training Step: 6244... Training loss: 1.4134... 0.0802 sec/batch
Epoch: 7/20... Training Step: 6245... Training loss: 1.3632... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6246... Training loss: 1.3588... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6247... Training loss: 1.4253... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6248... Training loss: 1.4359... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6249... Training loss: 1.3750... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6250... Training loss: 1.4870... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6251... Training loss: 1.4068... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6252... Training loss: 1.4741... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6253... Training loss: 1.4511... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6254... Training loss: 1.3582... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6255... Training loss: 1.3653... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6256... Training loss: 1.4379... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6257... Training loss: 1.4139... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6258... Training loss: 1.4976... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6259... Training loss: 1.3870... 0.0738 sec/batch
Epoch: 7/20... Training Step: 6260... Training loss: 1.3360... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6261... Training loss: 1.4910... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6262... Training loss: 1.3911... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6263... Training loss: 1.4929... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6264... Training loss: 1.4327... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6265... Training loss: 1.3817... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6266... Training loss: 1.4189... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6267... Training loss: 1.4334... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6268... Training loss: 1.3709... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6269... Training loss: 1.3763... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6270... Training loss: 1.3595... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6271... Training loss: 1.3853... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6272... Training loss: 1.4434... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6273... Training loss: 1.4243... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6274... Training loss: 1.3380... 0.0851 sec/batch
Epoch: 7/20... Training Step: 6275... Training loss: 1.3723... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6276... Training loss: 1.3894... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6277... Training loss: 1.4558... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6278... Training loss: 1.4039... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6279... Training loss: 1.4097... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6280... Training loss: 1.3857... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6281... Training loss: 1.3929... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6282... Training loss: 1.4423... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6283... Training loss: 1.4420... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6284... Training loss: 1.3713... 0.0863 sec/batch
Epoch: 7/20... Training Step: 6285... Training loss: 1.5121... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6286... Training loss: 1.4961... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6287... Training loss: 1.5163... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6288... Training loss: 1.4583... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6289... Training loss: 1.4313... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6290... Training loss: 1.4156... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6291... Training loss: 1.4169... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6292... Training loss: 1.3932... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6293... Training loss: 1.3962... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6294... Training loss: 1.3817... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6295... Training loss: 1.3662... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6296... Training loss: 1.4488... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6297... Training loss: 1.3871... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6298... Training loss: 1.4076... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6299... Training loss: 1.4577... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6300... Training loss: 1.4611... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6301... Training loss: 1.4599... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6302... Training loss: 1.4752... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6303... Training loss: 1.4349... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6304... Training loss: 1.3929... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6305... Training loss: 1.4185... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6306... Training loss: 1.3931... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6307... Training loss: 1.4244... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6308... Training loss: 1.4232... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6309... Training loss: 1.3711... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6310... Training loss: 1.4346... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6311... Training loss: 1.4183... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6312... Training loss: 1.4087... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6313... Training loss: 1.4507... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6314... Training loss: 1.5029... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6315... Training loss: 1.4139... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6316... Training loss: 1.4116... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6317... Training loss: 1.4004... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6318... Training loss: 1.3964... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6319... Training loss: 1.3816... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6320... Training loss: 1.3521... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6321... Training loss: 1.3792... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6322... Training loss: 1.3237... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6323... Training loss: 1.4724... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6324... Training loss: 1.3633... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6325... Training loss: 1.4230... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6326... Training loss: 1.3297... 0.0793 sec/batch
Epoch: 7/20... Training Step: 6327... Training loss: 1.4099... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6328... Training loss: 1.3864... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6329... Training loss: 1.4000... 0.0737 sec/batch
Epoch: 7/20... Training Step: 6330... Training loss: 1.4301... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6331... Training loss: 1.3822... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6332... Training loss: 1.4357... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6333... Training loss: 1.4209... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6334... Training loss: 1.3625... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6335... Training loss: 1.4403... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6336... Training loss: 1.4173... 0.0810 sec/batch
Epoch: 7/20... Training Step: 6337... Training loss: 1.4451... 0.0740 sec/batch
Epoch: 7/20... Training Step: 6338... Training loss: 1.4101... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6339... Training loss: 1.4625... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6340... Training loss: 1.4643... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6341... Training loss: 1.4485... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6342... Training loss: 1.3838... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6343... Training loss: 1.4100... 0.0843 sec/batch
Epoch: 7/20... Training Step: 6344... Training loss: 1.3803... 0.0811 sec/batch
Epoch: 7/20... Training Step: 6345... Training loss: 1.4524... 0.0821 sec/batch
Epoch: 7/20... Training Step: 6346... Training loss: 1.3860... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6347... Training loss: 1.4152... 0.0811 sec/batch
Epoch: 7/20... Training Step: 6348... Training loss: 1.3670... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6349... Training loss: 1.4015... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6350... Training loss: 1.4620... 0.0803 sec/batch
Epoch: 7/20... Training Step: 6351... Training loss: 1.4227... 0.0784 sec/batch
Epoch: 7/20... Training Step: 6352... Training loss: 1.3722... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6353... Training loss: 1.3704... 0.0805 sec/batch
Epoch: 7/20... Training Step: 6354... Training loss: 1.4190... 0.0936 sec/batch
Epoch: 7/20... Training Step: 6355... Training loss: 1.3661... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6356... Training loss: 1.4504... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6357... Training loss: 1.3911... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6358... Training loss: 1.4175... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6359... Training loss: 1.3617... 0.0798 sec/batch
Epoch: 7/20... Training Step: 6360... Training loss: 1.5098... 0.0808 sec/batch
Epoch: 7/20... Training Step: 6361... Training loss: 1.3581... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6362... Training loss: 1.3651... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6363... Training loss: 1.3325... 0.0811 sec/batch
Epoch: 7/20... Training Step: 6364... Training loss: 1.3471... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6365... Training loss: 1.3932... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6366... Training loss: 1.3243... 0.0811 sec/batch
Epoch: 7/20... Training Step: 6367... Training loss: 1.3777... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6368... Training loss: 1.4434... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6369... Training loss: 1.3365... 0.0845 sec/batch
Epoch: 7/20... Training Step: 6370... Training loss: 1.3660... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6371... Training loss: 1.4356... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6372... Training loss: 1.5077... 0.0796 sec/batch
Epoch: 7/20... Training Step: 6373... Training loss: 1.3383... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6374... Training loss: 1.4286... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6375... Training loss: 1.4092... 0.0797 sec/batch
Epoch: 7/20... Training Step: 6376... Training loss: 1.4089... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6377... Training loss: 1.4101... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6378... Training loss: 1.4076... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6379... Training loss: 1.5086... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6380... Training loss: 1.4998... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6381... Training loss: 1.4749... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6382... Training loss: 1.5203... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6383... Training loss: 1.4213... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6384... Training loss: 1.4354... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6385... Training loss: 1.4754... 0.0791 sec/batch
Epoch: 7/20... Training Step: 6386... Training loss: 1.4576... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6387... Training loss: 1.4695... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6388... Training loss: 1.3743... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6389... Training loss: 1.4690... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6390... Training loss: 1.5185... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6391... Training loss: 1.4256... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6392... Training loss: 1.4567... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6393... Training loss: 1.4084... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6394... Training loss: 1.4559... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6395... Training loss: 1.4425... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6396... Training loss: 1.4355... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6397... Training loss: 1.3976... 0.0796 sec/batch
Epoch: 7/20... Training Step: 6398... Training loss: 1.5034... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6399... Training loss: 1.3873... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6400... Training loss: 1.4248... 0.0791 sec/batch
Epoch: 7/20... Training Step: 6401... Training loss: 1.5001... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6402... Training loss: 1.5103... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6403... Training loss: 1.4104... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6404... Training loss: 1.4164... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6405... Training loss: 1.4606... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6406... Training loss: 1.4528... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6407... Training loss: 1.4516... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6408... Training loss: 1.4224... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6409... Training loss: 1.3641... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6410... Training loss: 1.4088... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6411... Training loss: 1.4149... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6412... Training loss: 1.4544... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6413... Training loss: 1.4504... 0.0789 sec/batch
Epoch: 7/20... Training Step: 6414... Training loss: 1.4120... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6415... Training loss: 1.3620... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6416... Training loss: 1.4345... 0.0813 sec/batch
Epoch: 7/20... Training Step: 6417... Training loss: 1.3816... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6418... Training loss: 1.3708... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6419... Training loss: 1.3301... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6420... Training loss: 1.4248... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6421... Training loss: 1.3970... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6422... Training loss: 1.4804... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6423... Training loss: 1.4325... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6424... Training loss: 1.4009... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6425... Training loss: 1.4117... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6426... Training loss: 1.4018... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6427... Training loss: 1.2791... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6428... Training loss: 1.4068... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6429... Training loss: 1.4171... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6430... Training loss: 1.4194... 0.0789 sec/batch
Epoch: 7/20... Training Step: 6431... Training loss: 1.3831... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6432... Training loss: 1.3543... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6433... Training loss: 1.3554... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6434... Training loss: 1.4251... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6435... Training loss: 1.4049... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6436... Training loss: 1.4105... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6437... Training loss: 1.3626... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6438... Training loss: 1.3697... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6439... Training loss: 1.4612... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6440... Training loss: 1.3759... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6441... Training loss: 1.3719... 0.0784 sec/batch
Epoch: 7/20... Training Step: 6442... Training loss: 1.3534... 0.0813 sec/batch
Epoch: 7/20... Training Step: 6443... Training loss: 1.3649... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6444... Training loss: 1.3540... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6445... Training loss: 1.3968... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6446... Training loss: 1.4003... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6447... Training loss: 1.3565... 0.0793 sec/batch
Epoch: 7/20... Training Step: 6448... Training loss: 1.3465... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6449... Training loss: 1.4235... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6450... Training loss: 1.3938... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6451... Training loss: 1.3988... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6452... Training loss: 1.4348... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6453... Training loss: 1.3853... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6454... Training loss: 1.3985... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6455... Training loss: 1.3959... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6456... Training loss: 1.3730... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6457... Training loss: 1.4154... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6458... Training loss: 1.3308... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6459... Training loss: 1.3934... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6460... Training loss: 1.3755... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6461... Training loss: 1.3711... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6462... Training loss: 1.3688... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6463... Training loss: 1.4302... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6464... Training loss: 1.4278... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6465... Training loss: 1.4558... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6466... Training loss: 1.4465... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6467... Training loss: 1.4270... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6468... Training loss: 1.4435... 0.0792 sec/batch
Epoch: 7/20... Training Step: 6469... Training loss: 1.4116... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6470... Training loss: 1.3995... 0.0833 sec/batch
Epoch: 7/20... Training Step: 6471... Training loss: 1.4073... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6472... Training loss: 1.3491... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6473... Training loss: 1.3892... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6474... Training loss: 1.3559... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6475... Training loss: 1.4823... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6476... Training loss: 1.3899... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6477... Training loss: 1.3940... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6478... Training loss: 1.3900... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6479... Training loss: 1.3952... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6480... Training loss: 1.3716... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6481... Training loss: 1.3900... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6482... Training loss: 1.4727... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6483... Training loss: 1.5164... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6484... Training loss: 1.4302... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6485... Training loss: 1.4725... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6486... Training loss: 1.4531... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6487... Training loss: 1.4325... 0.0798 sec/batch
Epoch: 7/20... Training Step: 6488... Training loss: 1.3782... 0.0843 sec/batch
Epoch: 7/20... Training Step: 6489... Training loss: 1.3637... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6490... Training loss: 1.3922... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6491... Training loss: 1.3865... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6492... Training loss: 1.3752... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6493... Training loss: 1.4943... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6494... Training loss: 1.3876... 0.0807 sec/batch
Epoch: 7/20... Training Step: 6495... Training loss: 1.3857... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6496... Training loss: 1.3890... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6497... Training loss: 1.3412... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6498... Training loss: 1.3700... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6499... Training loss: 1.4655... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6500... Training loss: 1.4269... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6501... Training loss: 1.3751... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6502... Training loss: 1.3831... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6503... Training loss: 1.4104... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6504... Training loss: 1.3379... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6505... Training loss: 1.3477... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6506... Training loss: 1.3940... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6507... Training loss: 1.4102... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6508... Training loss: 1.3513... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6509... Training loss: 1.3463... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6510... Training loss: 1.3790... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6511... Training loss: 1.4001... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6512... Training loss: 1.2969... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6513... Training loss: 1.3922... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6514... Training loss: 1.4034... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6515... Training loss: 1.3470... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6516... Training loss: 1.3617... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6517... Training loss: 1.4229... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6518... Training loss: 1.4155... 0.0831 sec/batch
Epoch: 7/20... Training Step: 6519... Training loss: 1.3586... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6520... Training loss: 1.3438... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6521... Training loss: 1.3066... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6522... Training loss: 1.3273... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6523... Training loss: 1.3409... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6524... Training loss: 1.3834... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6525... Training loss: 1.4487... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6526... Training loss: 1.4297... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6527... Training loss: 1.4328... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6528... Training loss: 1.4011... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6529... Training loss: 1.4096... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6530... Training loss: 1.5680... 0.0791 sec/batch
Epoch: 7/20... Training Step: 6531... Training loss: 1.4611... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6532... Training loss: 1.3762... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6533... Training loss: 1.3606... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6534... Training loss: 1.3606... 0.0741 sec/batch
Epoch: 7/20... Training Step: 6535... Training loss: 1.4685... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6536... Training loss: 1.3362... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6537... Training loss: 1.3838... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6538... Training loss: 1.4342... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6539... Training loss: 1.4125... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6540... Training loss: 1.4573... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6541... Training loss: 1.3877... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6542... Training loss: 1.3897... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6543... Training loss: 1.4148... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6544... Training loss: 1.4284... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6545... Training loss: 1.3626... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6546... Training loss: 1.3660... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6547... Training loss: 1.3442... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6548... Training loss: 1.4273... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6549... Training loss: 1.2939... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6550... Training loss: 1.3061... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6551... Training loss: 1.3667... 0.0833 sec/batch
Epoch: 7/20... Training Step: 6552... Training loss: 1.3776... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6553... Training loss: 1.4611... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6554... Training loss: 1.3754... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6555... Training loss: 1.3735... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6556... Training loss: 1.4191... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6557... Training loss: 1.3345... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6558... Training loss: 1.3585... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6559... Training loss: 1.4289... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6560... Training loss: 1.4253... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6561... Training loss: 1.4132... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6562... Training loss: 1.4391... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6563... Training loss: 1.3011... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6564... Training loss: 1.4170... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6565... Training loss: 1.4315... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6566... Training loss: 1.3785... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6567... Training loss: 1.4537... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6568... Training loss: 1.4001... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6569... Training loss: 1.4211... 0.0808 sec/batch
Epoch: 7/20... Training Step: 6570... Training loss: 1.4343... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6571... Training loss: 1.3368... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6572... Training loss: 1.3955... 0.0857 sec/batch
Epoch: 7/20... Training Step: 6573... Training loss: 1.3533... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6574... Training loss: 1.3285... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6575... Training loss: 1.4298... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6576... Training loss: 1.3940... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6577... Training loss: 1.3613... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6578... Training loss: 1.3844... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6579... Training loss: 1.3723... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6580... Training loss: 1.3977... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6581... Training loss: 1.2197... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6582... Training loss: 1.4088... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6583... Training loss: 1.3777... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6584... Training loss: 1.3547... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6585... Training loss: 1.3343... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6586... Training loss: 1.3721... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6587... Training loss: 1.3987... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6588... Training loss: 1.3827... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6589... Training loss: 1.4594... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6590... Training loss: 1.3624... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6591... Training loss: 1.4208... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6592... Training loss: 1.4895... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6593... Training loss: 1.4922... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6594... Training loss: 1.4701... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6595... Training loss: 1.3860... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6596... Training loss: 1.3863... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6597... Training loss: 1.3540... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6598... Training loss: 1.4377... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6599... Training loss: 1.4620... 0.0862 sec/batch
Epoch: 7/20... Training Step: 6600... Training loss: 1.3374... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6601... Training loss: 1.3557... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6602... Training loss: 1.3763... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6603... Training loss: 1.3914... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6604... Training loss: 1.4483... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6605... Training loss: 1.4086... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6606... Training loss: 1.4336... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6607... Training loss: 1.4624... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6608... Training loss: 1.4590... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6609... Training loss: 1.3348... 0.0790 sec/batch
Epoch: 7/20... Training Step: 6610... Training loss: 1.3631... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6611... Training loss: 1.3786... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6612... Training loss: 1.4857... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6613... Training loss: 1.4335... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6614... Training loss: 1.4845... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6615... Training loss: 1.3958... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6616... Training loss: 1.5035... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6617... Training loss: 1.5187... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6618... Training loss: 1.3785... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6619... Training loss: 1.5243... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6620... Training loss: 1.4422... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6621... Training loss: 1.4041... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6622... Training loss: 1.3702... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6623... Training loss: 1.4190... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6624... Training loss: 1.4385... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6625... Training loss: 1.4002... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6626... Training loss: 1.4018... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6627... Training loss: 1.3454... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6628... Training loss: 1.3729... 0.0848 sec/batch
Epoch: 7/20... Training Step: 6629... Training loss: 1.5125... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6630... Training loss: 1.3750... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6631... Training loss: 1.3335... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6632... Training loss: 1.4412... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6633... Training loss: 1.4717... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6634... Training loss: 1.4285... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6635... Training loss: 1.4446... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6636... Training loss: 1.3657... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6637... Training loss: 1.4708... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6638... Training loss: 1.4811... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6639... Training loss: 1.4051... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6640... Training loss: 1.4129... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6641... Training loss: 1.3642... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6642... Training loss: 1.4293... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6643... Training loss: 1.3956... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6644... Training loss: 1.4379... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6645... Training loss: 1.3819... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6646... Training loss: 1.3457... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6647... Training loss: 1.4107... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6648... Training loss: 1.4050... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6649... Training loss: 1.4754... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6650... Training loss: 1.4229... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6651... Training loss: 1.3836... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6652... Training loss: 1.4079... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6653... Training loss: 1.4199... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6654... Training loss: 1.4041... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6655... Training loss: 1.4117... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6656... Training loss: 1.3459... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6657... Training loss: 1.3461... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6658... Training loss: 1.3693... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6659... Training loss: 1.4053... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6660... Training loss: 1.4196... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6661... Training loss: 1.4456... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6662... Training loss: 1.3623... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6663... Training loss: 1.3963... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6664... Training loss: 1.4849... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6665... Training loss: 1.3831... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6666... Training loss: 1.3916... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6667... Training loss: 1.4268... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6668... Training loss: 1.4367... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6669... Training loss: 1.4385... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6670... Training loss: 1.4453... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6671... Training loss: 1.4534... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6672... Training loss: 1.5287... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6673... Training loss: 1.4345... 0.0789 sec/batch
Epoch: 7/20... Training Step: 6674... Training loss: 1.4349... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6675... Training loss: 1.4021... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6676... Training loss: 1.4450... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6677... Training loss: 1.4391... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6678... Training loss: 1.5054... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6679... Training loss: 1.3845... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6680... Training loss: 1.4625... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6681... Training loss: 1.4680... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6682... Training loss: 1.3709... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6683... Training loss: 1.3974... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6684... Training loss: 1.4241... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6685... Training loss: 1.3650... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6686... Training loss: 1.4392... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6687... Training loss: 1.3936... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6688... Training loss: 1.3869... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6689... Training loss: 1.3605... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6690... Training loss: 1.4911... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6691... Training loss: 1.3822... 0.0801 sec/batch
Epoch: 7/20... Training Step: 6692... Training loss: 1.4354... 0.0793 sec/batch
Epoch: 7/20... Training Step: 6693... Training loss: 1.3778... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6694... Training loss: 1.3915... 0.0830 sec/batch
Epoch: 7/20... Training Step: 6695... Training loss: 1.3703... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6696... Training loss: 1.4650... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6697... Training loss: 1.4270... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6698... Training loss: 1.3992... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6699... Training loss: 1.4168... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6700... Training loss: 1.4134... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6701... Training loss: 1.3532... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6702... Training loss: 1.3585... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6703... Training loss: 1.4930... 0.0792 sec/batch
Epoch: 7/20... Training Step: 6704... Training loss: 1.3686... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6705... Training loss: 1.4224... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6706... Training loss: 1.4388... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6707... Training loss: 1.3424... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6708... Training loss: 1.4292... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6709... Training loss: 1.4349... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6710... Training loss: 1.4656... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6711... Training loss: 1.4418... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6712... Training loss: 1.4695... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6713... Training loss: 1.4226... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6714... Training loss: 1.3460... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6715... Training loss: 1.4133... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6716... Training loss: 1.4141... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6717... Training loss: 1.4338... 0.0801 sec/batch
Epoch: 7/20... Training Step: 6718... Training loss: 1.4604... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6719... Training loss: 1.4231... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6720... Training loss: 1.3837... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6721... Training loss: 1.3811... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6722... Training loss: 1.4057... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6723... Training loss: 1.4178... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6724... Training loss: 1.4505... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6725... Training loss: 1.4493... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6726... Training loss: 1.4385... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6727... Training loss: 1.4157... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6728... Training loss: 1.2985... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6729... Training loss: 1.3668... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6730... Training loss: 1.3961... 0.0784 sec/batch
Epoch: 7/20... Training Step: 6731... Training loss: 1.3766... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6732... Training loss: 1.3836... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6733... Training loss: 1.3976... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6734... Training loss: 1.4099... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6735... Training loss: 1.4278... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6736... Training loss: 1.4096... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6737... Training loss: 1.4318... 0.0744 sec/batch
Epoch: 7/20... Training Step: 6738... Training loss: 1.4061... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6739... Training loss: 1.4214... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6740... Training loss: 1.4109... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6741... Training loss: 1.4101... 0.0795 sec/batch
Epoch: 7/20... Training Step: 6742... Training loss: 1.4384... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6743... Training loss: 1.3917... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6744... Training loss: 1.3684... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6745... Training loss: 1.4125... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6746... Training loss: 1.3701... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6747... Training loss: 1.3329... 0.0785 sec/batch
Epoch: 7/20... Training Step: 6748... Training loss: 1.3871... 0.0810 sec/batch
Epoch: 7/20... Training Step: 6749... Training loss: 1.3774... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6750... Training loss: 1.3949... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6751... Training loss: 1.3664... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6752... Training loss: 1.3563... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6753... Training loss: 1.4608... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6754... Training loss: 1.4522... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6755... Training loss: 1.4671... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6756... Training loss: 1.4134... 0.0748 sec/batch
Epoch: 7/20... Training Step: 6757... Training loss: 1.3550... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6758... Training loss: 1.4203... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6759... Training loss: 1.3906... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6760... Training loss: 1.4418... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6761... Training loss: 1.4580... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6762... Training loss: 1.4654... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6763... Training loss: 1.4087... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6764... Training loss: 1.4731... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6765... Training loss: 1.4071... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6766... Training loss: 1.4546... 0.0827 sec/batch
Epoch: 7/20... Training Step: 6767... Training loss: 1.4416... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6768... Training loss: 1.4655... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6769... Training loss: 1.4593... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6770... Training loss: 1.5351... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6771... Training loss: 1.4587... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6772... Training loss: 1.4148... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6773... Training loss: 1.4077... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6774... Training loss: 1.3424... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6775... Training loss: 1.4074... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6776... Training loss: 1.4164... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6777... Training loss: 1.4080... 0.0860 sec/batch
Epoch: 7/20... Training Step: 6778... Training loss: 1.4247... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6779... Training loss: 1.4378... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6780... Training loss: 1.4863... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6781... Training loss: 1.4421... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6782... Training loss: 1.3743... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6783... Training loss: 1.4602... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6784... Training loss: 1.4688... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6785... Training loss: 1.4958... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6786... Training loss: 1.4100... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6787... Training loss: 1.3406... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6788... Training loss: 1.3732... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6789... Training loss: 1.4130... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6790... Training loss: 1.4400... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6791... Training loss: 1.4547... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6792... Training loss: 1.4013... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6793... Training loss: 1.4428... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6794... Training loss: 1.5526... 0.0819 sec/batch
Epoch: 7/20... Training Step: 6795... Training loss: 1.4659... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6796... Training loss: 1.4725... 0.0881 sec/batch
Epoch: 7/20... Training Step: 6797... Training loss: 1.4769... 0.0755 sec/batch
Epoch: 7/20... Training Step: 6798... Training loss: 1.3560... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6799... Training loss: 1.4300... 0.0853 sec/batch
Epoch: 7/20... Training Step: 6800... Training loss: 1.4067... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6801... Training loss: 1.4186... 0.0745 sec/batch
Epoch: 7/20... Training Step: 6802... Training loss: 1.4411... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6803... Training loss: 1.3937... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6804... Training loss: 1.4914... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6805... Training loss: 1.4930... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6806... Training loss: 1.4213... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6807... Training loss: 1.4285... 0.0792 sec/batch
Epoch: 7/20... Training Step: 6808... Training loss: 1.4271... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6809... Training loss: 1.4414... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6810... Training loss: 1.4152... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6811... Training loss: 1.4504... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6812... Training loss: 1.4615... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6813... Training loss: 1.4375... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6814... Training loss: 1.3901... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6815... Training loss: 1.3571... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6816... Training loss: 1.4011... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6817... Training loss: 1.3745... 0.0747 sec/batch
Epoch: 7/20... Training Step: 6818... Training loss: 1.4132... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6819... Training loss: 1.4132... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6820... Training loss: 1.4742... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6821... Training loss: 1.4388... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6822... Training loss: 1.5122... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6823... Training loss: 1.4508... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6824... Training loss: 1.4311... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6825... Training loss: 1.4559... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6826... Training loss: 1.4201... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6827... Training loss: 1.4322... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6828... Training loss: 1.4460... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6829... Training loss: 1.4507... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6830... Training loss: 1.4706... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6831... Training loss: 1.4273... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6832... Training loss: 1.3924... 0.0785 sec/batch
Epoch: 7/20... Training Step: 6833... Training loss: 1.4084... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6834... Training loss: 1.3466... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6835... Training loss: 1.4412... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6836... Training loss: 1.4323... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6837... Training loss: 1.4199... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6838... Training loss: 1.4702... 0.0787 sec/batch
Epoch: 7/20... Training Step: 6839... Training loss: 1.4225... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6840... Training loss: 1.4687... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6841... Training loss: 1.4386... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6842... Training loss: 1.5043... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6843... Training loss: 1.4552... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6844... Training loss: 1.3471... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6845... Training loss: 1.3928... 0.0739 sec/batch
Epoch: 7/20... Training Step: 6846... Training loss: 1.3483... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6847... Training loss: 1.4655... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6848... Training loss: 1.3894... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6849... Training loss: 1.4240... 0.0789 sec/batch
Epoch: 7/20... Training Step: 6850... Training loss: 1.4494... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6851... Training loss: 1.3847... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6852... Training loss: 1.3853... 0.0798 sec/batch
Epoch: 7/20... Training Step: 6853... Training loss: 1.4003... 0.0742 sec/batch
Epoch: 7/20... Training Step: 6854... Training loss: 1.3850... 0.0758 sec/batch
Epoch: 7/20... Training Step: 6855... Training loss: 1.3842... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6856... Training loss: 1.3545... 0.0788 sec/batch
Epoch: 7/20... Training Step: 6857... Training loss: 1.3616... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6858... Training loss: 1.4231... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6859... Training loss: 1.3362... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6860... Training loss: 1.4369... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6861... Training loss: 1.3577... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6862... Training loss: 1.2975... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6863... Training loss: 1.3568... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6864... Training loss: 1.4198... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6865... Training loss: 1.5058... 0.0754 sec/batch
Epoch: 7/20... Training Step: 6866... Training loss: 1.4249... 0.0743 sec/batch
Epoch: 7/20... Training Step: 6867... Training loss: 1.3988... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6868... Training loss: 1.4032... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6869... Training loss: 1.3749... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6870... Training loss: 1.4566... 0.0792 sec/batch
Epoch: 7/20... Training Step: 6871... Training loss: 1.4059... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6872... Training loss: 1.4372... 0.0783 sec/batch
Epoch: 7/20... Training Step: 6873... Training loss: 1.5295... 0.0793 sec/batch
Epoch: 7/20... Training Step: 6874... Training loss: 1.4575... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6875... Training loss: 1.4368... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6876... Training loss: 1.3970... 0.0806 sec/batch
Epoch: 7/20... Training Step: 6877... Training loss: 1.4338... 0.0781 sec/batch
Epoch: 7/20... Training Step: 6878... Training loss: 1.4423... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6879... Training loss: 1.4051... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6880... Training loss: 1.4025... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6881... Training loss: 1.5184... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6882... Training loss: 1.4747... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6883... Training loss: 1.4321... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6884... Training loss: 1.4989... 0.0797 sec/batch
Epoch: 7/20... Training Step: 6885... Training loss: 1.4113... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6886... Training loss: 1.4339... 0.0798 sec/batch
Epoch: 7/20... Training Step: 6887... Training loss: 1.4286... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6888... Training loss: 1.4669... 0.0811 sec/batch
Epoch: 7/20... Training Step: 6889... Training loss: 1.3942... 0.0757 sec/batch
Epoch: 7/20... Training Step: 6890... Training loss: 1.4221... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6891... Training loss: 1.3920... 0.0784 sec/batch
Epoch: 7/20... Training Step: 6892... Training loss: 1.5110... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6893... Training loss: 1.4516... 0.0761 sec/batch
Epoch: 7/20... Training Step: 6894... Training loss: 1.4494... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6895... Training loss: 1.4902... 0.0762 sec/batch
Epoch: 7/20... Training Step: 6896... Training loss: 1.4765... 0.0749 sec/batch
Epoch: 7/20... Training Step: 6897... Training loss: 1.4639... 0.0756 sec/batch
Epoch: 7/20... Training Step: 6898... Training loss: 1.4076... 0.0778 sec/batch
Epoch: 7/20... Training Step: 6899... Training loss: 1.4143... 0.0759 sec/batch
Epoch: 7/20... Training Step: 6900... Training loss: 1.3718... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6901... Training loss: 1.4382... 0.0772 sec/batch
Epoch: 7/20... Training Step: 6902... Training loss: 1.3340... 0.0770 sec/batch
Epoch: 7/20... Training Step: 6903... Training loss: 1.4118... 0.0790 sec/batch
Epoch: 7/20... Training Step: 6904... Training loss: 1.3783... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6905... Training loss: 1.4215... 0.0753 sec/batch
Epoch: 7/20... Training Step: 6906... Training loss: 1.4571... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6907... Training loss: 1.4689... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6908... Training loss: 1.4354... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6909... Training loss: 1.4166... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6910... Training loss: 1.4460... 0.0777 sec/batch
Epoch: 7/20... Training Step: 6911... Training loss: 1.3763... 0.0768 sec/batch
Epoch: 7/20... Training Step: 6912... Training loss: 1.3477... 0.0767 sec/batch
Epoch: 7/20... Training Step: 6913... Training loss: 1.4204... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6914... Training loss: 1.3943... 0.0786 sec/batch
Epoch: 7/20... Training Step: 6915... Training loss: 1.4344... 0.0773 sec/batch
Epoch: 7/20... Training Step: 6916... Training loss: 1.3978... 0.0766 sec/batch
Epoch: 7/20... Training Step: 6917... Training loss: 1.4802... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6918... Training loss: 1.4752... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6919... Training loss: 1.4653... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6920... Training loss: 1.4603... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6921... Training loss: 1.4480... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6922... Training loss: 1.6081... 0.0752 sec/batch
Epoch: 7/20... Training Step: 6923... Training loss: 1.3920... 0.0774 sec/batch
Epoch: 7/20... Training Step: 6924... Training loss: 1.5117... 0.0750 sec/batch
Epoch: 7/20... Training Step: 6925... Training loss: 1.5098... 0.0763 sec/batch
Epoch: 7/20... Training Step: 6926... Training loss: 1.4156... 0.0746 sec/batch
Epoch: 7/20... Training Step: 6927... Training loss: 1.4719... 0.0821 sec/batch
Epoch: 7/20... Training Step: 6928... Training loss: 1.4736... 0.0785 sec/batch
Epoch: 7/20... Training Step: 6929... Training loss: 1.4946... 0.0776 sec/batch
Epoch: 7/20... Training Step: 6930... Training loss: 1.4811... 0.0769 sec/batch
Epoch: 7/20... Training Step: 6931... Training loss: 1.4521... 0.0764 sec/batch
Epoch: 7/20... Training Step: 6932... Training loss: 1.4008... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6933... Training loss: 1.3409... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6934... Training loss: 1.5153... 0.0760 sec/batch
Epoch: 7/20... Training Step: 6935... Training loss: 1.4490... 0.0765 sec/batch
Epoch: 7/20... Training Step: 6936... Training loss: 1.4247... 0.0782 sec/batch
Epoch: 7/20... Training Step: 6937... Training loss: 1.4299... 0.0775 sec/batch
Epoch: 7/20... Training Step: 6938... Training loss: 1.3751... 0.0751 sec/batch
Epoch: 7/20... Training Step: 6939... Training loss: 1.4039... 0.0780 sec/batch
Epoch: 7/20... Training Step: 6940... Training loss: 1.4007... 0.0792 sec/batch
Epoch: 7/20... Training Step: 6941... Training loss: 1.4430... 0.0794 sec/batch
Epoch: 7/20... Training Step: 6942... Training loss: 1.3228... 0.0779 sec/batch
Epoch: 7/20... Training Step: 6943... Training loss: 1.3814... 0.0771 sec/batch
Epoch: 7/20... Training Step: 6944... Training loss: 1.4437... 0.0750 sec/batch
Epoch: 8/20... Training Step: 6945... Training loss: 1.4989... 0.0751 sec/batch
Epoch: 8/20... Training Step: 6946... Training loss: 1.3731... 0.0774 sec/batch
Epoch: 8/20... Training Step: 6947... Training loss: 1.4683... 0.0745 sec/batch
Epoch: 8/20... Training Step: 6948... Training loss: 1.4287... 0.0757 sec/batch
Epoch: 8/20... Training Step: 6949... Training loss: 1.3303... 0.0749 sec/batch
Epoch: 8/20... Training Step: 6950... Training loss: 1.2809... 0.0757 sec/batch
Epoch: 8/20... Training Step: 6951... Training loss: 1.3928... 0.0770 sec/batch
Epoch: 8/20... Training Step: 6952... Training loss: 1.4366... 0.0752 sec/batch
Epoch: 8/20... Training Step: 6953... Training loss: 1.4335... 0.0771 sec/batch
Epoch: 8/20... Training Step: 6954... Training loss: 1.3071... 0.0782 sec/batch
Epoch: 8/20... Training Step: 6955... Training loss: 1.4779... 0.0784 sec/batch
Epoch: 8/20... Training Step: 6956... Training loss: 1.3543... 0.0746 sec/batch
Epoch: 8/20... Training Step: 6957... Training loss: 1.3996... 0.0764 sec/batch
Epoch: 8/20... Training Step: 6958... Training loss: 1.4225... 0.0773 sec/batch
Epoch: 8/20... Training Step: 6959... Training loss: 1.4628... 0.0750 sec/batch
Epoch: 8/20... Training Step: 6960... Training loss: 1.3536... 0.0821 sec/batch
Epoch: 8/20... Training Step: 6961... Training loss: 1.4107... 0.0773 sec/batch
Epoch: 8/20... Training Step: 6962... Training loss: 1.4189... 0.0750 sec/batch
Epoch: 8/20... Training Step: 6963... Training loss: 1.4025... 0.0785 sec/batch
Epoch: 8/20... Training Step: 6964... Training loss: 1.3748... 0.0757 sec/batch
Epoch: 8/20... Training Step: 6965... Training loss: 1.4070... 0.0749 sec/batch
Epoch: 8/20... Training Step: 6966... Training loss: 1.3728... 0.0773 sec/batch
Epoch: 8/20... Training Step: 6967... Training loss: 1.3811... 0.0746 sec/batch
Epoch: 8/20... Training Step: 6968... Training loss: 1.3357... 0.0770 sec/batch
Epoch: 8/20... Training Step: 6969... Training loss: 1.4427... 0.0774 sec/batch
Epoch: 8/20... Training Step: 6970... Training loss: 1.3391... 0.0752 sec/batch
Epoch: 8/20... Training Step: 6971... Training loss: 1.3896... 0.0749 sec/batch
Epoch: 8/20... Training Step: 6972... Training loss: 1.3202... 0.0764 sec/batch
Epoch: 8/20... Training Step: 6973... Training loss: 1.4074... 0.0748 sec/batch
Epoch: 8/20... Training Step: 6974... Training loss: 1.4425... 0.0749 sec/batch
Epoch: 8/20... Training Step: 6975... Training loss: 1.4306... 0.0758 sec/batch
Epoch: 8/20... Training Step: 6976... Training loss: 1.4344... 0.0768 sec/batch
Epoch: 8/20... Training Step: 6977... Training loss: 1.3919... 0.0756 sec/batch
Epoch: 8/20... Training Step: 6978... Training loss: 1.4365... 0.0811 sec/batch
Epoch: 8/20... Training Step: 6979... Training loss: 1.4922... 0.0754 sec/batch
Epoch: 8/20... Training Step: 6980... Training loss: 1.4629... 0.0782 sec/batch
Epoch: 8/20... Training Step: 6981... Training loss: 1.4100... 0.0750 sec/batch
Epoch: 8/20... Training Step: 6982... Training loss: 1.4639... 0.0741 sec/batch
Epoch: 8/20... Training Step: 6983... Training loss: 1.4154... 0.0757 sec/batch
Epoch: 8/20... Training Step: 6984... Training loss: 1.3789... 0.0781 sec/batch
Epoch: 8/20... Training Step: 6985... Training loss: 1.4175... 0.0768 sec/batch
Epoch: 8/20... Training Step: 6986... Training loss: 1.3887... 0.0784 sec/batch
Epoch: 8/20... Training Step: 6987... Training loss: 1.3418... 0.0766 sec/batch
Epoch: 8/20... Training Step: 6988... Training loss: 1.3828... 0.0780 sec/batch
Epoch: 8/20... Training Step: 6989... Training loss: 1.4394... 0.0749 sec/batch
Epoch: 8/20... Training Step: 6990... Training loss: 1.3992... 0.0787 sec/batch
Epoch: 8/20... Training Step: 6991... Training loss: 1.4395... 0.0765 sec/batch
Epoch: 8/20... Training Step: 6992... Training loss: 1.3753... 0.0779 sec/batch
Epoch: 8/20... Training Step: 6993... Training loss: 1.4384... 0.0756 sec/batch
Epoch: 8/20... Training Step: 6994... Training loss: 1.4197... 0.0765 sec/batch
Epoch: 8/20... Training Step: 6995... Training loss: 1.3178... 0.0763 sec/batch
Epoch: 8/20... Training Step: 6996... Training loss: 1.4371... 0.0784 sec/batch
Epoch: 8/20... Training Step: 6997... Training loss: 1.3445... 0.0784 sec/batch
Epoch: 8/20... Training Step: 6998... Training loss: 1.3732... 0.0742 sec/batch
Epoch: 8/20... Training Step: 6999... Training loss: 1.3800... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7000... Training loss: 1.3421... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7001... Training loss: 1.4297... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7002... Training loss: 1.4392... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7003... Training loss: 1.3887... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7004... Training loss: 1.4239... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7005... Training loss: 1.2998... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7006... Training loss: 1.3495... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7007... Training loss: 1.3801... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7008... Training loss: 1.3628... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7009... Training loss: 1.4338... 0.0740 sec/batch
Epoch: 8/20... Training Step: 7010... Training loss: 1.3821... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7011... Training loss: 1.3794... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7012... Training loss: 1.3970... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7013... Training loss: 1.4129... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7014... Training loss: 1.4010... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7015... Training loss: 1.4421... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7016... Training loss: 1.4432... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7017... Training loss: 1.4116... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7018... Training loss: 1.3636... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7019... Training loss: 1.3712... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7020... Training loss: 1.3436... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7021... Training loss: 1.3744... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7022... Training loss: 1.3671... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7023... Training loss: 1.3929... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7024... Training loss: 1.4391... 0.0798 sec/batch
Epoch: 8/20... Training Step: 7025... Training loss: 1.4100... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7026... Training loss: 1.3845... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7027... Training loss: 1.4280... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7028... Training loss: 1.4200... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7029... Training loss: 1.3639... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7030... Training loss: 1.4453... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7031... Training loss: 1.4013... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7032... Training loss: 1.4000... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7033... Training loss: 1.4316... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7034... Training loss: 1.4690... 0.0816 sec/batch
Epoch: 8/20... Training Step: 7035... Training loss: 1.4720... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7036... Training loss: 1.3720... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7037... Training loss: 1.3434... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7038... Training loss: 1.3225... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7039... Training loss: 1.3243... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7040... Training loss: 1.4224... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7041... Training loss: 1.3892... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7042... Training loss: 1.3431... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7043... Training loss: 1.3537... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7044... Training loss: 1.3727... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7045... Training loss: 1.3569... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7046... Training loss: 1.3668... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7047... Training loss: 1.4010... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7048... Training loss: 1.4029... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7049... Training loss: 1.4166... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7050... Training loss: 1.3954... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7051... Training loss: 1.3431... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7052... Training loss: 1.4560... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7053... Training loss: 1.3963... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7054... Training loss: 1.4570... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7055... Training loss: 1.4021... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7056... Training loss: 1.3908... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7057... Training loss: 1.4793... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7058... Training loss: 1.4325... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7059... Training loss: 1.3378... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7060... Training loss: 1.5236... 0.0789 sec/batch
Epoch: 8/20... Training Step: 7061... Training loss: 1.4375... 0.0815 sec/batch
Epoch: 8/20... Training Step: 7062... Training loss: 1.5191... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7063... Training loss: 1.4055... 0.0821 sec/batch
Epoch: 8/20... Training Step: 7064... Training loss: 1.3860... 0.0802 sec/batch
Epoch: 8/20... Training Step: 7065... Training loss: 1.5418... 0.0741 sec/batch
Epoch: 8/20... Training Step: 7066... Training loss: 1.3624... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7067... Training loss: 1.4445... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7068... Training loss: 1.4133... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7069... Training loss: 1.4778... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7070... Training loss: 1.4858... 0.0839 sec/batch
Epoch: 8/20... Training Step: 7071... Training loss: 1.4564... 0.0796 sec/batch
Epoch: 8/20... Training Step: 7072... Training loss: 1.4309... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7073... Training loss: 1.3549... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7074... Training loss: 1.4115... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7075... Training loss: 1.3741... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7076... Training loss: 1.4562... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7077... Training loss: 1.4246... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7078... Training loss: 1.3958... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7079... Training loss: 1.4632... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7080... Training loss: 1.4322... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7081... Training loss: 1.4418... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7082... Training loss: 1.4862... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7083... Training loss: 1.4702... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7084... Training loss: 1.4170... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7085... Training loss: 1.4131... 0.0793 sec/batch
Epoch: 8/20... Training Step: 7086... Training loss: 1.3789... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7087... Training loss: 1.3912... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7088... Training loss: 1.4002... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7089... Training loss: 1.2754... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7090... Training loss: 1.4553... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7091... Training loss: 1.4138... 0.0797 sec/batch
Epoch: 8/20... Training Step: 7092... Training loss: 1.3153... 0.0800 sec/batch
Epoch: 8/20... Training Step: 7093... Training loss: 1.3539... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7094... Training loss: 1.3398... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7095... Training loss: 1.2827... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7096... Training loss: 1.3990... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7097... Training loss: 1.3566... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7098... Training loss: 1.3915... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7099... Training loss: 1.3616... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7100... Training loss: 1.3553... 0.0826 sec/batch
Epoch: 8/20... Training Step: 7101... Training loss: 1.4000... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7102... Training loss: 1.3879... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7103... Training loss: 1.3908... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7104... Training loss: 1.4206... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7105... Training loss: 1.3625... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7106... Training loss: 1.3692... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7107... Training loss: 1.3846... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7108... Training loss: 1.4750... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7109... Training loss: 1.4289... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7110... Training loss: 1.4951... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7111... Training loss: 1.4279... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7112... Training loss: 1.4109... 0.0797 sec/batch
Epoch: 8/20... Training Step: 7113... Training loss: 1.4334... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7114... Training loss: 1.4084... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7115... Training loss: 1.4197... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7116... Training loss: 1.3701... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7117... Training loss: 1.3783... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7118... Training loss: 1.4636... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7119... Training loss: 1.3685... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7120... Training loss: 1.4107... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7121... Training loss: 1.4180... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7122... Training loss: 1.4889... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7123... Training loss: 1.4400... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7124... Training loss: 1.3769... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7125... Training loss: 1.4117... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7126... Training loss: 1.4738... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7127... Training loss: 1.3955... 0.0836 sec/batch
Epoch: 8/20... Training Step: 7128... Training loss: 1.4336... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7129... Training loss: 1.4221... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7130... Training loss: 1.4485... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7131... Training loss: 1.4334... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7132... Training loss: 1.3761... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7133... Training loss: 1.4056... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7134... Training loss: 1.4667... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7135... Training loss: 1.4946... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7136... Training loss: 1.4188... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7137... Training loss: 1.4453... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7138... Training loss: 1.5276... 0.0834 sec/batch
Epoch: 8/20... Training Step: 7139... Training loss: 1.4029... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7140... Training loss: 1.3498... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7141... Training loss: 1.4744... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7142... Training loss: 1.3992... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7143... Training loss: 1.4571... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7144... Training loss: 1.3978... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7145... Training loss: 1.4063... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7146... Training loss: 1.4080... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7147... Training loss: 1.4382... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7148... Training loss: 1.3932... 0.0800 sec/batch
Epoch: 8/20... Training Step: 7149... Training loss: 1.3753... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7150... Training loss: 1.4577... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7151... Training loss: 1.4004... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7152... Training loss: 1.3731... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7153... Training loss: 1.4403... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7154... Training loss: 1.4125... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7155... Training loss: 1.3734... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7156... Training loss: 1.4565... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7157... Training loss: 1.3747... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7158... Training loss: 1.4414... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7159... Training loss: 1.3840... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7160... Training loss: 1.4214... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7161... Training loss: 1.3756... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7162... Training loss: 1.4013... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7163... Training loss: 1.4279... 0.0837 sec/batch
Epoch: 8/20... Training Step: 7164... Training loss: 1.3390... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7165... Training loss: 1.3685... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7166... Training loss: 1.5085... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7167... Training loss: 1.4073... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7168... Training loss: 1.4300... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7169... Training loss: 1.4525... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7170... Training loss: 1.3787... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7171... Training loss: 1.4106... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7172... Training loss: 1.4389... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7173... Training loss: 1.3285... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7174... Training loss: 1.3576... 0.0741 sec/batch
Epoch: 8/20... Training Step: 7175... Training loss: 1.3828... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7176... Training loss: 1.4004... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7177... Training loss: 1.3975... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7178... Training loss: 1.4753... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7179... Training loss: 1.3886... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7180... Training loss: 1.4037... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7181... Training loss: 1.3707... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7182... Training loss: 1.4502... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7183... Training loss: 1.4532... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7184... Training loss: 1.3979... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7185... Training loss: 1.3951... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7186... Training loss: 1.4217... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7187... Training loss: 1.4386... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7188... Training loss: 1.4298... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7189... Training loss: 1.3832... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7190... Training loss: 1.4331... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7191... Training loss: 1.3582... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7192... Training loss: 1.3082... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7193... Training loss: 1.3918... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7194... Training loss: 1.4126... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7195... Training loss: 1.3710... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7196... Training loss: 1.3808... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7197... Training loss: 1.3900... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7198... Training loss: 1.3587... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7199... Training loss: 1.3325... 0.0803 sec/batch
Epoch: 8/20... Training Step: 7200... Training loss: 1.3565... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7201... Training loss: 1.4994... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7202... Training loss: 1.3910... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7203... Training loss: 1.4029... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7204... Training loss: 1.3990... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7205... Training loss: 1.4558... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7206... Training loss: 1.4407... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7207... Training loss: 1.4466... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7208... Training loss: 1.4187... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7209... Training loss: 1.4008... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7210... Training loss: 1.3302... 0.0805 sec/batch
Epoch: 8/20... Training Step: 7211... Training loss: 1.3627... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7212... Training loss: 1.4158... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7213... Training loss: 1.3339... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7214... Training loss: 1.3249... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7215... Training loss: 1.3891... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7216... Training loss: 1.3662... 0.0847 sec/batch
Epoch: 8/20... Training Step: 7217... Training loss: 1.2950... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7218... Training loss: 1.3253... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7219... Training loss: 1.3835... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7220... Training loss: 1.3952... 0.0793 sec/batch
Epoch: 8/20... Training Step: 7221... Training loss: 1.4217... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7222... Training loss: 1.2965... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7223... Training loss: 1.3497... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7224... Training loss: 1.4553... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7225... Training loss: 1.3873... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7226... Training loss: 1.3643... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7227... Training loss: 1.4274... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7228... Training loss: 1.3623... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7229... Training loss: 1.3091... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7230... Training loss: 1.3395... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7231... Training loss: 1.4140... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7232... Training loss: 1.3751... 0.0801 sec/batch
Epoch: 8/20... Training Step: 7233... Training loss: 1.3610... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7234... Training loss: 1.3429... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7235... Training loss: 1.3705... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7236... Training loss: 1.4023... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7237... Training loss: 1.3186... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7238... Training loss: 1.3563... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7239... Training loss: 1.3994... 0.0740 sec/batch
Epoch: 8/20... Training Step: 7240... Training loss: 1.4038... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7241... Training loss: 1.3408... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7242... Training loss: 1.4519... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7243... Training loss: 1.3955... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7244... Training loss: 1.4301... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7245... Training loss: 1.4339... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7246... Training loss: 1.3519... 0.0798 sec/batch
Epoch: 8/20... Training Step: 7247... Training loss: 1.3343... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7248... Training loss: 1.4184... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7249... Training loss: 1.3929... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7250... Training loss: 1.4631... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7251... Training loss: 1.3377... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7252... Training loss: 1.3169... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7253... Training loss: 1.4759... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7254... Training loss: 1.3805... 0.0812 sec/batch
Epoch: 8/20... Training Step: 7255... Training loss: 1.4626... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7256... Training loss: 1.4058... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7257... Training loss: 1.3584... 0.0835 sec/batch
Epoch: 8/20... Training Step: 7258... Training loss: 1.3839... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7259... Training loss: 1.4108... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7260... Training loss: 1.3483... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7261... Training loss: 1.3544... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7262... Training loss: 1.3324... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7263... Training loss: 1.3705... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7264... Training loss: 1.4080... 0.0866 sec/batch
Epoch: 8/20... Training Step: 7265... Training loss: 1.3930... 0.0799 sec/batch
Epoch: 8/20... Training Step: 7266... Training loss: 1.3227... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7267... Training loss: 1.3635... 0.0814 sec/batch
Epoch: 8/20... Training Step: 7268... Training loss: 1.3708... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7269... Training loss: 1.4317... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7270... Training loss: 1.3740... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7271... Training loss: 1.3958... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7272... Training loss: 1.3657... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7273... Training loss: 1.3789... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7274... Training loss: 1.4100... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7275... Training loss: 1.4279... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7276... Training loss: 1.3457... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7277... Training loss: 1.4802... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7278... Training loss: 1.4799... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7279... Training loss: 1.4502... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7280... Training loss: 1.4184... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7281... Training loss: 1.3722... 0.0806 sec/batch
Epoch: 8/20... Training Step: 7282... Training loss: 1.3883... 0.0792 sec/batch
Epoch: 8/20... Training Step: 7283... Training loss: 1.3937... 0.0835 sec/batch
Epoch: 8/20... Training Step: 7284... Training loss: 1.3650... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7285... Training loss: 1.3944... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7286... Training loss: 1.3533... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7287... Training loss: 1.3310... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7288... Training loss: 1.4274... 0.0789 sec/batch
Epoch: 8/20... Training Step: 7289... Training loss: 1.3831... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7290... Training loss: 1.4019... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7291... Training loss: 1.4370... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7292... Training loss: 1.4401... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7293... Training loss: 1.4348... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7294... Training loss: 1.4570... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7295... Training loss: 1.4172... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7296... Training loss: 1.3772... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7297... Training loss: 1.4013... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7298... Training loss: 1.3608... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7299... Training loss: 1.4211... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7300... Training loss: 1.4073... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7301... Training loss: 1.3440... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7302... Training loss: 1.3940... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7303... Training loss: 1.4020... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7304... Training loss: 1.3969... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7305... Training loss: 1.4412... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7306... Training loss: 1.4812... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7307... Training loss: 1.3812... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7308... Training loss: 1.4002... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7309... Training loss: 1.3872... 0.0837 sec/batch
Epoch: 8/20... Training Step: 7310... Training loss: 1.3504... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7311... Training loss: 1.3698... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7312... Training loss: 1.3149... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7313... Training loss: 1.3352... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7314... Training loss: 1.3310... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7315... Training loss: 1.4349... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7316... Training loss: 1.3314... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7317... Training loss: 1.3925... 0.0792 sec/batch
Epoch: 8/20... Training Step: 7318... Training loss: 1.3208... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7319... Training loss: 1.3689... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7320... Training loss: 1.3447... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7321... Training loss: 1.3690... 0.0826 sec/batch
Epoch: 8/20... Training Step: 7322... Training loss: 1.3924... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7323... Training loss: 1.3615... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7324... Training loss: 1.4097... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7325... Training loss: 1.3874... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7326... Training loss: 1.3597... 0.0790 sec/batch
Epoch: 8/20... Training Step: 7327... Training loss: 1.4166... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7328... Training loss: 1.4117... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7329... Training loss: 1.4387... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7330... Training loss: 1.3923... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7331... Training loss: 1.4314... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7332... Training loss: 1.4328... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7333... Training loss: 1.4308... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7334... Training loss: 1.3570... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7335... Training loss: 1.4208... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7336... Training loss: 1.3589... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7337... Training loss: 1.4307... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7338... Training loss: 1.3720... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7339... Training loss: 1.3864... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7340... Training loss: 1.3827... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7341... Training loss: 1.3858... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7342... Training loss: 1.4456... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7343... Training loss: 1.4005... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7344... Training loss: 1.3841... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7345... Training loss: 1.3569... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7346... Training loss: 1.4235... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7347... Training loss: 1.3516... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7348... Training loss: 1.4466... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7349... Training loss: 1.3740... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7350... Training loss: 1.3945... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7351... Training loss: 1.3227... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7352... Training loss: 1.4736... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7353... Training loss: 1.3488... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7354... Training loss: 1.3376... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7355... Training loss: 1.2892... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7356... Training loss: 1.3499... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7357... Training loss: 1.3867... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7358... Training loss: 1.3272... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7359... Training loss: 1.3418... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7360... Training loss: 1.4273... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7361... Training loss: 1.3331... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7362... Training loss: 1.3521... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7363... Training loss: 1.4027... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7364... Training loss: 1.4733... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7365... Training loss: 1.3486... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7366... Training loss: 1.4013... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7367... Training loss: 1.3883... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7368... Training loss: 1.3913... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7369... Training loss: 1.3922... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7370... Training loss: 1.3722... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7371... Training loss: 1.4554... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7372... Training loss: 1.4767... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7373... Training loss: 1.4384... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7374... Training loss: 1.5064... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7375... Training loss: 1.3809... 0.0798 sec/batch
Epoch: 8/20... Training Step: 7376... Training loss: 1.4267... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7377... Training loss: 1.4618... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7378... Training loss: 1.3873... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7379... Training loss: 1.4393... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7380... Training loss: 1.3624... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7381... Training loss: 1.4075... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7382... Training loss: 1.5106... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7383... Training loss: 1.4048... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7384... Training loss: 1.4313... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7385... Training loss: 1.3948... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7386... Training loss: 1.4297... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7387... Training loss: 1.4119... 0.0807 sec/batch
Epoch: 8/20... Training Step: 7388... Training loss: 1.4173... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7389... Training loss: 1.3876... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7390... Training loss: 1.5008... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7391... Training loss: 1.3854... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7392... Training loss: 1.3897... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7393... Training loss: 1.4876... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7394... Training loss: 1.4806... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7395... Training loss: 1.3763... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7396... Training loss: 1.4083... 0.0835 sec/batch
Epoch: 8/20... Training Step: 7397... Training loss: 1.4354... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7398... Training loss: 1.4257... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7399... Training loss: 1.4168... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7400... Training loss: 1.3877... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7401... Training loss: 1.3740... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7402... Training loss: 1.3756... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7403... Training loss: 1.3849... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7404... Training loss: 1.4161... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7405... Training loss: 1.4191... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7406... Training loss: 1.3955... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7407... Training loss: 1.3860... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7408... Training loss: 1.4004... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7409... Training loss: 1.3382... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7410... Training loss: 1.3537... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7411... Training loss: 1.2855... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7412... Training loss: 1.4020... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7413... Training loss: 1.3705... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7414... Training loss: 1.4662... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7415... Training loss: 1.4089... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7416... Training loss: 1.3923... 0.0839 sec/batch
Epoch: 8/20... Training Step: 7417... Training loss: 1.3853... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7418... Training loss: 1.3992... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7419... Training loss: 1.2818... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7420... Training loss: 1.3681... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7421... Training loss: 1.4066... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7422... Training loss: 1.4024... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7423... Training loss: 1.3490... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7424... Training loss: 1.3451... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7425... Training loss: 1.3361... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7426... Training loss: 1.4029... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7427... Training loss: 1.3840... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7428... Training loss: 1.3954... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7429... Training loss: 1.3065... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7430... Training loss: 1.3487... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7431... Training loss: 1.4351... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7432... Training loss: 1.3434... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7433... Training loss: 1.3498... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7434... Training loss: 1.3293... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7435... Training loss: 1.3594... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7436... Training loss: 1.3276... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7437... Training loss: 1.3938... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7438... Training loss: 1.3822... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7439... Training loss: 1.3472... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7440... Training loss: 1.3314... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7441... Training loss: 1.4061... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7442... Training loss: 1.3668... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7443... Training loss: 1.3685... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7444... Training loss: 1.4081... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7445... Training loss: 1.3499... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7446... Training loss: 1.4005... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7447... Training loss: 1.3802... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7448... Training loss: 1.3754... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7449... Training loss: 1.3976... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7450... Training loss: 1.2931... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7451... Training loss: 1.3499... 0.0741 sec/batch
Epoch: 8/20... Training Step: 7452... Training loss: 1.3297... 0.0796 sec/batch
Epoch: 8/20... Training Step: 7453... Training loss: 1.3447... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7454... Training loss: 1.3613... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7455... Training loss: 1.4278... 0.0805 sec/batch
Epoch: 8/20... Training Step: 7456... Training loss: 1.3929... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7457... Training loss: 1.4551... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7458... Training loss: 1.4317... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7459... Training loss: 1.3722... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7460... Training loss: 1.3996... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7461... Training loss: 1.3891... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7462... Training loss: 1.3774... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7463... Training loss: 1.4143... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7464... Training loss: 1.3436... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7465... Training loss: 1.3758... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7466... Training loss: 1.3348... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7467... Training loss: 1.4556... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7468... Training loss: 1.3569... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7469... Training loss: 1.3721... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7470... Training loss: 1.3675... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7471... Training loss: 1.3722... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7472... Training loss: 1.3456... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7473... Training loss: 1.3561... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7474... Training loss: 1.4247... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7475... Training loss: 1.4972... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7476... Training loss: 1.4156... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7477... Training loss: 1.4426... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7478... Training loss: 1.4071... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7479... Training loss: 1.3743... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7480... Training loss: 1.3652... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7481... Training loss: 1.3540... 0.0798 sec/batch
Epoch: 8/20... Training Step: 7482... Training loss: 1.3560... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7483... Training loss: 1.3583... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7484... Training loss: 1.3602... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7485... Training loss: 1.4832... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7486... Training loss: 1.3560... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7487... Training loss: 1.3672... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7488... Training loss: 1.3635... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7489... Training loss: 1.3351... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7490... Training loss: 1.3573... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7491... Training loss: 1.4444... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7492... Training loss: 1.4169... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7493... Training loss: 1.3495... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7494... Training loss: 1.3418... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7495... Training loss: 1.3674... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7496... Training loss: 1.3184... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7497... Training loss: 1.3058... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7498... Training loss: 1.3766... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7499... Training loss: 1.3956... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7500... Training loss: 1.3246... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7501... Training loss: 1.3234... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7502... Training loss: 1.3580... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7503... Training loss: 1.4052... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7504... Training loss: 1.3106... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7505... Training loss: 1.3607... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7506... Training loss: 1.3804... 0.0793 sec/batch
Epoch: 8/20... Training Step: 7507... Training loss: 1.3477... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7508... Training loss: 1.3416... 0.0740 sec/batch
Epoch: 8/20... Training Step: 7509... Training loss: 1.4056... 0.0844 sec/batch
Epoch: 8/20... Training Step: 7510... Training loss: 1.3967... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7511... Training loss: 1.3341... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7512... Training loss: 1.3141... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7513... Training loss: 1.3046... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7514... Training loss: 1.3305... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7515... Training loss: 1.3185... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7516... Training loss: 1.3491... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7517... Training loss: 1.4406... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7518... Training loss: 1.4094... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7519... Training loss: 1.4157... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7520... Training loss: 1.3912... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7521... Training loss: 1.3733... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7522... Training loss: 1.5441... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7523... Training loss: 1.4190... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7524... Training loss: 1.3695... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7525... Training loss: 1.2992... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7526... Training loss: 1.3589... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7527... Training loss: 1.4365... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7528... Training loss: 1.3156... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7529... Training loss: 1.3911... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7530... Training loss: 1.4192... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7531... Training loss: 1.3979... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7532... Training loss: 1.4325... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7533... Training loss: 1.3709... 0.0809 sec/batch
Epoch: 8/20... Training Step: 7534... Training loss: 1.3823... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7535... Training loss: 1.3849... 0.0793 sec/batch
Epoch: 8/20... Training Step: 7536... Training loss: 1.3996... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7537... Training loss: 1.3157... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7538... Training loss: 1.3425... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7539... Training loss: 1.3057... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7540... Training loss: 1.4083... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7541... Training loss: 1.2618... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7542... Training loss: 1.2914... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7543... Training loss: 1.3443... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7544... Training loss: 1.3465... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7545... Training loss: 1.4050... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7546... Training loss: 1.3777... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7547... Training loss: 1.3654... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7548... Training loss: 1.4009... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7549... Training loss: 1.3313... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7550... Training loss: 1.3442... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7551... Training loss: 1.3954... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7552... Training loss: 1.4031... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7553... Training loss: 1.3776... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7554... Training loss: 1.4325... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7555... Training loss: 1.3121... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7556... Training loss: 1.4370... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7557... Training loss: 1.4170... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7558... Training loss: 1.3503... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7559... Training loss: 1.4263... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7560... Training loss: 1.3860... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7561... Training loss: 1.4028... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7562... Training loss: 1.4082... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7563... Training loss: 1.3124... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7564... Training loss: 1.3736... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7565... Training loss: 1.3138... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7566... Training loss: 1.2905... 0.0821 sec/batch
Epoch: 8/20... Training Step: 7567... Training loss: 1.3971... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7568... Training loss: 1.4183... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7569... Training loss: 1.3210... 0.0805 sec/batch
Epoch: 8/20... Training Step: 7570... Training loss: 1.3827... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7571... Training loss: 1.3484... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7572... Training loss: 1.3659... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7573... Training loss: 1.1997... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7574... Training loss: 1.3912... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7575... Training loss: 1.3984... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7576... Training loss: 1.3417... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7577... Training loss: 1.3095... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7578... Training loss: 1.3525... 0.0802 sec/batch
Epoch: 8/20... Training Step: 7579... Training loss: 1.4027... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7580... Training loss: 1.3558... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7581... Training loss: 1.4284... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7582... Training loss: 1.3674... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7583... Training loss: 1.4032... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7584... Training loss: 1.4517... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7585... Training loss: 1.4852... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7586... Training loss: 1.4305... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7587... Training loss: 1.3613... 0.0800 sec/batch
Epoch: 8/20... Training Step: 7588... Training loss: 1.3801... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7589... Training loss: 1.3329... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7590... Training loss: 1.4020... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7591... Training loss: 1.4132... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7592... Training loss: 1.2987... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7593... Training loss: 1.3147... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7594... Training loss: 1.3866... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7595... Training loss: 1.3561... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7596... Training loss: 1.4156... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7597... Training loss: 1.3682... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7598... Training loss: 1.4492... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7599... Training loss: 1.4584... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7600... Training loss: 1.4334... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7601... Training loss: 1.3119... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7602... Training loss: 1.3497... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7603... Training loss: 1.3661... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7604... Training loss: 1.4851... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7605... Training loss: 1.4018... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7606... Training loss: 1.4595... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7607... Training loss: 1.3829... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7608... Training loss: 1.4790... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7609... Training loss: 1.5010... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7610... Training loss: 1.3486... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7611... Training loss: 1.4866... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7612... Training loss: 1.4286... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7613... Training loss: 1.3787... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7614... Training loss: 1.3496... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7615... Training loss: 1.4188... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7616... Training loss: 1.4495... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7617... Training loss: 1.3736... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7618... Training loss: 1.4025... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7619... Training loss: 1.3339... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7620... Training loss: 1.3746... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7621... Training loss: 1.4561... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7622... Training loss: 1.3689... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7623... Training loss: 1.3382... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7624... Training loss: 1.4190... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7625... Training loss: 1.4333... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7626... Training loss: 1.4246... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7627... Training loss: 1.4212... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7628... Training loss: 1.3584... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7629... Training loss: 1.4588... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7630... Training loss: 1.4425... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7631... Training loss: 1.3799... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7632... Training loss: 1.3591... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7633... Training loss: 1.3314... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7634... Training loss: 1.4208... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7635... Training loss: 1.3616... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7636... Training loss: 1.4010... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7637... Training loss: 1.3622... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7638... Training loss: 1.3239... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7639... Training loss: 1.4101... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7640... Training loss: 1.3825... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7641... Training loss: 1.4597... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7642... Training loss: 1.4066... 0.0741 sec/batch
Epoch: 8/20... Training Step: 7643... Training loss: 1.3860... 0.0816 sec/batch
Epoch: 8/20... Training Step: 7644... Training loss: 1.3925... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7645... Training loss: 1.4061... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7646... Training loss: 1.3859... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7647... Training loss: 1.3797... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7648... Training loss: 1.3130... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7649... Training loss: 1.3070... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7650... Training loss: 1.3444... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7651... Training loss: 1.3939... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7652... Training loss: 1.3904... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7653... Training loss: 1.4168... 0.0844 sec/batch
Epoch: 8/20... Training Step: 7654... Training loss: 1.3662... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7655... Training loss: 1.3661... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7656... Training loss: 1.4493... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7657... Training loss: 1.3368... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7658... Training loss: 1.3755... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7659... Training loss: 1.4133... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7660... Training loss: 1.4179... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7661... Training loss: 1.4331... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7662... Training loss: 1.4229... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7663... Training loss: 1.4219... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7664... Training loss: 1.4893... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7665... Training loss: 1.4130... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7666... Training loss: 1.3807... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7667... Training loss: 1.3868... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7668... Training loss: 1.4224... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7669... Training loss: 1.4434... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7670... Training loss: 1.4651... 0.0787 sec/batch
Epoch: 8/20... Training Step: 7671... Training loss: 1.3679... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7672... Training loss: 1.4332... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7673... Training loss: 1.4267... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7674... Training loss: 1.3517... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7675... Training loss: 1.3710... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7676... Training loss: 1.4099... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7677... Training loss: 1.3590... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7678... Training loss: 1.4223... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7679... Training loss: 1.3796... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7680... Training loss: 1.3637... 0.0818 sec/batch
Epoch: 8/20... Training Step: 7681... Training loss: 1.3498... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7682... Training loss: 1.4490... 0.0795 sec/batch
Epoch: 8/20... Training Step: 7683... Training loss: 1.3856... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7684... Training loss: 1.3910... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7685... Training loss: 1.3439... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7686... Training loss: 1.3857... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7687... Training loss: 1.3595... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7688... Training loss: 1.4360... 0.0804 sec/batch
Epoch: 8/20... Training Step: 7689... Training loss: 1.4064... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7690... Training loss: 1.3624... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7691... Training loss: 1.3797... 0.0789 sec/batch
Epoch: 8/20... Training Step: 7692... Training loss: 1.3963... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7693... Training loss: 1.3188... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7694... Training loss: 1.3276... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7695... Training loss: 1.4824... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7696... Training loss: 1.3308... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7697... Training loss: 1.4104... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7698... Training loss: 1.4030... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7699... Training loss: 1.2961... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7700... Training loss: 1.4050... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7701... Training loss: 1.4242... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7702... Training loss: 1.4354... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7703... Training loss: 1.4411... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7704... Training loss: 1.4558... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7705... Training loss: 1.3829... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7706... Training loss: 1.3365... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7707... Training loss: 1.3962... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7708... Training loss: 1.3774... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7709... Training loss: 1.4209... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7710... Training loss: 1.4314... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7711... Training loss: 1.3780... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7712... Training loss: 1.3517... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7713... Training loss: 1.3498... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7714... Training loss: 1.3989... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7715... Training loss: 1.4114... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7716... Training loss: 1.4634... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7717... Training loss: 1.4281... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7718... Training loss: 1.4203... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7719... Training loss: 1.3888... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7720... Training loss: 1.2940... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7721... Training loss: 1.3393... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7722... Training loss: 1.3794... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7723... Training loss: 1.3596... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7724... Training loss: 1.3654... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7725... Training loss: 1.3930... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7726... Training loss: 1.4014... 0.0772 sec/batch
Epoch: 8/20... Training Step: 7727... Training loss: 1.4168... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7728... Training loss: 1.4103... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7729... Training loss: 1.4074... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7730... Training loss: 1.3762... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7731... Training loss: 1.4000... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7732... Training loss: 1.4082... 0.0737 sec/batch
Epoch: 8/20... Training Step: 7733... Training loss: 1.4064... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7734... Training loss: 1.4255... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7735... Training loss: 1.3724... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7736... Training loss: 1.3523... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7737... Training loss: 1.3986... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7738... Training loss: 1.3416... 0.0782 sec/batch
Epoch: 8/20... Training Step: 7739... Training loss: 1.2807... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7740... Training loss: 1.3685... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7741... Training loss: 1.3418... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7742... Training loss: 1.3551... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7743... Training loss: 1.3556... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7744... Training loss: 1.3590... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7745... Training loss: 1.4100... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7746... Training loss: 1.4253... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7747... Training loss: 1.4444... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7748... Training loss: 1.4040... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7749... Training loss: 1.3381... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7750... Training loss: 1.3747... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7751... Training loss: 1.3698... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7752... Training loss: 1.4300... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7753... Training loss: 1.4280... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7754... Training loss: 1.4405... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7755... Training loss: 1.3915... 0.0796 sec/batch
Epoch: 8/20... Training Step: 7756... Training loss: 1.4625... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7757... Training loss: 1.3861... 0.0780 sec/batch
Epoch: 8/20... Training Step: 7758... Training loss: 1.4466... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7759... Training loss: 1.4376... 0.0778 sec/batch
Epoch: 8/20... Training Step: 7760... Training loss: 1.4255... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7761... Training loss: 1.4538... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7762... Training loss: 1.5211... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7763... Training loss: 1.4234... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7764... Training loss: 1.3773... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7765... Training loss: 1.3830... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7766... Training loss: 1.3313... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7767... Training loss: 1.3913... 0.0790 sec/batch
Epoch: 8/20... Training Step: 7768... Training loss: 1.3840... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7769... Training loss: 1.3846... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7770... Training loss: 1.4223... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7771... Training loss: 1.4164... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7772... Training loss: 1.4477... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7773... Training loss: 1.4501... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7774... Training loss: 1.3722... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7775... Training loss: 1.4585... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7776... Training loss: 1.4136... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7777... Training loss: 1.4956... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7778... Training loss: 1.3923... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7779... Training loss: 1.3379... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7780... Training loss: 1.3771... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7781... Training loss: 1.4296... 0.0794 sec/batch
Epoch: 8/20... Training Step: 7782... Training loss: 1.4174... 0.0785 sec/batch
Epoch: 8/20... Training Step: 7783... Training loss: 1.4207... 0.0786 sec/batch
Epoch: 8/20... Training Step: 7784... Training loss: 1.3864... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7785... Training loss: 1.4221... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7786... Training loss: 1.5494... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7787... Training loss: 1.4364... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7788... Training loss: 1.4336... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7789... Training loss: 1.4735... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7790... Training loss: 1.3381... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7791... Training loss: 1.4047... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7792... Training loss: 1.3737... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7793... Training loss: 1.4040... 0.0791 sec/batch
Epoch: 8/20... Training Step: 7794... Training loss: 1.3982... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7795... Training loss: 1.3771... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7796... Training loss: 1.4521... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7797... Training loss: 1.4597... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7798... Training loss: 1.4075... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7799... Training loss: 1.4032... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7800... Training loss: 1.3912... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7801... Training loss: 1.4245... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7802... Training loss: 1.3909... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7803... Training loss: 1.4191... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7804... Training loss: 1.4454... 0.0837 sec/batch
Epoch: 8/20... Training Step: 7805... Training loss: 1.4073... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7806... Training loss: 1.3623... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7807... Training loss: 1.3434... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7808... Training loss: 1.3772... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7809... Training loss: 1.3475... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7810... Training loss: 1.4047... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7811... Training loss: 1.3985... 0.0793 sec/batch
Epoch: 8/20... Training Step: 7812... Training loss: 1.4660... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7813... Training loss: 1.4039... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7814... Training loss: 1.4641... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7815... Training loss: 1.4467... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7816... Training loss: 1.4245... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7817... Training loss: 1.4095... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7818... Training loss: 1.3980... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7819... Training loss: 1.3911... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7820... Training loss: 1.4263... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7821... Training loss: 1.4067... 0.0740 sec/batch
Epoch: 8/20... Training Step: 7822... Training loss: 1.4666... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7823... Training loss: 1.3914... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7824... Training loss: 1.3677... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7825... Training loss: 1.3933... 0.0837 sec/batch
Epoch: 8/20... Training Step: 7826... Training loss: 1.3341... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7827... Training loss: 1.4216... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7828... Training loss: 1.3969... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7829... Training loss: 1.4218... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7830... Training loss: 1.4489... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7831... Training loss: 1.4127... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7832... Training loss: 1.4598... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7833... Training loss: 1.4281... 0.0803 sec/batch
Epoch: 8/20... Training Step: 7834... Training loss: 1.4773... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7835... Training loss: 1.4358... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7836... Training loss: 1.3385... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7837... Training loss: 1.3974... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7838... Training loss: 1.3305... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7839... Training loss: 1.4323... 0.0788 sec/batch
Epoch: 8/20... Training Step: 7840... Training loss: 1.3478... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7841... Training loss: 1.4225... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7842... Training loss: 1.4311... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7843... Training loss: 1.3817... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7844... Training loss: 1.3371... 0.0743 sec/batch
Epoch: 8/20... Training Step: 7845... Training loss: 1.3610... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7846... Training loss: 1.3596... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7847... Training loss: 1.3505... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7848... Training loss: 1.3185... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7849... Training loss: 1.3255... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7850... Training loss: 1.4023... 0.0790 sec/batch
Epoch: 8/20... Training Step: 7851... Training loss: 1.3264... 0.0798 sec/batch
Epoch: 8/20... Training Step: 7852... Training loss: 1.4053... 0.0789 sec/batch
Epoch: 8/20... Training Step: 7853... Training loss: 1.3329... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7854... Training loss: 1.2903... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7855... Training loss: 1.3125... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7856... Training loss: 1.3912... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7857... Training loss: 1.4882... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7858... Training loss: 1.3984... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7859... Training loss: 1.3755... 0.0748 sec/batch
Epoch: 8/20... Training Step: 7860... Training loss: 1.3653... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7861... Training loss: 1.3908... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7862... Training loss: 1.4309... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7863... Training loss: 1.3758... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7864... Training loss: 1.4209... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7865... Training loss: 1.4829... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7866... Training loss: 1.4432... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7867... Training loss: 1.3976... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7868... Training loss: 1.3743... 0.0745 sec/batch
Epoch: 8/20... Training Step: 7869... Training loss: 1.4100... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7870... Training loss: 1.4408... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7871... Training loss: 1.3929... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7872... Training loss: 1.3952... 0.0764 sec/batch
Epoch: 8/20... Training Step: 7873... Training loss: 1.4968... 0.0758 sec/batch
Epoch: 8/20... Training Step: 7874... Training loss: 1.4439... 0.0781 sec/batch
Epoch: 8/20... Training Step: 7875... Training loss: 1.4247... 0.0777 sec/batch
Epoch: 8/20... Training Step: 7876... Training loss: 1.4890... 0.0792 sec/batch
Epoch: 8/20... Training Step: 7877... Training loss: 1.3941... 0.0845 sec/batch
Epoch: 8/20... Training Step: 7878... Training loss: 1.4021... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7879... Training loss: 1.4405... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7880... Training loss: 1.4193... 0.0740 sec/batch
Epoch: 8/20... Training Step: 7881... Training loss: 1.3737... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7882... Training loss: 1.3985... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7883... Training loss: 1.3836... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7884... Training loss: 1.4819... 0.0784 sec/batch
Epoch: 8/20... Training Step: 7885... Training loss: 1.4025... 0.0763 sec/batch
Epoch: 8/20... Training Step: 7886... Training loss: 1.4374... 0.0760 sec/batch
Epoch: 8/20... Training Step: 7887... Training loss: 1.4824... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7888... Training loss: 1.4221... 0.0752 sec/batch
Epoch: 8/20... Training Step: 7889... Training loss: 1.4656... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7890... Training loss: 1.3964... 0.0779 sec/batch
Epoch: 8/20... Training Step: 7891... Training loss: 1.3825... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7892... Training loss: 1.3516... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7893... Training loss: 1.4163... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7894... Training loss: 1.2888... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7895... Training loss: 1.4024... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7896... Training loss: 1.3364... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7897... Training loss: 1.3814... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7898... Training loss: 1.4102... 0.0741 sec/batch
Epoch: 8/20... Training Step: 7899... Training loss: 1.4379... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7900... Training loss: 1.4044... 0.0847 sec/batch
Epoch: 8/20... Training Step: 7901... Training loss: 1.3952... 0.0757 sec/batch
Epoch: 8/20... Training Step: 7902... Training loss: 1.4334... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7903... Training loss: 1.3508... 0.0767 sec/batch
Epoch: 8/20... Training Step: 7904... Training loss: 1.3589... 0.0747 sec/batch
Epoch: 8/20... Training Step: 7905... Training loss: 1.4048... 0.0750 sec/batch
Epoch: 8/20... Training Step: 7906... Training loss: 1.4015... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7907... Training loss: 1.4190... 0.0756 sec/batch
Epoch: 8/20... Training Step: 7908... Training loss: 1.3841... 0.0776 sec/batch
Epoch: 8/20... Training Step: 7909... Training loss: 1.4497... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7910... Training loss: 1.4703... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7911... Training loss: 1.4426... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7912... Training loss: 1.4495... 0.0768 sec/batch
Epoch: 8/20... Training Step: 7913... Training loss: 1.4352... 0.0744 sec/batch
Epoch: 8/20... Training Step: 7914... Training loss: 1.5879... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7915... Training loss: 1.3738... 0.0766 sec/batch
Epoch: 8/20... Training Step: 7916... Training loss: 1.4955... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7917... Training loss: 1.4945... 0.0742 sec/batch
Epoch: 8/20... Training Step: 7918... Training loss: 1.3731... 0.0754 sec/batch
Epoch: 8/20... Training Step: 7919... Training loss: 1.4397... 0.0765 sec/batch
Epoch: 8/20... Training Step: 7920... Training loss: 1.4588... 0.0773 sec/batch
Epoch: 8/20... Training Step: 7921... Training loss: 1.4823... 0.0774 sec/batch
Epoch: 8/20... Training Step: 7922... Training loss: 1.4504... 0.0749 sec/batch
Epoch: 8/20... Training Step: 7923... Training loss: 1.4435... 0.0746 sec/batch
Epoch: 8/20... Training Step: 7924... Training loss: 1.3751... 0.0771 sec/batch
Epoch: 8/20... Training Step: 7925... Training loss: 1.3205... 0.0753 sec/batch
Epoch: 8/20... Training Step: 7926... Training loss: 1.4965... 0.0775 sec/batch
Epoch: 8/20... Training Step: 7927... Training loss: 1.4307... 0.0751 sec/batch
Epoch: 8/20... Training Step: 7928... Training loss: 1.4006... 0.0755 sec/batch
Epoch: 8/20... Training Step: 7929... Training loss: 1.4171... 0.0769 sec/batch
Epoch: 8/20... Training Step: 7930... Training loss: 1.3796... 0.0783 sec/batch
Epoch: 8/20... Training Step: 7931... Training loss: 1.3751... 0.0762 sec/batch
Epoch: 8/20... Training Step: 7932... Training loss: 1.3700... 0.0759 sec/batch
Epoch: 8/20... Training Step: 7933... Training loss: 1.3936... 0.0761 sec/batch
Epoch: 8/20... Training Step: 7934... Training loss: 1.2786... 0.0792 sec/batch
Epoch: 8/20... Training Step: 7935... Training loss: 1.3699... 0.0770 sec/batch
Epoch: 8/20... Training Step: 7936... Training loss: 1.4331... 0.0794 sec/batch
Epoch: 9/20... Training Step: 7937... Training loss: 1.4676... 0.0778 sec/batch
Epoch: 9/20... Training Step: 7938... Training loss: 1.3675... 0.0754 sec/batch
Epoch: 9/20... Training Step: 7939... Training loss: 1.4477... 0.0773 sec/batch
Epoch: 9/20... Training Step: 7940... Training loss: 1.3982... 0.0766 sec/batch
Epoch: 9/20... Training Step: 7941... Training loss: 1.3152... 0.0767 sec/batch
Epoch: 9/20... Training Step: 7942... Training loss: 1.2642... 0.0767 sec/batch
Epoch: 9/20... Training Step: 7943... Training loss: 1.3827... 0.0765 sec/batch
Epoch: 9/20... Training Step: 7944... Training loss: 1.3892... 0.0768 sec/batch
Epoch: 9/20... Training Step: 7945... Training loss: 1.3979... 0.0765 sec/batch
Epoch: 9/20... Training Step: 7946... Training loss: 1.3069... 0.0766 sec/batch
Epoch: 9/20... Training Step: 7947... Training loss: 1.4439... 0.0755 sec/batch
Epoch: 9/20... Training Step: 7948... Training loss: 1.3452... 0.0757 sec/batch
Epoch: 9/20... Training Step: 7949... Training loss: 1.3884... 0.0765 sec/batch
Epoch: 9/20... Training Step: 7950... Training loss: 1.3773... 0.0755 sec/batch
Epoch: 9/20... Training Step: 7951... Training loss: 1.4280... 0.0758 sec/batch
Epoch: 9/20... Training Step: 7952... Training loss: 1.3229... 0.0764 sec/batch
Epoch: 9/20... Training Step: 7953... Training loss: 1.3711... 0.0745 sec/batch
Epoch: 9/20... Training Step: 7954... Training loss: 1.3977... 0.0780 sec/batch
Epoch: 9/20... Training Step: 7955... Training loss: 1.3665... 0.0756 sec/batch
Epoch: 9/20... Training Step: 7956... Training loss: 1.3483... 0.0816 sec/batch
Epoch: 9/20... Training Step: 7957... Training loss: 1.3820... 0.0746 sec/batch
Epoch: 9/20... Training Step: 7958... Training loss: 1.3548... 0.0758 sec/batch
Epoch: 9/20... Training Step: 7959... Training loss: 1.3461... 0.0764 sec/batch
Epoch: 9/20... Training Step: 7960... Training loss: 1.3121... 0.0760 sec/batch
Epoch: 9/20... Training Step: 7961... Training loss: 1.4036... 0.0744 sec/batch
Epoch: 9/20... Training Step: 7962... Training loss: 1.3265... 0.0758 sec/batch
Epoch: 9/20... Training Step: 7963... Training loss: 1.3619... 0.0755 sec/batch
Epoch: 9/20... Training Step: 7964... Training loss: 1.3192... 0.0750 sec/batch
Epoch: 9/20... Training Step: 7965... Training loss: 1.3576... 0.0779 sec/batch
Epoch: 9/20... Training Step: 7966... Training loss: 1.4513... 0.0755 sec/batch
Epoch: 9/20... Training Step: 7967... Training loss: 1.4164... 0.0749 sec/batch
Epoch: 9/20... Training Step: 7968... Training loss: 1.4271... 0.0746 sec/batch
Epoch: 9/20... Training Step: 7969... Training loss: 1.3681... 0.0782 sec/batch
Epoch: 9/20... Training Step: 7970... Training loss: 1.4224... 0.0757 sec/batch
Epoch: 9/20... Training Step: 7971... Training loss: 1.4906... 0.0754 sec/batch
Epoch: 9/20... Training Step: 7972... Training loss: 1.4604... 0.0752 sec/batch
Epoch: 9/20... Training Step: 7973... Training loss: 1.3915... 0.0778 sec/batch
Epoch: 9/20... Training Step: 7974... Training loss: 1.4275... 0.0774 sec/batch
Epoch: 9/20... Training Step: 7975... Training loss: 1.3847... 0.0759 sec/batch
Epoch: 9/20... Training Step: 7976... Training loss: 1.3651... 0.0778 sec/batch
Epoch: 9/20... Training Step: 7977... Training loss: 1.3853... 0.0771 sec/batch
Epoch: 9/20... Training Step: 7978... Training loss: 1.3599... 0.0781 sec/batch
Epoch: 9/20... Training Step: 7979... Training loss: 1.3515... 0.0759 sec/batch
Epoch: 9/20... Training Step: 7980... Training loss: 1.3481... 0.0777 sec/batch
Epoch: 9/20... Training Step: 7981... Training loss: 1.3978... 0.0758 sec/batch
Epoch: 9/20... Training Step: 7982... Training loss: 1.3609... 0.0796 sec/batch
Epoch: 9/20... Training Step: 7983... Training loss: 1.4231... 0.0747 sec/batch
Epoch: 9/20... Training Step: 7984... Training loss: 1.3518... 0.0763 sec/batch
Epoch: 9/20... Training Step: 7985... Training loss: 1.4084... 0.0760 sec/batch
Epoch: 9/20... Training Step: 7986... Training loss: 1.3921... 0.0740 sec/batch
Epoch: 9/20... Training Step: 7987... Training loss: 1.2969... 0.0768 sec/batch
Epoch: 9/20... Training Step: 7988... Training loss: 1.4245... 0.0743 sec/batch
Epoch: 9/20... Training Step: 7989... Training loss: 1.3061... 0.0756 sec/batch
Epoch: 9/20... Training Step: 7990... Training loss: 1.3407... 0.0754 sec/batch
Epoch: 9/20... Training Step: 7991... Training loss: 1.3597... 0.0765 sec/batch
Epoch: 9/20... Training Step: 7992... Training loss: 1.3338... 0.0767 sec/batch
Epoch: 9/20... Training Step: 7993... Training loss: 1.3874... 0.0747 sec/batch
Epoch: 9/20... Training Step: 7994... Training loss: 1.4260... 0.0771 sec/batch
Epoch: 9/20... Training Step: 7995... Training loss: 1.3655... 0.0742 sec/batch
Epoch: 9/20... Training Step: 7996... Training loss: 1.4116... 0.0775 sec/batch
Epoch: 9/20... Training Step: 7997... Training loss: 1.2755... 0.0773 sec/batch
Epoch: 9/20... Training Step: 7998... Training loss: 1.3119... 0.0760 sec/batch
Epoch: 9/20... Training Step: 7999... Training loss: 1.3668... 0.0795 sec/batch
Epoch: 9/20... Training Step: 8000... Training loss: 1.3165... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8001... Training loss: 1.4012... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8002... Training loss: 1.3529... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8003... Training loss: 1.3581... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8004... Training loss: 1.3632... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8005... Training loss: 1.3986... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8006... Training loss: 1.3845... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8007... Training loss: 1.4075... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8008... Training loss: 1.4413... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8009... Training loss: 1.3959... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8010... Training loss: 1.3349... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8011... Training loss: 1.3644... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8012... Training loss: 1.3240... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8013... Training loss: 1.3546... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8014... Training loss: 1.3644... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8015... Training loss: 1.3752... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8016... Training loss: 1.4367... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8017... Training loss: 1.3974... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8018... Training loss: 1.3516... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8019... Training loss: 1.3884... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8020... Training loss: 1.3820... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8021... Training loss: 1.3498... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8022... Training loss: 1.4339... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8023... Training loss: 1.3808... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8024... Training loss: 1.3847... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8025... Training loss: 1.4002... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8026... Training loss: 1.4384... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8027... Training loss: 1.4560... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8028... Training loss: 1.3784... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8029... Training loss: 1.3317... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8030... Training loss: 1.3096... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8031... Training loss: 1.3178... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8032... Training loss: 1.4014... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8033... Training loss: 1.3518... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8034... Training loss: 1.3215... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8035... Training loss: 1.3469... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8036... Training loss: 1.3344... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8037... Training loss: 1.3408... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8038... Training loss: 1.3412... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8039... Training loss: 1.4027... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8040... Training loss: 1.4070... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8041... Training loss: 1.3921... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8042... Training loss: 1.3818... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8043... Training loss: 1.3363... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8044... Training loss: 1.4253... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8045... Training loss: 1.3620... 0.0742 sec/batch
Epoch: 9/20... Training Step: 8046... Training loss: 1.4190... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8047... Training loss: 1.3894... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8048... Training loss: 1.3616... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8049... Training loss: 1.4612... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8050... Training loss: 1.4003... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8051... Training loss: 1.3222... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8052... Training loss: 1.4970... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8053... Training loss: 1.4267... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8054... Training loss: 1.4771... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8055... Training loss: 1.4019... 0.0789 sec/batch
Epoch: 9/20... Training Step: 8056... Training loss: 1.3944... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8057... Training loss: 1.4864... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8058... Training loss: 1.3339... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8059... Training loss: 1.4047... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8060... Training loss: 1.3806... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8061... Training loss: 1.4501... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8062... Training loss: 1.4445... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8063... Training loss: 1.4331... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8064... Training loss: 1.3997... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8065... Training loss: 1.3215... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8066... Training loss: 1.4015... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8067... Training loss: 1.3458... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8068... Training loss: 1.4347... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8069... Training loss: 1.4024... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8070... Training loss: 1.3837... 0.0821 sec/batch
Epoch: 9/20... Training Step: 8071... Training loss: 1.4174... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8072... Training loss: 1.4051... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8073... Training loss: 1.4238... 0.0836 sec/batch
Epoch: 9/20... Training Step: 8074... Training loss: 1.4642... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8075... Training loss: 1.4354... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8076... Training loss: 1.3889... 0.0790 sec/batch
Epoch: 9/20... Training Step: 8077... Training loss: 1.4006... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8078... Training loss: 1.3698... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8079... Training loss: 1.3714... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8080... Training loss: 1.3678... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8081... Training loss: 1.2562... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8082... Training loss: 1.4245... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8083... Training loss: 1.3818... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8084... Training loss: 1.3091... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8085... Training loss: 1.3527... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8086... Training loss: 1.3260... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8087... Training loss: 1.2566... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8088... Training loss: 1.3667... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8089... Training loss: 1.3279... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8090... Training loss: 1.3741... 0.0738 sec/batch
Epoch: 9/20... Training Step: 8091... Training loss: 1.3509... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8092... Training loss: 1.3355... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8093... Training loss: 1.3758... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8094... Training loss: 1.3624... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8095... Training loss: 1.3844... 0.0736 sec/batch
Epoch: 9/20... Training Step: 8096... Training loss: 1.4347... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8097... Training loss: 1.3602... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8098... Training loss: 1.3512... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8099... Training loss: 1.3663... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8100... Training loss: 1.4277... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8101... Training loss: 1.3932... 0.0739 sec/batch
Epoch: 9/20... Training Step: 8102... Training loss: 1.4565... 0.0789 sec/batch
Epoch: 9/20... Training Step: 8103... Training loss: 1.4297... 0.0797 sec/batch
Epoch: 9/20... Training Step: 8104... Training loss: 1.3877... 0.0737 sec/batch
Epoch: 9/20... Training Step: 8105... Training loss: 1.4204... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8106... Training loss: 1.3639... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8107... Training loss: 1.4136... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8108... Training loss: 1.3618... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8109... Training loss: 1.3678... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8110... Training loss: 1.4357... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8111... Training loss: 1.3687... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8112... Training loss: 1.4077... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8113... Training loss: 1.4122... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8114... Training loss: 1.4746... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8115... Training loss: 1.4374... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8116... Training loss: 1.3512... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8117... Training loss: 1.3952... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8118... Training loss: 1.4582... 0.0805 sec/batch
Epoch: 9/20... Training Step: 8119... Training loss: 1.3829... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8120... Training loss: 1.4108... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8121... Training loss: 1.4071... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8122... Training loss: 1.4333... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8123... Training loss: 1.4132... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8124... Training loss: 1.3845... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8125... Training loss: 1.3846... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8126... Training loss: 1.4639... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8127... Training loss: 1.4766... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8128... Training loss: 1.4100... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8129... Training loss: 1.4203... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8130... Training loss: 1.4941... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8131... Training loss: 1.3753... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8132... Training loss: 1.3552... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8133... Training loss: 1.4421... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8134... Training loss: 1.3913... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8135... Training loss: 1.4192... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8136... Training loss: 1.3832... 0.0848 sec/batch
Epoch: 9/20... Training Step: 8137... Training loss: 1.3756... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8138... Training loss: 1.3952... 0.0791 sec/batch
Epoch: 9/20... Training Step: 8139... Training loss: 1.3962... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8140... Training loss: 1.3543... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8141... Training loss: 1.3418... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8142... Training loss: 1.4387... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8143... Training loss: 1.3705... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8144... Training loss: 1.3719... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8145... Training loss: 1.4239... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8146... Training loss: 1.4049... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8147... Training loss: 1.3451... 0.0740 sec/batch
Epoch: 9/20... Training Step: 8148... Training loss: 1.4333... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8149... Training loss: 1.3568... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8150... Training loss: 1.4218... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8151... Training loss: 1.3546... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8152... Training loss: 1.3908... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8153... Training loss: 1.3527... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8154... Training loss: 1.4105... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8155... Training loss: 1.3821... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8156... Training loss: 1.3330... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8157... Training loss: 1.3412... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8158... Training loss: 1.5037... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8159... Training loss: 1.3973... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8160... Training loss: 1.3918... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8161... Training loss: 1.4273... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8162... Training loss: 1.3378... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8163... Training loss: 1.4048... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8164... Training loss: 1.4156... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8165... Training loss: 1.3059... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8166... Training loss: 1.3518... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8167... Training loss: 1.3594... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8168... Training loss: 1.3939... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8169... Training loss: 1.3711... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8170... Training loss: 1.4581... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8171... Training loss: 1.3536... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8172... Training loss: 1.3842... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8173... Training loss: 1.3543... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8174... Training loss: 1.4202... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8175... Training loss: 1.4469... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8176... Training loss: 1.3774... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8177... Training loss: 1.3675... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8178... Training loss: 1.3785... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8179... Training loss: 1.4257... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8180... Training loss: 1.4188... 0.0794 sec/batch
Epoch: 9/20... Training Step: 8181... Training loss: 1.3744... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8182... Training loss: 1.4198... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8183... Training loss: 1.3359... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8184... Training loss: 1.3169... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8185... Training loss: 1.4059... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8186... Training loss: 1.3830... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8187... Training loss: 1.3529... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8188... Training loss: 1.3482... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8189... Training loss: 1.3339... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8190... Training loss: 1.3314... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8191... Training loss: 1.3245... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8192... Training loss: 1.3662... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8193... Training loss: 1.4648... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8194... Training loss: 1.3607... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8195... Training loss: 1.3626... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8196... Training loss: 1.3914... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8197... Training loss: 1.4447... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8198... Training loss: 1.4346... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8199... Training loss: 1.4161... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8200... Training loss: 1.4112... 0.0844 sec/batch
Epoch: 9/20... Training Step: 8201... Training loss: 1.4174... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8202... Training loss: 1.3065... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8203... Training loss: 1.3498... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8204... Training loss: 1.3858... 0.0844 sec/batch
Epoch: 9/20... Training Step: 8205... Training loss: 1.3118... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8206... Training loss: 1.3456... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8207... Training loss: 1.3727... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8208... Training loss: 1.3526... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8209... Training loss: 1.2889... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8210... Training loss: 1.3109... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8211... Training loss: 1.3485... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8212... Training loss: 1.3551... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8213... Training loss: 1.3781... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8214... Training loss: 1.2732... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8215... Training loss: 1.3283... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8216... Training loss: 1.4372... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8217... Training loss: 1.3695... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8218... Training loss: 1.3347... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8219... Training loss: 1.3999... 0.0812 sec/batch
Epoch: 9/20... Training Step: 8220... Training loss: 1.3309... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8221... Training loss: 1.2797... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8222... Training loss: 1.3385... 0.0787 sec/batch
Epoch: 9/20... Training Step: 8223... Training loss: 1.3982... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8224... Training loss: 1.3419... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8225... Training loss: 1.3780... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8226... Training loss: 1.3298... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8227... Training loss: 1.3347... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8228... Training loss: 1.3605... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8229... Training loss: 1.2949... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8230... Training loss: 1.3352... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8231... Training loss: 1.3716... 0.0792 sec/batch
Epoch: 9/20... Training Step: 8232... Training loss: 1.4004... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8233... Training loss: 1.3623... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8234... Training loss: 1.4541... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8235... Training loss: 1.3674... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8236... Training loss: 1.4064... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8237... Training loss: 1.4303... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8238... Training loss: 1.3330... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8239... Training loss: 1.3190... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8240... Training loss: 1.3979... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8241... Training loss: 1.3718... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8242... Training loss: 1.4229... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8243... Training loss: 1.3345... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8244... Training loss: 1.3125... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8245... Training loss: 1.4507... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8246... Training loss: 1.3554... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8247... Training loss: 1.4336... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8248... Training loss: 1.3678... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8249... Training loss: 1.3561... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8250... Training loss: 1.3627... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8251... Training loss: 1.3745... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8252... Training loss: 1.3309... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8253... Training loss: 1.3239... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8254... Training loss: 1.3241... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8255... Training loss: 1.3663... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8256... Training loss: 1.4009... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8257... Training loss: 1.3866... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8258... Training loss: 1.2848... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8259... Training loss: 1.3228... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8260... Training loss: 1.3345... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8261... Training loss: 1.4246... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8262... Training loss: 1.3420... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8263... Training loss: 1.3770... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8264... Training loss: 1.3556... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8265... Training loss: 1.3840... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8266... Training loss: 1.3948... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8267... Training loss: 1.4092... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8268... Training loss: 1.3198... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8269... Training loss: 1.4830... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8270... Training loss: 1.4522... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8271... Training loss: 1.4413... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8272... Training loss: 1.4132... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8273... Training loss: 1.3652... 0.0825 sec/batch
Epoch: 9/20... Training Step: 8274... Training loss: 1.3641... 0.0798 sec/batch
Epoch: 9/20... Training Step: 8275... Training loss: 1.3709... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8276... Training loss: 1.3626... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8277... Training loss: 1.3693... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8278... Training loss: 1.3439... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8279... Training loss: 1.3117... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8280... Training loss: 1.3830... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8281... Training loss: 1.3609... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8282... Training loss: 1.3771... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8283... Training loss: 1.4050... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8284... Training loss: 1.4295... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8285... Training loss: 1.4147... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8286... Training loss: 1.4166... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8287... Training loss: 1.3855... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8288... Training loss: 1.3455... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8289... Training loss: 1.3561... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8290... Training loss: 1.3560... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8291... Training loss: 1.3903... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8292... Training loss: 1.3904... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8293... Training loss: 1.3469... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8294... Training loss: 1.3666... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8295... Training loss: 1.3823... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8296... Training loss: 1.3830... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8297... Training loss: 1.4046... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8298... Training loss: 1.4612... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8299... Training loss: 1.3702... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8300... Training loss: 1.3845... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8301... Training loss: 1.3700... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8302... Training loss: 1.3397... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8303... Training loss: 1.3579... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8304... Training loss: 1.2983... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8305... Training loss: 1.3165... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8306... Training loss: 1.3170... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8307... Training loss: 1.4440... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8308... Training loss: 1.3032... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8309... Training loss: 1.3625... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8310... Training loss: 1.2937... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8311... Training loss: 1.3775... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8312... Training loss: 1.3103... 0.0803 sec/batch
Epoch: 9/20... Training Step: 8313... Training loss: 1.3516... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8314... Training loss: 1.3761... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8315... Training loss: 1.3324... 0.0794 sec/batch
Epoch: 9/20... Training Step: 8316... Training loss: 1.3778... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8317... Training loss: 1.3740... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8318... Training loss: 1.3474... 0.0790 sec/batch
Epoch: 9/20... Training Step: 8319... Training loss: 1.4144... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8320... Training loss: 1.4076... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8321... Training loss: 1.4136... 0.0851 sec/batch
Epoch: 9/20... Training Step: 8322... Training loss: 1.3718... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8323... Training loss: 1.4116... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8324... Training loss: 1.3971... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8325... Training loss: 1.4334... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8326... Training loss: 1.3379... 0.0800 sec/batch
Epoch: 9/20... Training Step: 8327... Training loss: 1.3901... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8328... Training loss: 1.3538... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8329... Training loss: 1.3986... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8330... Training loss: 1.3512... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8331... Training loss: 1.3650... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8332... Training loss: 1.3486... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8333... Training loss: 1.3436... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8334... Training loss: 1.4184... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8335... Training loss: 1.3821... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8336... Training loss: 1.3486... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8337... Training loss: 1.3190... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8338... Training loss: 1.3902... 0.0806 sec/batch
Epoch: 9/20... Training Step: 8339... Training loss: 1.3132... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8340... Training loss: 1.4230... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8341... Training loss: 1.3534... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8342... Training loss: 1.3812... 0.0818 sec/batch
Epoch: 9/20... Training Step: 8343... Training loss: 1.3051... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8344... Training loss: 1.4450... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8345... Training loss: 1.3356... 0.0738 sec/batch
Epoch: 9/20... Training Step: 8346... Training loss: 1.3034... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8347... Training loss: 1.2860... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8348... Training loss: 1.3429... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8349... Training loss: 1.3766... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8350... Training loss: 1.3072... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8351... Training loss: 1.3275... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8352... Training loss: 1.4028... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8353... Training loss: 1.2976... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8354... Training loss: 1.3595... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8355... Training loss: 1.3928... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8356... Training loss: 1.4497... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8357... Training loss: 1.2933... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8358... Training loss: 1.3956... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8359... Training loss: 1.3670... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8360... Training loss: 1.3871... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8361... Training loss: 1.3664... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8362... Training loss: 1.3746... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8363... Training loss: 1.4391... 0.0794 sec/batch
Epoch: 9/20... Training Step: 8364... Training loss: 1.4669... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8365... Training loss: 1.4340... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8366... Training loss: 1.4788... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8367... Training loss: 1.3749... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8368... Training loss: 1.4057... 0.0742 sec/batch
Epoch: 9/20... Training Step: 8369... Training loss: 1.4321... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8370... Training loss: 1.3877... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8371... Training loss: 1.4245... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8372... Training loss: 1.3607... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8373... Training loss: 1.4015... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8374... Training loss: 1.5000... 0.0740 sec/batch
Epoch: 9/20... Training Step: 8375... Training loss: 1.3953... 0.0791 sec/batch
Epoch: 9/20... Training Step: 8376... Training loss: 1.4103... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8377... Training loss: 1.3661... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8378... Training loss: 1.4007... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8379... Training loss: 1.4029... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8380... Training loss: 1.4044... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8381... Training loss: 1.3836... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8382... Training loss: 1.4732... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8383... Training loss: 1.3742... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8384... Training loss: 1.3795... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8385... Training loss: 1.4705... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8386... Training loss: 1.4718... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8387... Training loss: 1.3741... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8388... Training loss: 1.3726... 0.0919 sec/batch
Epoch: 9/20... Training Step: 8389... Training loss: 1.4038... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8390... Training loss: 1.4148... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8391... Training loss: 1.4165... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8392... Training loss: 1.3648... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8393... Training loss: 1.3420... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8394... Training loss: 1.3543... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8395... Training loss: 1.3611... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8396... Training loss: 1.4058... 0.0791 sec/batch
Epoch: 9/20... Training Step: 8397... Training loss: 1.4139... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8398... Training loss: 1.3772... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8399... Training loss: 1.3424... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8400... Training loss: 1.3870... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8401... Training loss: 1.3270... 0.0742 sec/batch
Epoch: 9/20... Training Step: 8402... Training loss: 1.3486... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8403... Training loss: 1.2866... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8404... Training loss: 1.3767... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8405... Training loss: 1.3543... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8406... Training loss: 1.4492... 0.0852 sec/batch
Epoch: 9/20... Training Step: 8407... Training loss: 1.3944... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8408... Training loss: 1.3719... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8409... Training loss: 1.3644... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8410... Training loss: 1.3780... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8411... Training loss: 1.2511... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8412... Training loss: 1.3504... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8413... Training loss: 1.3735... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8414... Training loss: 1.3814... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8415... Training loss: 1.3019... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8416... Training loss: 1.3241... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8417... Training loss: 1.3137... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8418... Training loss: 1.3717... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8419... Training loss: 1.3622... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8420... Training loss: 1.3848... 0.0800 sec/batch
Epoch: 9/20... Training Step: 8421... Training loss: 1.3159... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8422... Training loss: 1.3310... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8423... Training loss: 1.3990... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8424... Training loss: 1.3420... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8425... Training loss: 1.3389... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8426... Training loss: 1.3251... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8427... Training loss: 1.3278... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8428... Training loss: 1.3179... 0.0789 sec/batch
Epoch: 9/20... Training Step: 8429... Training loss: 1.3711... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8430... Training loss: 1.3460... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8431... Training loss: 1.3193... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8432... Training loss: 1.3245... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8433... Training loss: 1.3630... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8434... Training loss: 1.3631... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8435... Training loss: 1.3438... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8436... Training loss: 1.3565... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8437... Training loss: 1.3344... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8438... Training loss: 1.3639... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8439... Training loss: 1.3409... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8440... Training loss: 1.3590... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8441... Training loss: 1.3633... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8442... Training loss: 1.3161... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8443... Training loss: 1.3443... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8444... Training loss: 1.3252... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8445... Training loss: 1.3215... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8446... Training loss: 1.3398... 0.0797 sec/batch
Epoch: 9/20... Training Step: 8447... Training loss: 1.3996... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8448... Training loss: 1.3617... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8449... Training loss: 1.4268... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8450... Training loss: 1.4100... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8451... Training loss: 1.3689... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8452... Training loss: 1.3857... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8453... Training loss: 1.3884... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8454... Training loss: 1.3845... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8455... Training loss: 1.3653... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8456... Training loss: 1.3085... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8457... Training loss: 1.3755... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8458... Training loss: 1.3329... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8459... Training loss: 1.4460... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8460... Training loss: 1.3390... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8461... Training loss: 1.3586... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8462... Training loss: 1.3523... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8463... Training loss: 1.3473... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8464... Training loss: 1.3303... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8465... Training loss: 1.3631... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8466... Training loss: 1.4347... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8467... Training loss: 1.4953... 0.0827 sec/batch
Epoch: 9/20... Training Step: 8468... Training loss: 1.3836... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8469... Training loss: 1.4490... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8470... Training loss: 1.4194... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8471... Training loss: 1.3913... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8472... Training loss: 1.3617... 0.0788 sec/batch
Epoch: 9/20... Training Step: 8473... Training loss: 1.3363... 0.0844 sec/batch
Epoch: 9/20... Training Step: 8474... Training loss: 1.3322... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8475... Training loss: 1.3400... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8476... Training loss: 1.3444... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8477... Training loss: 1.4526... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8478... Training loss: 1.3414... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8479... Training loss: 1.3455... 0.0787 sec/batch
Epoch: 9/20... Training Step: 8480... Training loss: 1.3509... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8481... Training loss: 1.3127... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8482... Training loss: 1.3303... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8483... Training loss: 1.4213... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8484... Training loss: 1.3934... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8485... Training loss: 1.3158... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8486... Training loss: 1.3430... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8487... Training loss: 1.3581... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8488... Training loss: 1.2945... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8489... Training loss: 1.3100... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8490... Training loss: 1.3355... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8491... Training loss: 1.3969... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8492... Training loss: 1.3400... 0.0739 sec/batch
Epoch: 9/20... Training Step: 8493... Training loss: 1.3295... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8494... Training loss: 1.3512... 0.0838 sec/batch
Epoch: 9/20... Training Step: 8495... Training loss: 1.3659... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8496... Training loss: 1.3060... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8497... Training loss: 1.3256... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8498... Training loss: 1.3572... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8499... Training loss: 1.3098... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8500... Training loss: 1.3422... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8501... Training loss: 1.3794... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8502... Training loss: 1.3923... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8503... Training loss: 1.3467... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8504... Training loss: 1.2925... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8505... Training loss: 1.2848... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8506... Training loss: 1.2947... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8507... Training loss: 1.3167... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8508... Training loss: 1.3340... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8509... Training loss: 1.4193... 0.0790 sec/batch
Epoch: 9/20... Training Step: 8510... Training loss: 1.3848... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8511... Training loss: 1.3885... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8512... Training loss: 1.3683... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8513... Training loss: 1.3468... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8514... Training loss: 1.5012... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8515... Training loss: 1.4081... 0.0841 sec/batch
Epoch: 9/20... Training Step: 8516... Training loss: 1.3479... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8517... Training loss: 1.2880... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8518... Training loss: 1.3234... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8519... Training loss: 1.4178... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8520... Training loss: 1.3175... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8521... Training loss: 1.3369... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8522... Training loss: 1.3682... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8523... Training loss: 1.3810... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8524... Training loss: 1.4151... 0.0817 sec/batch
Epoch: 9/20... Training Step: 8525... Training loss: 1.3419... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8526... Training loss: 1.3298... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8527... Training loss: 1.3591... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8528... Training loss: 1.3787... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8529... Training loss: 1.3141... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8530... Training loss: 1.2951... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8531... Training loss: 1.2907... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8532... Training loss: 1.3684... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8533... Training loss: 1.2379... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8534... Training loss: 1.2738... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8535... Training loss: 1.3315... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8536... Training loss: 1.3226... 0.0789 sec/batch
Epoch: 9/20... Training Step: 8537... Training loss: 1.4129... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8538... Training loss: 1.3445... 0.0797 sec/batch
Epoch: 9/20... Training Step: 8539... Training loss: 1.3220... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8540... Training loss: 1.3493... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8541... Training loss: 1.3216... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8542... Training loss: 1.3210... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8543... Training loss: 1.3856... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8544... Training loss: 1.3847... 0.0742 sec/batch
Epoch: 9/20... Training Step: 8545... Training loss: 1.3619... 0.0830 sec/batch
Epoch: 9/20... Training Step: 8546... Training loss: 1.4246... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8547... Training loss: 1.2971... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8548... Training loss: 1.3809... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8549... Training loss: 1.3862... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8550... Training loss: 1.3265... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8551... Training loss: 1.4066... 0.0819 sec/batch
Epoch: 9/20... Training Step: 8552... Training loss: 1.3676... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8553... Training loss: 1.3870... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8554... Training loss: 1.4184... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8555... Training loss: 1.3227... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8556... Training loss: 1.3315... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8557... Training loss: 1.3143... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8558... Training loss: 1.2982... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8559... Training loss: 1.4083... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8560... Training loss: 1.3803... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8561... Training loss: 1.3015... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8562... Training loss: 1.3295... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8563... Training loss: 1.3445... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8564... Training loss: 1.3575... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8565... Training loss: 1.1971... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8566... Training loss: 1.3715... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8567... Training loss: 1.3523... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8568... Training loss: 1.3382... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8569... Training loss: 1.3093... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8570... Training loss: 1.3333... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8571... Training loss: 1.3699... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8572... Training loss: 1.3490... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8573... Training loss: 1.4304... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8574... Training loss: 1.3504... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8575... Training loss: 1.3653... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8576... Training loss: 1.4303... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8577... Training loss: 1.4499... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8578... Training loss: 1.4273... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8579... Training loss: 1.3425... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8580... Training loss: 1.3738... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8581... Training loss: 1.3220... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8582... Training loss: 1.4125... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8583... Training loss: 1.3892... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8584... Training loss: 1.2659... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8585... Training loss: 1.3224... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8586... Training loss: 1.3727... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8587... Training loss: 1.3435... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8588... Training loss: 1.3866... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8589... Training loss: 1.3601... 0.0742 sec/batch
Epoch: 9/20... Training Step: 8590... Training loss: 1.4145... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8591... Training loss: 1.4080... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8592... Training loss: 1.3951... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8593... Training loss: 1.2938... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8594... Training loss: 1.3577... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8595... Training loss: 1.3356... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8596... Training loss: 1.4330... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8597... Training loss: 1.4123... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8598... Training loss: 1.4360... 0.0740 sec/batch
Epoch: 9/20... Training Step: 8599... Training loss: 1.3875... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8600... Training loss: 1.4653... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8601... Training loss: 1.4598... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8602... Training loss: 1.3172... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8603... Training loss: 1.4894... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8604... Training loss: 1.3824... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8605... Training loss: 1.3835... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8606... Training loss: 1.3242... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8607... Training loss: 1.3931... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8608... Training loss: 1.4287... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8609... Training loss: 1.3647... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8610... Training loss: 1.3708... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8611... Training loss: 1.3040... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8612... Training loss: 1.3353... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8613... Training loss: 1.4532... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8614... Training loss: 1.3491... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8615... Training loss: 1.2905... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8616... Training loss: 1.4095... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8617... Training loss: 1.3930... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8618... Training loss: 1.4142... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8619... Training loss: 1.4016... 0.0788 sec/batch
Epoch: 9/20... Training Step: 8620... Training loss: 1.3349... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8621... Training loss: 1.4399... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8622... Training loss: 1.4310... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8623... Training loss: 1.3858... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8624... Training loss: 1.3475... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8625... Training loss: 1.3391... 0.0801 sec/batch
Epoch: 9/20... Training Step: 8626... Training loss: 1.4039... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8627... Training loss: 1.3569... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8628... Training loss: 1.3834... 0.0788 sec/batch
Epoch: 9/20... Training Step: 8629... Training loss: 1.3564... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8630... Training loss: 1.2909... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8631... Training loss: 1.3958... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8632... Training loss: 1.3627... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8633... Training loss: 1.4281... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8634... Training loss: 1.3795... 0.0804 sec/batch
Epoch: 9/20... Training Step: 8635... Training loss: 1.3634... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8636... Training loss: 1.3795... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8637... Training loss: 1.3653... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8638... Training loss: 1.3665... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8639... Training loss: 1.3591... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8640... Training loss: 1.3076... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8641... Training loss: 1.2981... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8642... Training loss: 1.3308... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8643... Training loss: 1.3564... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8644... Training loss: 1.3805... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8645... Training loss: 1.3893... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8646... Training loss: 1.3526... 0.0792 sec/batch
Epoch: 9/20... Training Step: 8647... Training loss: 1.3409... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8648... Training loss: 1.4379... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8649... Training loss: 1.3187... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8650... Training loss: 1.3720... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8651... Training loss: 1.3850... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8652... Training loss: 1.3939... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8653... Training loss: 1.4205... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8654... Training loss: 1.4022... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8655... Training loss: 1.4050... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8656... Training loss: 1.4754... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8657... Training loss: 1.3912... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8658... Training loss: 1.3708... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8659... Training loss: 1.3846... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8660... Training loss: 1.4051... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8661... Training loss: 1.4071... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8662... Training loss: 1.4455... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8663... Training loss: 1.3568... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8664... Training loss: 1.4149... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8665... Training loss: 1.4251... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8666... Training loss: 1.3232... 0.0782 sec/batch
Epoch: 9/20... Training Step: 8667... Training loss: 1.3739... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8668... Training loss: 1.3994... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8669... Training loss: 1.3312... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8670... Training loss: 1.4122... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8671... Training loss: 1.3739... 0.0793 sec/batch
Epoch: 9/20... Training Step: 8672... Training loss: 1.3376... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8673... Training loss: 1.3483... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8674... Training loss: 1.4364... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8675... Training loss: 1.3440... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8676... Training loss: 1.3696... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8677... Training loss: 1.3167... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8678... Training loss: 1.3483... 0.0741 sec/batch
Epoch: 9/20... Training Step: 8679... Training loss: 1.3542... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8680... Training loss: 1.4167... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8681... Training loss: 1.3881... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8682... Training loss: 1.3657... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8683... Training loss: 1.3628... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8684... Training loss: 1.3697... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8685... Training loss: 1.3147... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8686... Training loss: 1.3171... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8687... Training loss: 1.4579... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8688... Training loss: 1.3224... 0.0772 sec/batch
Epoch: 9/20... Training Step: 8689... Training loss: 1.4035... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8690... Training loss: 1.3990... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8691... Training loss: 1.3097... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8692... Training loss: 1.3807... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8693... Training loss: 1.3873... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8694... Training loss: 1.4344... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8695... Training loss: 1.3958... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8696... Training loss: 1.4654... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8697... Training loss: 1.3774... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8698... Training loss: 1.3173... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8699... Training loss: 1.3855... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8700... Training loss: 1.3679... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8701... Training loss: 1.4073... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8702... Training loss: 1.4135... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8703... Training loss: 1.3689... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8704... Training loss: 1.3260... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8705... Training loss: 1.3382... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8706... Training loss: 1.3675... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8707... Training loss: 1.3939... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8708... Training loss: 1.4334... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8709... Training loss: 1.4263... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8710... Training loss: 1.4072... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8711... Training loss: 1.3477... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8712... Training loss: 1.2633... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8713... Training loss: 1.3364... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8714... Training loss: 1.3863... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8715... Training loss: 1.3531... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8716... Training loss: 1.3557... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8717... Training loss: 1.3746... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8718... Training loss: 1.3654... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8719... Training loss: 1.4096... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8720... Training loss: 1.3918... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8721... Training loss: 1.3804... 0.0813 sec/batch
Epoch: 9/20... Training Step: 8722... Training loss: 1.3567... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8723... Training loss: 1.3507... 0.0785 sec/batch
Epoch: 9/20... Training Step: 8724... Training loss: 1.3896... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8725... Training loss: 1.3883... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8726... Training loss: 1.3934... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8727... Training loss: 1.3597... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8728... Training loss: 1.3479... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8729... Training loss: 1.3588... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8730... Training loss: 1.3355... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8731... Training loss: 1.2828... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8732... Training loss: 1.3531... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8733... Training loss: 1.3251... 0.0795 sec/batch
Epoch: 9/20... Training Step: 8734... Training loss: 1.3547... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8735... Training loss: 1.3473... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8736... Training loss: 1.3269... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8737... Training loss: 1.3977... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8738... Training loss: 1.4399... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8739... Training loss: 1.4418... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8740... Training loss: 1.3738... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8741... Training loss: 1.3246... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8742... Training loss: 1.3631... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8743... Training loss: 1.3532... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8744... Training loss: 1.4086... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8745... Training loss: 1.4122... 0.0786 sec/batch
Epoch: 9/20... Training Step: 8746... Training loss: 1.4300... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8747... Training loss: 1.3697... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8748... Training loss: 1.4395... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8749... Training loss: 1.3791... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8750... Training loss: 1.4241... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8751... Training loss: 1.4299... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8752... Training loss: 1.3936... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8753... Training loss: 1.4381... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8754... Training loss: 1.4863... 0.0830 sec/batch
Epoch: 9/20... Training Step: 8755... Training loss: 1.4041... 0.0743 sec/batch
Epoch: 9/20... Training Step: 8756... Training loss: 1.3735... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8757... Training loss: 1.3667... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8758... Training loss: 1.3187... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8759... Training loss: 1.3779... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8760... Training loss: 1.3905... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8761... Training loss: 1.3608... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8762... Training loss: 1.4043... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8763... Training loss: 1.3941... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8764... Training loss: 1.4397... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8765... Training loss: 1.3988... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8766... Training loss: 1.3358... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8767... Training loss: 1.4166... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8768... Training loss: 1.4087... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8769... Training loss: 1.4517... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8770... Training loss: 1.3840... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8771... Training loss: 1.2994... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8772... Training loss: 1.3545... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8773... Training loss: 1.3815... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8774... Training loss: 1.4041... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8775... Training loss: 1.4125... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8776... Training loss: 1.3857... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8777... Training loss: 1.3731... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8778... Training loss: 1.5287... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8779... Training loss: 1.4241... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8780... Training loss: 1.4440... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8781... Training loss: 1.4231... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8782... Training loss: 1.3223... 0.0790 sec/batch
Epoch: 9/20... Training Step: 8783... Training loss: 1.3792... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8784... Training loss: 1.3635... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8785... Training loss: 1.3834... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8786... Training loss: 1.4162... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8787... Training loss: 1.3480... 0.0778 sec/batch
Epoch: 9/20... Training Step: 8788... Training loss: 1.4723... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8789... Training loss: 1.4542... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8790... Training loss: 1.3739... 0.0834 sec/batch
Epoch: 9/20... Training Step: 8791... Training loss: 1.3769... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8792... Training loss: 1.3946... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8793... Training loss: 1.3948... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8794... Training loss: 1.3557... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8795... Training loss: 1.4252... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8796... Training loss: 1.4431... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8797... Training loss: 1.3934... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8798... Training loss: 1.3171... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8799... Training loss: 1.3310... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8800... Training loss: 1.3588... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8801... Training loss: 1.3051... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8802... Training loss: 1.3930... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8803... Training loss: 1.3776... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8804... Training loss: 1.4327... 0.0789 sec/batch
Epoch: 9/20... Training Step: 8805... Training loss: 1.3952... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8806... Training loss: 1.4406... 0.0740 sec/batch
Epoch: 9/20... Training Step: 8807... Training loss: 1.4114... 0.0798 sec/batch
Epoch: 9/20... Training Step: 8808... Training loss: 1.3901... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8809... Training loss: 1.3981... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8810... Training loss: 1.3896... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8811... Training loss: 1.4009... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8812... Training loss: 1.4259... 0.0759 sec/batch
Epoch: 9/20... Training Step: 8813... Training loss: 1.3903... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8814... Training loss: 1.4485... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8815... Training loss: 1.3970... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8816... Training loss: 1.3744... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8817... Training loss: 1.3695... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8818... Training loss: 1.3035... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8819... Training loss: 1.4143... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8820... Training loss: 1.3766... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8821... Training loss: 1.3763... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8822... Training loss: 1.4447... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8823... Training loss: 1.3929... 0.0771 sec/batch
Epoch: 9/20... Training Step: 8824... Training loss: 1.4142... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8825... Training loss: 1.4181... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8826... Training loss: 1.4327... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8827... Training loss: 1.4316... 0.0758 sec/batch
Epoch: 9/20... Training Step: 8828... Training loss: 1.3049... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8829... Training loss: 1.3466... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8830... Training loss: 1.3202... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8831... Training loss: 1.4163... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8832... Training loss: 1.3467... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8833... Training loss: 1.4080... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8834... Training loss: 1.4166... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8835... Training loss: 1.3432... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8836... Training loss: 1.3289... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8837... Training loss: 1.3377... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8838... Training loss: 1.3536... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8839... Training loss: 1.3194... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8840... Training loss: 1.3204... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8841... Training loss: 1.3078... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8842... Training loss: 1.3850... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8843... Training loss: 1.3122... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8844... Training loss: 1.3710... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8845... Training loss: 1.3113... 0.0766 sec/batch
Epoch: 9/20... Training Step: 8846... Training loss: 1.2594... 0.0781 sec/batch
Epoch: 9/20... Training Step: 8847... Training loss: 1.3080... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8848... Training loss: 1.3825... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8849... Training loss: 1.5041... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8850... Training loss: 1.3778... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8851... Training loss: 1.3712... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8852... Training loss: 1.3862... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8853... Training loss: 1.3457... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8854... Training loss: 1.4199... 0.0756 sec/batch
Epoch: 9/20... Training Step: 8855... Training loss: 1.3779... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8856... Training loss: 1.3871... 0.0749 sec/batch
Epoch: 9/20... Training Step: 8857... Training loss: 1.4832... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8858... Training loss: 1.4100... 0.0827 sec/batch
Epoch: 9/20... Training Step: 8859... Training loss: 1.3744... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8860... Training loss: 1.3335... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8861... Training loss: 1.3890... 0.0796 sec/batch
Epoch: 9/20... Training Step: 8862... Training loss: 1.4104... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8863... Training loss: 1.3739... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8864... Training loss: 1.3563... 0.0796 sec/batch
Epoch: 9/20... Training Step: 8865... Training loss: 1.4632... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8866... Training loss: 1.4349... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8867... Training loss: 1.3904... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8868... Training loss: 1.4774... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8869... Training loss: 1.3792... 0.0763 sec/batch
Epoch: 9/20... Training Step: 8870... Training loss: 1.3911... 0.0770 sec/batch
Epoch: 9/20... Training Step: 8871... Training loss: 1.4007... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8872... Training loss: 1.4093... 0.0751 sec/batch
Epoch: 9/20... Training Step: 8873... Training loss: 1.3412... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8874... Training loss: 1.3641... 0.0739 sec/batch
Epoch: 9/20... Training Step: 8875... Training loss: 1.3773... 0.0750 sec/batch
Epoch: 9/20... Training Step: 8876... Training loss: 1.4657... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8877... Training loss: 1.4034... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8878... Training loss: 1.4087... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8879... Training loss: 1.4433... 0.0840 sec/batch
Epoch: 9/20... Training Step: 8880... Training loss: 1.4172... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8881... Training loss: 1.4293... 0.0774 sec/batch
Epoch: 9/20... Training Step: 8882... Training loss: 1.3727... 0.0797 sec/batch
Epoch: 9/20... Training Step: 8883... Training loss: 1.3724... 0.0745 sec/batch
Epoch: 9/20... Training Step: 8884... Training loss: 1.3441... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8885... Training loss: 1.3809... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8886... Training loss: 1.3083... 0.0765 sec/batch
Epoch: 9/20... Training Step: 8887... Training loss: 1.3750... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8888... Training loss: 1.3224... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8889... Training loss: 1.3569... 0.0760 sec/batch
Epoch: 9/20... Training Step: 8890... Training loss: 1.3984... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8891... Training loss: 1.4055... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8892... Training loss: 1.3954... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8893... Training loss: 1.3579... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8894... Training loss: 1.4032... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8895... Training loss: 1.3380... 0.0815 sec/batch
Epoch: 9/20... Training Step: 8896... Training loss: 1.3504... 0.0764 sec/batch
Epoch: 9/20... Training Step: 8897... Training loss: 1.3844... 0.0779 sec/batch
Epoch: 9/20... Training Step: 8898... Training loss: 1.3692... 0.0783 sec/batch
Epoch: 9/20... Training Step: 8899... Training loss: 1.3650... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8900... Training loss: 1.3872... 0.0834 sec/batch
Epoch: 9/20... Training Step: 8901... Training loss: 1.4352... 0.0780 sec/batch
Epoch: 9/20... Training Step: 8902... Training loss: 1.4169... 0.0762 sec/batch
Epoch: 9/20... Training Step: 8903... Training loss: 1.4331... 0.0790 sec/batch
Epoch: 9/20... Training Step: 8904... Training loss: 1.4296... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8905... Training loss: 1.4247... 0.0752 sec/batch
Epoch: 9/20... Training Step: 8906... Training loss: 1.5497... 0.0784 sec/batch
Epoch: 9/20... Training Step: 8907... Training loss: 1.3342... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8908... Training loss: 1.4706... 0.0757 sec/batch
Epoch: 9/20... Training Step: 8909... Training loss: 1.5030... 0.0767 sec/batch
Epoch: 9/20... Training Step: 8910... Training loss: 1.3423... 0.0773 sec/batch
Epoch: 9/20... Training Step: 8911... Training loss: 1.4305... 0.0755 sec/batch
Epoch: 9/20... Training Step: 8912... Training loss: 1.4240... 0.0744 sec/batch
Epoch: 9/20... Training Step: 8913... Training loss: 1.4521... 0.0754 sec/batch
Epoch: 9/20... Training Step: 8914... Training loss: 1.4262... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8915... Training loss: 1.4138... 0.0769 sec/batch
Epoch: 9/20... Training Step: 8916... Training loss: 1.3620... 0.0788 sec/batch
Epoch: 9/20... Training Step: 8917... Training loss: 1.2971... 0.0748 sec/batch
Epoch: 9/20... Training Step: 8918... Training loss: 1.4739... 0.0768 sec/batch
Epoch: 9/20... Training Step: 8919... Training loss: 1.4082... 0.0746 sec/batch
Epoch: 9/20... Training Step: 8920... Training loss: 1.3901... 0.0777 sec/batch
Epoch: 9/20... Training Step: 8921... Training loss: 1.3905... 0.0816 sec/batch
Epoch: 9/20... Training Step: 8922... Training loss: 1.3329... 0.0787 sec/batch
Epoch: 9/20... Training Step: 8923... Training loss: 1.3598... 0.0775 sec/batch
Epoch: 9/20... Training Step: 8924... Training loss: 1.3862... 0.0753 sec/batch
Epoch: 9/20... Training Step: 8925... Training loss: 1.4031... 0.0761 sec/batch
Epoch: 9/20... Training Step: 8926... Training loss: 1.2646... 0.0747 sec/batch
Epoch: 9/20... Training Step: 8927... Training loss: 1.3674... 0.0776 sec/batch
Epoch: 9/20... Training Step: 8928... Training loss: 1.4109... 0.0760 sec/batch
Epoch: 10/20... Training Step: 8929... Training loss: 1.4626... 0.0743 sec/batch
Epoch: 10/20... Training Step: 8930... Training loss: 1.3631... 0.0745 sec/batch
Epoch: 10/20... Training Step: 8931... Training loss: 1.4426... 0.0779 sec/batch
Epoch: 10/20... Training Step: 8932... Training loss: 1.3944... 0.0753 sec/batch
Epoch: 10/20... Training Step: 8933... Training loss: 1.2995... 0.0749 sec/batch
Epoch: 10/20... Training Step: 8934... Training loss: 1.2536... 0.0770 sec/batch
Epoch: 10/20... Training Step: 8935... Training loss: 1.3517... 0.0787 sec/batch
Epoch: 10/20... Training Step: 8936... Training loss: 1.3979... 0.0842 sec/batch
Epoch: 10/20... Training Step: 8937... Training loss: 1.3856... 0.0752 sec/batch
Epoch: 10/20... Training Step: 8938... Training loss: 1.2826... 0.0750 sec/batch
Epoch: 10/20... Training Step: 8939... Training loss: 1.4193... 0.0791 sec/batch
Epoch: 10/20... Training Step: 8940... Training loss: 1.3287... 0.0758 sec/batch
Epoch: 10/20... Training Step: 8941... Training loss: 1.3627... 0.0758 sec/batch
Epoch: 10/20... Training Step: 8942... Training loss: 1.3698... 0.0772 sec/batch
Epoch: 10/20... Training Step: 8943... Training loss: 1.4076... 0.0752 sec/batch
Epoch: 10/20... Training Step: 8944... Training loss: 1.2962... 0.0765 sec/batch
Epoch: 10/20... Training Step: 8945... Training loss: 1.3434... 0.0770 sec/batch
Epoch: 10/20... Training Step: 8946... Training loss: 1.3634... 0.0746 sec/batch
Epoch: 10/20... Training Step: 8947... Training loss: 1.3413... 0.0784 sec/batch
Epoch: 10/20... Training Step: 8948... Training loss: 1.3365... 0.0778 sec/batch
Epoch: 10/20... Training Step: 8949... Training loss: 1.3504... 0.0750 sec/batch
Epoch: 10/20... Training Step: 8950... Training loss: 1.3292... 0.0769 sec/batch
Epoch: 10/20... Training Step: 8951... Training loss: 1.3530... 0.0780 sec/batch
Epoch: 10/20... Training Step: 8952... Training loss: 1.2985... 0.0751 sec/batch
Epoch: 10/20... Training Step: 8953... Training loss: 1.4027... 0.0757 sec/batch
Epoch: 10/20... Training Step: 8954... Training loss: 1.3146... 0.0746 sec/batch
Epoch: 10/20... Training Step: 8955... Training loss: 1.3514... 0.0757 sec/batch
Epoch: 10/20... Training Step: 8956... Training loss: 1.2908... 0.0746 sec/batch
Epoch: 10/20... Training Step: 8957... Training loss: 1.3700... 0.0766 sec/batch
Epoch: 10/20... Training Step: 8958... Training loss: 1.3929... 0.0751 sec/batch
Epoch: 10/20... Training Step: 8959... Training loss: 1.3980... 0.0762 sec/batch
Epoch: 10/20... Training Step: 8960... Training loss: 1.4069... 0.0745 sec/batch
Epoch: 10/20... Training Step: 8961... Training loss: 1.3788... 0.0796 sec/batch
Epoch: 10/20... Training Step: 8962... Training loss: 1.4268... 0.0760 sec/batch
Epoch: 10/20... Training Step: 8963... Training loss: 1.4443... 0.0751 sec/batch
Epoch: 10/20... Training Step: 8964... Training loss: 1.4200... 0.0773 sec/batch
Epoch: 10/20... Training Step: 8965... Training loss: 1.3584... 0.0750 sec/batch
Epoch: 10/20... Training Step: 8966... Training loss: 1.4278... 0.0746 sec/batch
Epoch: 10/20... Training Step: 8967... Training loss: 1.3800... 0.0743 sec/batch
Epoch: 10/20... Training Step: 8968... Training loss: 1.3659... 0.0791 sec/batch
Epoch: 10/20... Training Step: 8969... Training loss: 1.3705... 0.0774 sec/batch
Epoch: 10/20... Training Step: 8970... Training loss: 1.3516... 0.0750 sec/batch
Epoch: 10/20... Training Step: 8971... Training loss: 1.3185... 0.0761 sec/batch
Epoch: 10/20... Training Step: 8972... Training loss: 1.3334... 0.0772 sec/batch
Epoch: 10/20... Training Step: 8973... Training loss: 1.3771... 0.0755 sec/batch
Epoch: 10/20... Training Step: 8974... Training loss: 1.3534... 0.0746 sec/batch
Epoch: 10/20... Training Step: 8975... Training loss: 1.4197... 0.0767 sec/batch
Epoch: 10/20... Training Step: 8976... Training loss: 1.3403... 0.0781 sec/batch
Epoch: 10/20... Training Step: 8977... Training loss: 1.3978... 0.0763 sec/batch
Epoch: 10/20... Training Step: 8978... Training loss: 1.3899... 0.0764 sec/batch
Epoch: 10/20... Training Step: 8979... Training loss: 1.2858... 0.0748 sec/batch
Epoch: 10/20... Training Step: 8980... Training loss: 1.4029... 0.0734 sec/batch
Epoch: 10/20... Training Step: 8981... Training loss: 1.2811... 0.0768 sec/batch
Epoch: 10/20... Training Step: 8982... Training loss: 1.3318... 0.0766 sec/batch
Epoch: 10/20... Training Step: 8983... Training loss: 1.3508... 0.0763 sec/batch
Epoch: 10/20... Training Step: 8984... Training loss: 1.3069... 0.0778 sec/batch
Epoch: 10/20... Training Step: 8985... Training loss: 1.3707... 0.0743 sec/batch
Epoch: 10/20... Training Step: 8986... Training loss: 1.4206... 0.0759 sec/batch
Epoch: 10/20... Training Step: 8987... Training loss: 1.3539... 0.0780 sec/batch
Epoch: 10/20... Training Step: 8988... Training loss: 1.3795... 0.0754 sec/batch
Epoch: 10/20... Training Step: 8989... Training loss: 1.2498... 0.0786 sec/batch
Epoch: 10/20... Training Step: 8990... Training loss: 1.2826... 0.0753 sec/batch
Epoch: 10/20... Training Step: 8991... Training loss: 1.3471... 0.0766 sec/batch
Epoch: 10/20... Training Step: 8992... Training loss: 1.3072... 0.0750 sec/batch
Epoch: 10/20... Training Step: 8993... Training loss: 1.3876... 0.0749 sec/batch
Epoch: 10/20... Training Step: 8994... Training loss: 1.3530... 0.0759 sec/batch
Epoch: 10/20... Training Step: 8995... Training loss: 1.3207... 0.0760 sec/batch
Epoch: 10/20... Training Step: 8996... Training loss: 1.3429... 0.0771 sec/batch
Epoch: 10/20... Training Step: 8997... Training loss: 1.3863... 0.0768 sec/batch
Epoch: 10/20... Training Step: 8998... Training loss: 1.3819... 0.0781 sec/batch
Epoch: 10/20... Training Step: 8999... Training loss: 1.3881... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9000... Training loss: 1.4453... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9001... Training loss: 1.3719... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9002... Training loss: 1.3177... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9003... Training loss: 1.3543... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9004... Training loss: 1.3229... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9005... Training loss: 1.3413... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9006... Training loss: 1.3496... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9007... Training loss: 1.3653... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9008... Training loss: 1.4092... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9009... Training loss: 1.4005... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9010... Training loss: 1.3470... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9011... Training loss: 1.3778... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9012... Training loss: 1.3663... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9013... Training loss: 1.3263... 0.0801 sec/batch
Epoch: 10/20... Training Step: 9014... Training loss: 1.4096... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9015... Training loss: 1.3569... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9016... Training loss: 1.3649... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9017... Training loss: 1.3901... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9018... Training loss: 1.3980... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9019... Training loss: 1.4227... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9020... Training loss: 1.3400... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9021... Training loss: 1.3155... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9022... Training loss: 1.2766... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9023... Training loss: 1.2683... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9024... Training loss: 1.3807... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9025... Training loss: 1.3487... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9026... Training loss: 1.3028... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9027... Training loss: 1.3248... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9028... Training loss: 1.3387... 0.0816 sec/batch
Epoch: 10/20... Training Step: 9029... Training loss: 1.3094... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9030... Training loss: 1.3383... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9031... Training loss: 1.3723... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9032... Training loss: 1.3830... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9033... Training loss: 1.3780... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9034... Training loss: 1.3627... 0.0792 sec/batch
Epoch: 10/20... Training Step: 9035... Training loss: 1.2841... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9036... Training loss: 1.4013... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9037... Training loss: 1.3677... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9038... Training loss: 1.3963... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9039... Training loss: 1.3830... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9040... Training loss: 1.3521... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9041... Training loss: 1.4365... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9042... Training loss: 1.4070... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9043... Training loss: 1.3340... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9044... Training loss: 1.4734... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9045... Training loss: 1.4049... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9046... Training loss: 1.4688... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9047... Training loss: 1.3923... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9048... Training loss: 1.3666... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9049... Training loss: 1.4473... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9050... Training loss: 1.3259... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9051... Training loss: 1.3800... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9052... Training loss: 1.3512... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9053... Training loss: 1.4310... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9054... Training loss: 1.4510... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9055... Training loss: 1.4099... 0.0793 sec/batch
Epoch: 10/20... Training Step: 9056... Training loss: 1.3784... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9057... Training loss: 1.3103... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9058... Training loss: 1.3684... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9059... Training loss: 1.3363... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9060... Training loss: 1.4070... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9061... Training loss: 1.3955... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9062... Training loss: 1.3706... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9063... Training loss: 1.4149... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9064... Training loss: 1.4012... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9065... Training loss: 1.4083... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9066... Training loss: 1.4636... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9067... Training loss: 1.4165... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9068... Training loss: 1.3711... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9069... Training loss: 1.3859... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9070... Training loss: 1.3422... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9071... Training loss: 1.3420... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9072... Training loss: 1.3629... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9073... Training loss: 1.2514... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9074... Training loss: 1.4072... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9075... Training loss: 1.3414... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9076... Training loss: 1.2933... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9077... Training loss: 1.3345... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9078... Training loss: 1.3118... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9079... Training loss: 1.2578... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9080... Training loss: 1.3616... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9081... Training loss: 1.2930... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9082... Training loss: 1.3499... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9083... Training loss: 1.3266... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9084... Training loss: 1.3176... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9085... Training loss: 1.3397... 0.0792 sec/batch
Epoch: 10/20... Training Step: 9086... Training loss: 1.3498... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9087... Training loss: 1.3545... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9088... Training loss: 1.3871... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9089... Training loss: 1.3334... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9090... Training loss: 1.3478... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9091... Training loss: 1.3665... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9092... Training loss: 1.4279... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9093... Training loss: 1.3646... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9094... Training loss: 1.4553... 0.0854 sec/batch
Epoch: 10/20... Training Step: 9095... Training loss: 1.4110... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9096... Training loss: 1.3593... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9097... Training loss: 1.4073... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9098... Training loss: 1.3619... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9099... Training loss: 1.3659... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9100... Training loss: 1.3479... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9101... Training loss: 1.3604... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9102... Training loss: 1.3936... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9103... Training loss: 1.3648... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9104... Training loss: 1.3901... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9105... Training loss: 1.3758... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9106... Training loss: 1.4745... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9107... Training loss: 1.4057... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9108... Training loss: 1.3477... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9109... Training loss: 1.3835... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9110... Training loss: 1.4190... 0.0793 sec/batch
Epoch: 10/20... Training Step: 9111... Training loss: 1.3562... 0.0799 sec/batch
Epoch: 10/20... Training Step: 9112... Training loss: 1.3863... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9113... Training loss: 1.3871... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9114... Training loss: 1.4103... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9115... Training loss: 1.3923... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9116... Training loss: 1.3709... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9117... Training loss: 1.3568... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9118... Training loss: 1.4373... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9119... Training loss: 1.4555... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9120... Training loss: 1.4021... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9121... Training loss: 1.4123... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9122... Training loss: 1.4830... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9123... Training loss: 1.3579... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9124... Training loss: 1.3313... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9125... Training loss: 1.4129... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9126... Training loss: 1.3863... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9127... Training loss: 1.3839... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9128... Training loss: 1.3601... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9129... Training loss: 1.3487... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9130... Training loss: 1.3668... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9131... Training loss: 1.3817... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9132... Training loss: 1.3489... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9133... Training loss: 1.3405... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9134... Training loss: 1.4071... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9135... Training loss: 1.3576... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9136... Training loss: 1.3525... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9137... Training loss: 1.4049... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9138... Training loss: 1.3901... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9139... Training loss: 1.3420... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9140... Training loss: 1.4083... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9141... Training loss: 1.3297... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9142... Training loss: 1.4162... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9143... Training loss: 1.3616... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9144... Training loss: 1.3842... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9145... Training loss: 1.3390... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9146... Training loss: 1.3903... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9147... Training loss: 1.3559... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9148... Training loss: 1.3235... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9149... Training loss: 1.3304... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9150... Training loss: 1.4638... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9151... Training loss: 1.3820... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9152... Training loss: 1.3861... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9153... Training loss: 1.4018... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9154... Training loss: 1.3458... 0.0790 sec/batch
Epoch: 10/20... Training Step: 9155... Training loss: 1.3757... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9156... Training loss: 1.3899... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9157... Training loss: 1.2861... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9158... Training loss: 1.3035... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9159... Training loss: 1.3366... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9160... Training loss: 1.3788... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9161... Training loss: 1.3615... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9162... Training loss: 1.4285... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9163... Training loss: 1.3444... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9164... Training loss: 1.3686... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9165... Training loss: 1.3270... 0.0736 sec/batch
Epoch: 10/20... Training Step: 9166... Training loss: 1.3931... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9167... Training loss: 1.4122... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9168... Training loss: 1.3567... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9169... Training loss: 1.3494... 0.0794 sec/batch
Epoch: 10/20... Training Step: 9170... Training loss: 1.3642... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9171... Training loss: 1.4050... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9172... Training loss: 1.4197... 0.0802 sec/batch
Epoch: 10/20... Training Step: 9173... Training loss: 1.3301... 0.0795 sec/batch
Epoch: 10/20... Training Step: 9174... Training loss: 1.4192... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9175... Training loss: 1.3193... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9176... Training loss: 1.2903... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9177... Training loss: 1.3695... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9178... Training loss: 1.3884... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9179... Training loss: 1.3461... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9180... Training loss: 1.3589... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9181... Training loss: 1.3232... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9182... Training loss: 1.3314... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9183... Training loss: 1.2867... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9184... Training loss: 1.3427... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9185... Training loss: 1.4340... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9186... Training loss: 1.3453... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9187... Training loss: 1.3448... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9188... Training loss: 1.3522... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9189... Training loss: 1.3901... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9190... Training loss: 1.4211... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9191... Training loss: 1.3916... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9192... Training loss: 1.3963... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9193... Training loss: 1.4011... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9194... Training loss: 1.2787... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9195... Training loss: 1.3364... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9196... Training loss: 1.3748... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9197... Training loss: 1.3011... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9198... Training loss: 1.3171... 0.0817 sec/batch
Epoch: 10/20... Training Step: 9199... Training loss: 1.3594... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9200... Training loss: 1.3332... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9201... Training loss: 1.2791... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9202... Training loss: 1.2851... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9203... Training loss: 1.3429... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9204... Training loss: 1.3260... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9205... Training loss: 1.3737... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9206... Training loss: 1.2588... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9207... Training loss: 1.3098... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9208... Training loss: 1.3995... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9209... Training loss: 1.3446... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9210... Training loss: 1.3345... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9211... Training loss: 1.3683... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9212... Training loss: 1.3144... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9213... Training loss: 1.2610... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9214... Training loss: 1.3217... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9215... Training loss: 1.3855... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9216... Training loss: 1.3440... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9217... Training loss: 1.3157... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9218... Training loss: 1.3367... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9219... Training loss: 1.3338... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9220... Training loss: 1.3474... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9221... Training loss: 1.2717... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9222... Training loss: 1.3065... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9223... Training loss: 1.3732... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9224... Training loss: 1.3802... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9225... Training loss: 1.3145... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9226... Training loss: 1.4535... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9227... Training loss: 1.3625... 0.0802 sec/batch
Epoch: 10/20... Training Step: 9228... Training loss: 1.3853... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9229... Training loss: 1.4256... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9230... Training loss: 1.3340... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9231... Training loss: 1.3134... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9232... Training loss: 1.3853... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9233... Training loss: 1.3412... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9234... Training loss: 1.4355... 0.0801 sec/batch
Epoch: 10/20... Training Step: 9235... Training loss: 1.3159... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9236... Training loss: 1.2799... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9237... Training loss: 1.4238... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9238... Training loss: 1.3346... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9239... Training loss: 1.4406... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9240... Training loss: 1.3365... 0.0796 sec/batch
Epoch: 10/20... Training Step: 9241... Training loss: 1.3363... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9242... Training loss: 1.3657... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9243... Training loss: 1.3768... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9244... Training loss: 1.3118... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9245... Training loss: 1.3203... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9246... Training loss: 1.3052... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9247... Training loss: 1.3301... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9248... Training loss: 1.3553... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9249... Training loss: 1.3603... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9250... Training loss: 1.2798... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9251... Training loss: 1.3095... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9252... Training loss: 1.3487... 0.0820 sec/batch
Epoch: 10/20... Training Step: 9253... Training loss: 1.4109... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9254... Training loss: 1.3455... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9255... Training loss: 1.3476... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9256... Training loss: 1.3251... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9257... Training loss: 1.3261... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9258... Training loss: 1.3654... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9259... Training loss: 1.4017... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9260... Training loss: 1.3012... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9261... Training loss: 1.4327... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9262... Training loss: 1.4511... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9263... Training loss: 1.4396... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9264... Training loss: 1.3788... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9265... Training loss: 1.3249... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9266... Training loss: 1.3377... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9267... Training loss: 1.3461... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9268... Training loss: 1.3335... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9269... Training loss: 1.3503... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9270... Training loss: 1.3086... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9271... Training loss: 1.3176... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9272... Training loss: 1.3792... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9273... Training loss: 1.3247... 0.0834 sec/batch
Epoch: 10/20... Training Step: 9274... Training loss: 1.3489... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9275... Training loss: 1.3821... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9276... Training loss: 1.3946... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9277... Training loss: 1.4047... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9278... Training loss: 1.4155... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9279... Training loss: 1.3740... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9280... Training loss: 1.3183... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9281... Training loss: 1.3548... 0.0790 sec/batch
Epoch: 10/20... Training Step: 9282... Training loss: 1.3332... 0.0804 sec/batch
Epoch: 10/20... Training Step: 9283... Training loss: 1.3639... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9284... Training loss: 1.3642... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9285... Training loss: 1.3205... 0.0795 sec/batch
Epoch: 10/20... Training Step: 9286... Training loss: 1.3648... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9287... Training loss: 1.3588... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9288... Training loss: 1.3509... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9289... Training loss: 1.4095... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9290... Training loss: 1.4328... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9291... Training loss: 1.3380... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9292... Training loss: 1.3555... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9293... Training loss: 1.3557... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9294... Training loss: 1.3226... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9295... Training loss: 1.3281... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9296... Training loss: 1.2935... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9297... Training loss: 1.2981... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9298... Training loss: 1.2927... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9299... Training loss: 1.4108... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9300... Training loss: 1.3063... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9301... Training loss: 1.3539... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9302... Training loss: 1.2951... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9303... Training loss: 1.3508... 0.0791 sec/batch
Epoch: 10/20... Training Step: 9304... Training loss: 1.3221... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9305... Training loss: 1.3471... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9306... Training loss: 1.3557... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9307... Training loss: 1.3306... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9308... Training loss: 1.3689... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9309... Training loss: 1.3541... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9310... Training loss: 1.3466... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9311... Training loss: 1.3909... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9312... Training loss: 1.3916... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9313... Training loss: 1.3764... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9314... Training loss: 1.3624... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9315... Training loss: 1.3813... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9316... Training loss: 1.4059... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9317... Training loss: 1.3952... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9318... Training loss: 1.3283... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9319... Training loss: 1.3762... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9320... Training loss: 1.3464... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9321... Training loss: 1.3815... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9322... Training loss: 1.3199... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9323... Training loss: 1.3493... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9324... Training loss: 1.3441... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9325... Training loss: 1.3337... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9326... Training loss: 1.4029... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9327... Training loss: 1.3870... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9328... Training loss: 1.3355... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9329... Training loss: 1.3112... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9330... Training loss: 1.4041... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9331... Training loss: 1.3183... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9332... Training loss: 1.3855... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9333... Training loss: 1.3426... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9334... Training loss: 1.3620... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9335... Training loss: 1.2896... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9336... Training loss: 1.4431... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9337... Training loss: 1.3100... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9338... Training loss: 1.3078... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9339... Training loss: 1.2518... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9340... Training loss: 1.3193... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9341... Training loss: 1.3326... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9342... Training loss: 1.2847... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9343... Training loss: 1.3119... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9344... Training loss: 1.3920... 0.0805 sec/batch
Epoch: 10/20... Training Step: 9345... Training loss: 1.2676... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9346... Training loss: 1.3109... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9347... Training loss: 1.3966... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9348... Training loss: 1.4471... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9349... Training loss: 1.2895... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9350... Training loss: 1.3713... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9351... Training loss: 1.3426... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9352... Training loss: 1.3727... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9353... Training loss: 1.3578... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9354... Training loss: 1.3702... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9355... Training loss: 1.4068... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9356... Training loss: 1.4361... 0.0794 sec/batch
Epoch: 10/20... Training Step: 9357... Training loss: 1.4129... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9358... Training loss: 1.4553... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9359... Training loss: 1.3620... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9360... Training loss: 1.3822... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9361... Training loss: 1.4247... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9362... Training loss: 1.3816... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9363... Training loss: 1.4028... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9364... Training loss: 1.3426... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9365... Training loss: 1.4118... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9366... Training loss: 1.4729... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9367... Training loss: 1.3902... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9368... Training loss: 1.3931... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9369... Training loss: 1.3608... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9370... Training loss: 1.4079... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9371... Training loss: 1.3704... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9372... Training loss: 1.3930... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9373... Training loss: 1.3561... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9374... Training loss: 1.4590... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9375... Training loss: 1.3681... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9376... Training loss: 1.3774... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9377... Training loss: 1.4301... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9378... Training loss: 1.4399... 0.0788 sec/batch
Epoch: 10/20... Training Step: 9379... Training loss: 1.3650... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9380... Training loss: 1.3431... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9381... Training loss: 1.3976... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9382... Training loss: 1.3864... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9383... Training loss: 1.4028... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9384... Training loss: 1.3640... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9385... Training loss: 1.3187... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9386... Training loss: 1.3318... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9387... Training loss: 1.3491... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9388... Training loss: 1.3627... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9389... Training loss: 1.4011... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9390... Training loss: 1.3524... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9391... Training loss: 1.3163... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9392... Training loss: 1.3939... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9393... Training loss: 1.3257... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9394... Training loss: 1.3207... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9395... Training loss: 1.2566... 0.0814 sec/batch
Epoch: 10/20... Training Step: 9396... Training loss: 1.3549... 0.0805 sec/batch
Epoch: 10/20... Training Step: 9397... Training loss: 1.3552... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9398... Training loss: 1.4079... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9399... Training loss: 1.3783... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9400... Training loss: 1.3389... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9401... Training loss: 1.3540... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9402... Training loss: 1.3637... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9403... Training loss: 1.2492... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9404... Training loss: 1.3404... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9405... Training loss: 1.3697... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9406... Training loss: 1.3436... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9407... Training loss: 1.3241... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9408... Training loss: 1.2978... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9409... Training loss: 1.3061... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9410... Training loss: 1.3648... 0.0829 sec/batch
Epoch: 10/20... Training Step: 9411... Training loss: 1.3463... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9412... Training loss: 1.3563... 0.0791 sec/batch
Epoch: 10/20... Training Step: 9413... Training loss: 1.2697... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9414... Training loss: 1.3275... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9415... Training loss: 1.4068... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9416... Training loss: 1.3323... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9417... Training loss: 1.3323... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9418... Training loss: 1.2877... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9419... Training loss: 1.3094... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9420... Training loss: 1.3080... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9421... Training loss: 1.3395... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9422... Training loss: 1.3342... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9423... Training loss: 1.3131... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9424... Training loss: 1.3184... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9425... Training loss: 1.3510... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9426... Training loss: 1.3486... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9427... Training loss: 1.3228... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9428... Training loss: 1.3736... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9429... Training loss: 1.3241... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9430... Training loss: 1.3525... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9431... Training loss: 1.3479... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9432... Training loss: 1.3378... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9433... Training loss: 1.3484... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9434... Training loss: 1.2952... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9435... Training loss: 1.3245... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9436... Training loss: 1.2938... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9437... Training loss: 1.2984... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9438... Training loss: 1.3136... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9439... Training loss: 1.3699... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9440... Training loss: 1.3562... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9441... Training loss: 1.3981... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9442... Training loss: 1.3870... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9443... Training loss: 1.3462... 0.0798 sec/batch
Epoch: 10/20... Training Step: 9444... Training loss: 1.3669... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9445... Training loss: 1.3589... 0.0737 sec/batch
Epoch: 10/20... Training Step: 9446... Training loss: 1.3698... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9447... Training loss: 1.3608... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9448... Training loss: 1.2933... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9449... Training loss: 1.3307... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9450... Training loss: 1.3026... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9451... Training loss: 1.4447... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9452... Training loss: 1.3331... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9453... Training loss: 1.3411... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9454... Training loss: 1.3524... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9455... Training loss: 1.3326... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9456... Training loss: 1.3243... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9457... Training loss: 1.3243... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9458... Training loss: 1.3968... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9459... Training loss: 1.4567... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9460... Training loss: 1.3733... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9461... Training loss: 1.3996... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9462... Training loss: 1.3979... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9463... Training loss: 1.3627... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9464... Training loss: 1.3247... 0.0810 sec/batch
Epoch: 10/20... Training Step: 9465... Training loss: 1.3100... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9466... Training loss: 1.3158... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9467... Training loss: 1.3211... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9468... Training loss: 1.3247... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9469... Training loss: 1.4395... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9470... Training loss: 1.3173... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9471... Training loss: 1.3158... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9472... Training loss: 1.3405... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9473... Training loss: 1.2913... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9474... Training loss: 1.3188... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9475... Training loss: 1.4099... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9476... Training loss: 1.3820... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9477... Training loss: 1.3038... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9478... Training loss: 1.3258... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9479... Training loss: 1.3329... 0.0832 sec/batch
Epoch: 10/20... Training Step: 9480... Training loss: 1.2900... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9481... Training loss: 1.2908... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9482... Training loss: 1.3339... 0.0796 sec/batch
Epoch: 10/20... Training Step: 9483... Training loss: 1.3684... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9484... Training loss: 1.3254... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9485... Training loss: 1.2899... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9486... Training loss: 1.3343... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9487... Training loss: 1.3554... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9488... Training loss: 1.2636... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9489... Training loss: 1.3074... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9490... Training loss: 1.3400... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9491... Training loss: 1.3139... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9492... Training loss: 1.3126... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9493... Training loss: 1.3440... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9494... Training loss: 1.3560... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9495... Training loss: 1.3201... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9496... Training loss: 1.2953... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9497... Training loss: 1.2584... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9498... Training loss: 1.2845... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9499... Training loss: 1.2821... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9500... Training loss: 1.3013... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9501... Training loss: 1.3923... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9502... Training loss: 1.3744... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9503... Training loss: 1.3713... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9504... Training loss: 1.3530... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9505... Training loss: 1.3513... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9506... Training loss: 1.4924... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9507... Training loss: 1.3975... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9508... Training loss: 1.3021... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9509... Training loss: 1.2695... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9510... Training loss: 1.3036... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9511... Training loss: 1.4039... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9512... Training loss: 1.2699... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9513... Training loss: 1.3160... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9514... Training loss: 1.3518... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9515... Training loss: 1.3638... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9516... Training loss: 1.4093... 0.0788 sec/batch
Epoch: 10/20... Training Step: 9517... Training loss: 1.3424... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9518... Training loss: 1.3477... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9519... Training loss: 1.3623... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9520... Training loss: 1.3669... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9521... Training loss: 1.2949... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9522... Training loss: 1.3140... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9523... Training loss: 1.2883... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9524... Training loss: 1.3766... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9525... Training loss: 1.2165... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9526... Training loss: 1.2349... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9527... Training loss: 1.3186... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9528... Training loss: 1.3103... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9529... Training loss: 1.3921... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9530... Training loss: 1.3318... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9531... Training loss: 1.3227... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9532... Training loss: 1.3607... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9533... Training loss: 1.3047... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9534... Training loss: 1.3313... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9535... Training loss: 1.3959... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9536... Training loss: 1.3590... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9537... Training loss: 1.3604... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9538... Training loss: 1.3910... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9539... Training loss: 1.2689... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9540... Training loss: 1.3891... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9541... Training loss: 1.3903... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9542... Training loss: 1.3131... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9543... Training loss: 1.4073... 0.0811 sec/batch
Epoch: 10/20... Training Step: 9544... Training loss: 1.3498... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9545... Training loss: 1.3604... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9546... Training loss: 1.3724... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9547... Training loss: 1.2889... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9548... Training loss: 1.3353... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9549... Training loss: 1.3132... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9550... Training loss: 1.2665... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9551... Training loss: 1.4038... 0.0809 sec/batch
Epoch: 10/20... Training Step: 9552... Training loss: 1.3470... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9553... Training loss: 1.3084... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9554... Training loss: 1.3044... 0.0788 sec/batch
Epoch: 10/20... Training Step: 9555... Training loss: 1.3227... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9556... Training loss: 1.3669... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9557... Training loss: 1.1545... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9558... Training loss: 1.3553... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9559... Training loss: 1.3371... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9560... Training loss: 1.3123... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9561... Training loss: 1.2850... 0.0737 sec/batch
Epoch: 10/20... Training Step: 9562... Training loss: 1.3202... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9563... Training loss: 1.3657... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9564... Training loss: 1.3262... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9565... Training loss: 1.4069... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9566... Training loss: 1.3250... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9567... Training loss: 1.3435... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9568... Training loss: 1.4154... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9569... Training loss: 1.4442... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9570... Training loss: 1.4247... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9571... Training loss: 1.3279... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9572... Training loss: 1.3513... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9573... Training loss: 1.3092... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9574... Training loss: 1.3659... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9575... Training loss: 1.3766... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9576... Training loss: 1.2689... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9577... Training loss: 1.3024... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9578... Training loss: 1.3381... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9579... Training loss: 1.3326... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9580... Training loss: 1.3809... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9581... Training loss: 1.3406... 0.0788 sec/batch
Epoch: 10/20... Training Step: 9582... Training loss: 1.3974... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9583... Training loss: 1.4091... 0.0797 sec/batch
Epoch: 10/20... Training Step: 9584... Training loss: 1.3786... 0.0796 sec/batch
Epoch: 10/20... Training Step: 9585... Training loss: 1.2771... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9586... Training loss: 1.3082... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9587... Training loss: 1.3208... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9588... Training loss: 1.4196... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9589... Training loss: 1.3856... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9590... Training loss: 1.4289... 0.0791 sec/batch
Epoch: 10/20... Training Step: 9591... Training loss: 1.3478... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9592... Training loss: 1.4372... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9593... Training loss: 1.4457... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9594... Training loss: 1.3044... 0.0791 sec/batch
Epoch: 10/20... Training Step: 9595... Training loss: 1.4776... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9596... Training loss: 1.3944... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9597... Training loss: 1.3464... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9598... Training loss: 1.3137... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9599... Training loss: 1.3778... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9600... Training loss: 1.4204... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9601... Training loss: 1.3539... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9602... Training loss: 1.3415... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9603... Training loss: 1.3056... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9604... Training loss: 1.3333... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9605... Training loss: 1.4354... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9606... Training loss: 1.3456... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9607... Training loss: 1.2966... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9608... Training loss: 1.4102... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9609... Training loss: 1.4042... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9610... Training loss: 1.3747... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9611... Training loss: 1.3766... 0.0798 sec/batch
Epoch: 10/20... Training Step: 9612... Training loss: 1.3321... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9613... Training loss: 1.4124... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9614... Training loss: 1.3874... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9615... Training loss: 1.3514... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9616... Training loss: 1.3652... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9617... Training loss: 1.3014... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9618... Training loss: 1.3723... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9619... Training loss: 1.3489... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9620... Training loss: 1.3457... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9621... Training loss: 1.3488... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9622... Training loss: 1.2835... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9623... Training loss: 1.3620... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9624... Training loss: 1.3563... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9625... Training loss: 1.4478... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9626... Training loss: 1.3673... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9627... Training loss: 1.3289... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9628... Training loss: 1.3368... 0.0792 sec/batch
Epoch: 10/20... Training Step: 9629... Training loss: 1.3674... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9630... Training loss: 1.3314... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9631... Training loss: 1.3705... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9632... Training loss: 1.2818... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9633... Training loss: 1.2868... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9634... Training loss: 1.3039... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9635... Training loss: 1.3516... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9636... Training loss: 1.3581... 0.0805 sec/batch
Epoch: 10/20... Training Step: 9637... Training loss: 1.3759... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9638... Training loss: 1.3219... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9639... Training loss: 1.3313... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9640... Training loss: 1.4316... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9641... Training loss: 1.3070... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9642... Training loss: 1.3436... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9643... Training loss: 1.3626... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9644... Training loss: 1.3876... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9645... Training loss: 1.4008... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9646... Training loss: 1.3904... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9647... Training loss: 1.3929... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9648... Training loss: 1.4611... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9649... Training loss: 1.3923... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9650... Training loss: 1.3679... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9651... Training loss: 1.3425... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9652... Training loss: 1.3661... 0.0822 sec/batch
Epoch: 10/20... Training Step: 9653... Training loss: 1.4198... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9654... Training loss: 1.4425... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9655... Training loss: 1.3535... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9656... Training loss: 1.3976... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9657... Training loss: 1.4050... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9658... Training loss: 1.3136... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9659... Training loss: 1.3527... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9660... Training loss: 1.3682... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9661... Training loss: 1.3375... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9662... Training loss: 1.4070... 0.0793 sec/batch
Epoch: 10/20... Training Step: 9663... Training loss: 1.3475... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9664... Training loss: 1.3277... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9665... Training loss: 1.3077... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9666... Training loss: 1.4336... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9667... Training loss: 1.3172... 0.0799 sec/batch
Epoch: 10/20... Training Step: 9668... Training loss: 1.3455... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9669... Training loss: 1.3146... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9670... Training loss: 1.3491... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9671... Training loss: 1.3276... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9672... Training loss: 1.4167... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9673... Training loss: 1.3811... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9674... Training loss: 1.3445... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9675... Training loss: 1.3678... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9676... Training loss: 1.3482... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9677... Training loss: 1.2979... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9678... Training loss: 1.2994... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9679... Training loss: 1.4469... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9680... Training loss: 1.3154... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9681... Training loss: 1.3859... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9682... Training loss: 1.3757... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9683... Training loss: 1.2787... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9684... Training loss: 1.3440... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9685... Training loss: 1.3755... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9686... Training loss: 1.4012... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9687... Training loss: 1.4121... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9688... Training loss: 1.4375... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9689... Training loss: 1.3691... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9690... Training loss: 1.3030... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9691... Training loss: 1.3845... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9692... Training loss: 1.3433... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9693... Training loss: 1.3857... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9694... Training loss: 1.3804... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9695... Training loss: 1.3557... 0.0740 sec/batch
Epoch: 10/20... Training Step: 9696... Training loss: 1.3186... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9697... Training loss: 1.3316... 0.0760 sec/batch
Epoch: 10/20... Training Step: 9698... Training loss: 1.3730... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9699... Training loss: 1.3828... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9700... Training loss: 1.4088... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9701... Training loss: 1.3996... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9702... Training loss: 1.4002... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9703... Training loss: 1.3481... 0.0821 sec/batch
Epoch: 10/20... Training Step: 9704... Training loss: 1.2753... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9705... Training loss: 1.3116... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9706... Training loss: 1.3442... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9707... Training loss: 1.3219... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9708... Training loss: 1.3144... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9709... Training loss: 1.3555... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9710... Training loss: 1.3568... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9711... Training loss: 1.3830... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9712... Training loss: 1.3617... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9713... Training loss: 1.3787... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9714... Training loss: 1.3746... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9715... Training loss: 1.3665... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9716... Training loss: 1.3862... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9717... Training loss: 1.3831... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9718... Training loss: 1.3773... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9719... Training loss: 1.3471... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9720... Training loss: 1.3482... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9721... Training loss: 1.3499... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9722... Training loss: 1.3105... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9723... Training loss: 1.2760... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9724... Training loss: 1.3492... 0.0811 sec/batch
Epoch: 10/20... Training Step: 9725... Training loss: 1.3207... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9726... Training loss: 1.3300... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9727... Training loss: 1.3127... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9728... Training loss: 1.3227... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9729... Training loss: 1.3746... 0.0774 sec/batch
Epoch: 10/20... Training Step: 9730... Training loss: 1.3967... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9731... Training loss: 1.4208... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9732... Training loss: 1.3586... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9733... Training loss: 1.3044... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9734... Training loss: 1.3448... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9735... Training loss: 1.3424... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9736... Training loss: 1.3783... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9737... Training loss: 1.4226... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9738... Training loss: 1.4132... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9739... Training loss: 1.3699... 0.0772 sec/batch
Epoch: 10/20... Training Step: 9740... Training loss: 1.4048... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9741... Training loss: 1.3706... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9742... Training loss: 1.4030... 0.0755 sec/batch
Epoch: 10/20... Training Step: 9743... Training loss: 1.4047... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9744... Training loss: 1.4077... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9745... Training loss: 1.4005... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9746... Training loss: 1.4818... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9747... Training loss: 1.3866... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9748... Training loss: 1.3571... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9749... Training loss: 1.3395... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9750... Training loss: 1.3217... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9751... Training loss: 1.3706... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9752... Training loss: 1.3445... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9753... Training loss: 1.3699... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9754... Training loss: 1.3937... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9755... Training loss: 1.3776... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9756... Training loss: 1.4195... 0.0744 sec/batch
Epoch: 10/20... Training Step: 9757... Training loss: 1.4001... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9758... Training loss: 1.3281... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9759... Training loss: 1.4148... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9760... Training loss: 1.3951... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9761... Training loss: 1.4326... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9762... Training loss: 1.3855... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9763... Training loss: 1.2988... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9764... Training loss: 1.3471... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9765... Training loss: 1.3875... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9766... Training loss: 1.3841... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9767... Training loss: 1.4059... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9768... Training loss: 1.3607... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9769... Training loss: 1.3812... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9770... Training loss: 1.5012... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9771... Training loss: 1.4069... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9772... Training loss: 1.4252... 0.0786 sec/batch
Epoch: 10/20... Training Step: 9773... Training loss: 1.3865... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9774... Training loss: 1.2896... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9775... Training loss: 1.3745... 0.0793 sec/batch
Epoch: 10/20... Training Step: 9776... Training loss: 1.3507... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9777... Training loss: 1.3581... 0.0767 sec/batch
Epoch: 10/20... Training Step: 9778... Training loss: 1.3937... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9779... Training loss: 1.3521... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9780... Training loss: 1.4606... 0.0741 sec/batch
Epoch: 10/20... Training Step: 9781... Training loss: 1.4354... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9782... Training loss: 1.3719... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9783... Training loss: 1.3642... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9784... Training loss: 1.3536... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9785... Training loss: 1.3510... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9786... Training loss: 1.3482... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9787... Training loss: 1.3957... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9788... Training loss: 1.4163... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9789... Training loss: 1.3641... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9790... Training loss: 1.3211... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9791... Training loss: 1.3129... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9792... Training loss: 1.3631... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9793... Training loss: 1.3073... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9794... Training loss: 1.3690... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9795... Training loss: 1.3844... 0.0769 sec/batch
Epoch: 10/20... Training Step: 9796... Training loss: 1.4338... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9797... Training loss: 1.3843... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9798... Training loss: 1.4396... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9799... Training loss: 1.4186... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9800... Training loss: 1.3783... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9801... Training loss: 1.3818... 0.0763 sec/batch
Epoch: 10/20... Training Step: 9802... Training loss: 1.3633... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9803... Training loss: 1.3730... 0.0736 sec/batch
Epoch: 10/20... Training Step: 9804... Training loss: 1.4150... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9805... Training loss: 1.3874... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9806... Training loss: 1.4353... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9807... Training loss: 1.3588... 0.0810 sec/batch
Epoch: 10/20... Training Step: 9808... Training loss: 1.3518... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9809... Training loss: 1.3740... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9810... Training loss: 1.3004... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9811... Training loss: 1.3868... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9812... Training loss: 1.3469... 0.0766 sec/batch
Epoch: 10/20... Training Step: 9813... Training loss: 1.3638... 0.0791 sec/batch
Epoch: 10/20... Training Step: 9814... Training loss: 1.4140... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9815... Training loss: 1.3670... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9816... Training loss: 1.4095... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9817... Training loss: 1.3927... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9818... Training loss: 1.4521... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9819... Training loss: 1.4185... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9820... Training loss: 1.2920... 0.0787 sec/batch
Epoch: 10/20... Training Step: 9821... Training loss: 1.3520... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9822... Training loss: 1.3046... 0.0840 sec/batch
Epoch: 10/20... Training Step: 9823... Training loss: 1.4141... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9824... Training loss: 1.3218... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9825... Training loss: 1.3903... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9826... Training loss: 1.4051... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9827... Training loss: 1.3486... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9828... Training loss: 1.3060... 0.0792 sec/batch
Epoch: 10/20... Training Step: 9829... Training loss: 1.3412... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9830... Training loss: 1.3138... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9831... Training loss: 1.3222... 0.0775 sec/batch
Epoch: 10/20... Training Step: 9832... Training loss: 1.3055... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9833... Training loss: 1.2878... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9834... Training loss: 1.3618... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9835... Training loss: 1.3040... 0.0810 sec/batch
Epoch: 10/20... Training Step: 9836... Training loss: 1.3541... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9837... Training loss: 1.3012... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9838... Training loss: 1.2623... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9839... Training loss: 1.2951... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9840... Training loss: 1.3538... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9841... Training loss: 1.4409... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9842... Training loss: 1.3495... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9843... Training loss: 1.3460... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9844... Training loss: 1.3413... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9845... Training loss: 1.3503... 0.0762 sec/batch
Epoch: 10/20... Training Step: 9846... Training loss: 1.4120... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9847... Training loss: 1.3600... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9848... Training loss: 1.4013... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9849... Training loss: 1.4462... 0.0781 sec/batch
Epoch: 10/20... Training Step: 9850... Training loss: 1.4166... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9851... Training loss: 1.3839... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9852... Training loss: 1.3141... 0.0778 sec/batch
Epoch: 10/20... Training Step: 9853... Training loss: 1.3741... 0.0780 sec/batch
Epoch: 10/20... Training Step: 9854... Training loss: 1.3882... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9855... Training loss: 1.3513... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9856... Training loss: 1.3537... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9857... Training loss: 1.4656... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9858... Training loss: 1.4107... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9859... Training loss: 1.3742... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9860... Training loss: 1.4491... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9861... Training loss: 1.3373... 0.0789 sec/batch
Epoch: 10/20... Training Step: 9862... Training loss: 1.3819... 0.0770 sec/batch
Epoch: 10/20... Training Step: 9863... Training loss: 1.3671... 0.0742 sec/batch
Epoch: 10/20... Training Step: 9864... Training loss: 1.3783... 0.0794 sec/batch
Epoch: 10/20... Training Step: 9865... Training loss: 1.3545... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9866... Training loss: 1.3775... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9867... Training loss: 1.3629... 0.0784 sec/batch
Epoch: 10/20... Training Step: 9868... Training loss: 1.4654... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9869... Training loss: 1.3749... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9870... Training loss: 1.3888... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9871... Training loss: 1.4261... 0.0771 sec/batch
Epoch: 10/20... Training Step: 9872... Training loss: 1.3932... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9873... Training loss: 1.4345... 0.0810 sec/batch
Epoch: 10/20... Training Step: 9874... Training loss: 1.3358... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9875... Training loss: 1.3573... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9876... Training loss: 1.3384... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9877... Training loss: 1.3775... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9878... Training loss: 1.2918... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9879... Training loss: 1.3579... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9880... Training loss: 1.3217... 0.0749 sec/batch
Epoch: 10/20... Training Step: 9881... Training loss: 1.3255... 0.0764 sec/batch
Epoch: 10/20... Training Step: 9882... Training loss: 1.3815... 0.0768 sec/batch
Epoch: 10/20... Training Step: 9883... Training loss: 1.3921... 0.0747 sec/batch
Epoch: 10/20... Training Step: 9884... Training loss: 1.3734... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9885... Training loss: 1.3549... 0.0753 sec/batch
Epoch: 10/20... Training Step: 9886... Training loss: 1.3847... 0.0757 sec/batch
Epoch: 10/20... Training Step: 9887... Training loss: 1.3124... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9888... Training loss: 1.3172... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9889... Training loss: 1.3844... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9890... Training loss: 1.3557... 0.0743 sec/batch
Epoch: 10/20... Training Step: 9891... Training loss: 1.3768... 0.0761 sec/batch
Epoch: 10/20... Training Step: 9892... Training loss: 1.3477... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9893... Training loss: 1.4246... 0.0756 sec/batch
Epoch: 10/20... Training Step: 9894... Training loss: 1.4125... 0.0785 sec/batch
Epoch: 10/20... Training Step: 9895... Training loss: 1.4309... 0.0748 sec/batch
Epoch: 10/20... Training Step: 9896... Training loss: 1.4226... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9897... Training loss: 1.4116... 0.0765 sec/batch
Epoch: 10/20... Training Step: 9898... Training loss: 1.5538... 0.0752 sec/batch
Epoch: 10/20... Training Step: 9899... Training loss: 1.3456... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9900... Training loss: 1.4566... 0.0758 sec/batch
Epoch: 10/20... Training Step: 9901... Training loss: 1.4557... 0.0783 sec/batch
Epoch: 10/20... Training Step: 9902... Training loss: 1.3313... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9903... Training loss: 1.4016... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9904... Training loss: 1.4215... 0.0779 sec/batch
Epoch: 10/20... Training Step: 9905... Training loss: 1.4431... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9906... Training loss: 1.3934... 0.0746 sec/batch
Epoch: 10/20... Training Step: 9907... Training loss: 1.4273... 0.0738 sec/batch
Epoch: 10/20... Training Step: 9908... Training loss: 1.3535... 0.0745 sec/batch
Epoch: 10/20... Training Step: 9909... Training loss: 1.3091... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9910... Training loss: 1.4592... 0.0750 sec/batch
Epoch: 10/20... Training Step: 9911... Training loss: 1.3667... 0.0773 sec/batch
Epoch: 10/20... Training Step: 9912... Training loss: 1.4043... 0.0782 sec/batch
Epoch: 10/20... Training Step: 9913... Training loss: 1.3881... 0.0801 sec/batch
Epoch: 10/20... Training Step: 9914... Training loss: 1.3167... 0.0754 sec/batch
Epoch: 10/20... Training Step: 9915... Training loss: 1.3307... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9916... Training loss: 1.3562... 0.0759 sec/batch
Epoch: 10/20... Training Step: 9917... Training loss: 1.3789... 0.0776 sec/batch
Epoch: 10/20... Training Step: 9918... Training loss: 1.2580... 0.0777 sec/batch
Epoch: 10/20... Training Step: 9919... Training loss: 1.3482... 0.0751 sec/batch
Epoch: 10/20... Training Step: 9920... Training loss: 1.4185... 0.0763 sec/batch
Epoch: 11/20... Training Step: 9921... Training loss: 1.4388... 0.0795 sec/batch
Epoch: 11/20... Training Step: 9922... Training loss: 1.3288... 0.0777 sec/batch
Epoch: 11/20... Training Step: 9923... Training loss: 1.4311... 0.0765 sec/batch
Epoch: 11/20... Training Step: 9924... Training loss: 1.3699... 0.0761 sec/batch
Epoch: 11/20... Training Step: 9925... Training loss: 1.2949... 0.0744 sec/batch
Epoch: 11/20... Training Step: 9926... Training loss: 1.2347... 0.0785 sec/batch
Epoch: 11/20... Training Step: 9927... Training loss: 1.3508... 0.0750 sec/batch
Epoch: 11/20... Training Step: 9928... Training loss: 1.3796... 0.0756 sec/batch
Epoch: 11/20... Training Step: 9929... Training loss: 1.3898... 0.0777 sec/batch
Epoch: 11/20... Training Step: 9930... Training loss: 1.2727... 0.0745 sec/batch
Epoch: 11/20... Training Step: 9931... Training loss: 1.4140... 0.0776 sec/batch
Epoch: 11/20... Training Step: 9932... Training loss: 1.3196... 0.0740 sec/batch
Epoch: 11/20... Training Step: 9933... Training loss: 1.3375... 0.0753 sec/batch
Epoch: 11/20... Training Step: 9934... Training loss: 1.3682... 0.0751 sec/batch
Epoch: 11/20... Training Step: 9935... Training loss: 1.4139... 0.0745 sec/batch
Epoch: 11/20... Training Step: 9936... Training loss: 1.3067... 0.0768 sec/batch
Epoch: 11/20... Training Step: 9937... Training loss: 1.3495... 0.0776 sec/batch
Epoch: 11/20... Training Step: 9938... Training loss: 1.3740... 0.0760 sec/batch
Epoch: 11/20... Training Step: 9939... Training loss: 1.3238... 0.0755 sec/batch
Epoch: 11/20... Training Step: 9940... Training loss: 1.3098... 0.0757 sec/batch
Epoch: 11/20... Training Step: 9941... Training loss: 1.3570... 0.0752 sec/batch
Epoch: 11/20... Training Step: 9942... Training loss: 1.3085... 0.0771 sec/batch
Epoch: 11/20... Training Step: 9943... Training loss: 1.3347... 0.0775 sec/batch
Epoch: 11/20... Training Step: 9944... Training loss: 1.2853... 0.0764 sec/batch
Epoch: 11/20... Training Step: 9945... Training loss: 1.3922... 0.0759 sec/batch
Epoch: 11/20... Training Step: 9946... Training loss: 1.2937... 0.0752 sec/batch
Epoch: 11/20... Training Step: 9947... Training loss: 1.3372... 0.0774 sec/batch
Epoch: 11/20... Training Step: 9948... Training loss: 1.2931... 0.0760 sec/batch
Epoch: 11/20... Training Step: 9949... Training loss: 1.3529... 0.0771 sec/batch
Epoch: 11/20... Training Step: 9950... Training loss: 1.4021... 0.0773 sec/batch
Epoch: 11/20... Training Step: 9951... Training loss: 1.3771... 0.0761 sec/batch
Epoch: 11/20... Training Step: 9952... Training loss: 1.4058... 0.0755 sec/batch
Epoch: 11/20... Training Step: 9953... Training loss: 1.3365... 0.0780 sec/batch
Epoch: 11/20... Training Step: 9954... Training loss: 1.4073... 0.0783 sec/batch
Epoch: 11/20... Training Step: 9955... Training loss: 1.4296... 0.0747 sec/batch
Epoch: 11/20... Training Step: 9956... Training loss: 1.4052... 0.0766 sec/batch
Epoch: 11/20... Training Step: 9957... Training loss: 1.3550... 0.0750 sec/batch
Epoch: 11/20... Training Step: 9958... Training loss: 1.4035... 0.0753 sec/batch
Epoch: 11/20... Training Step: 9959... Training loss: 1.3608... 0.0740 sec/batch
Epoch: 11/20... Training Step: 9960... Training loss: 1.3304... 0.0780 sec/batch
Epoch: 11/20... Training Step: 9961... Training loss: 1.3924... 0.0751 sec/batch
Epoch: 11/20... Training Step: 9962... Training loss: 1.3301... 0.0750 sec/batch
Epoch: 11/20... Training Step: 9963... Training loss: 1.3054... 0.0761 sec/batch
Epoch: 11/20... Training Step: 9964... Training loss: 1.3282... 0.0756 sec/batch
Epoch: 11/20... Training Step: 9965... Training loss: 1.3845... 0.0779 sec/batch
Epoch: 11/20... Training Step: 9966... Training loss: 1.3305... 0.0777 sec/batch
Epoch: 11/20... Training Step: 9967... Training loss: 1.3976... 0.0747 sec/batch
Epoch: 11/20... Training Step: 9968... Training loss: 1.3153... 0.0765 sec/batch
Epoch: 11/20... Training Step: 9969... Training loss: 1.3888... 0.0751 sec/batch
Epoch: 11/20... Training Step: 9970... Training loss: 1.3979... 0.0751 sec/batch
Epoch: 11/20... Training Step: 9971... Training loss: 1.2654... 0.0754 sec/batch
Epoch: 11/20... Training Step: 9972... Training loss: 1.3654... 0.0777 sec/batch
Epoch: 11/20... Training Step: 9973... Training loss: 1.2703... 0.0749 sec/batch
Epoch: 11/20... Training Step: 9974... Training loss: 1.2978... 0.0778 sec/batch
Epoch: 11/20... Training Step: 9975... Training loss: 1.3224... 0.0755 sec/batch
Epoch: 11/20... Training Step: 9976... Training loss: 1.3022... 0.0754 sec/batch
Epoch: 11/20... Training Step: 9977... Training loss: 1.3637... 0.0760 sec/batch
Epoch: 11/20... Training Step: 9978... Training loss: 1.4028... 0.0788 sec/batch
Epoch: 11/20... Training Step: 9979... Training loss: 1.3471... 0.0761 sec/batch
Epoch: 11/20... Training Step: 9980... Training loss: 1.3872... 0.0777 sec/batch
Epoch: 11/20... Training Step: 9981... Training loss: 1.2559... 0.0762 sec/batch
Epoch: 11/20... Training Step: 9982... Training loss: 1.2846... 0.0778 sec/batch
Epoch: 11/20... Training Step: 9983... Training loss: 1.3242... 0.0753 sec/batch
Epoch: 11/20... Training Step: 9984... Training loss: 1.2855... 0.0760 sec/batch
Epoch: 11/20... Training Step: 9985... Training loss: 1.3470... 0.0766 sec/batch
Epoch: 11/20... Training Step: 9986... Training loss: 1.3489... 0.0764 sec/batch
Epoch: 11/20... Training Step: 9987... Training loss: 1.3240... 0.0759 sec/batch
Epoch: 11/20... Training Step: 9988... Training loss: 1.3302... 0.0751 sec/batch
Epoch: 11/20... Training Step: 9989... Training loss: 1.3833... 0.0745 sec/batch
Epoch: 11/20... Training Step: 9990... Training loss: 1.3537... 0.0781 sec/batch
Epoch: 11/20... Training Step: 9991... Training loss: 1.3853... 0.0750 sec/batch
Epoch: 11/20... Training Step: 9992... Training loss: 1.4128... 0.0802 sec/batch
Epoch: 11/20... Training Step: 9993... Training loss: 1.3559... 0.0753 sec/batch
Epoch: 11/20... Training Step: 9994... Training loss: 1.3211... 0.0752 sec/batch
Epoch: 11/20... Training Step: 9995... Training loss: 1.3235... 0.0748 sec/batch
Epoch: 11/20... Training Step: 9996... Training loss: 1.2897... 0.0760 sec/batch
Epoch: 11/20... Training Step: 9997... Training loss: 1.3266... 0.0746 sec/batch
Epoch: 11/20... Training Step: 9998... Training loss: 1.3323... 0.0757 sec/batch
Epoch: 11/20... Training Step: 9999... Training loss: 1.3507... 0.0791 sec/batch
Epoch: 11/20... Training Step: 10000... Training loss: 1.3986... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10001... Training loss: 1.3744... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10002... Training loss: 1.3351... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10003... Training loss: 1.3581... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10004... Training loss: 1.3759... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10005... Training loss: 1.3299... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10006... Training loss: 1.3951... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10007... Training loss: 1.3460... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10008... Training loss: 1.3537... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10009... Training loss: 1.3826... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10010... Training loss: 1.3867... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10011... Training loss: 1.4117... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10012... Training loss: 1.3421... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10013... Training loss: 1.3038... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10014... Training loss: 1.2603... 0.0864 sec/batch
Epoch: 11/20... Training Step: 10015... Training loss: 1.2483... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10016... Training loss: 1.3631... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10017... Training loss: 1.3317... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10018... Training loss: 1.2947... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10019... Training loss: 1.3113... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10020... Training loss: 1.3272... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10021... Training loss: 1.3059... 0.0799 sec/batch
Epoch: 11/20... Training Step: 10022... Training loss: 1.3101... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10023... Training loss: 1.3438... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10024... Training loss: 1.3884... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10025... Training loss: 1.3702... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10026... Training loss: 1.3297... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10027... Training loss: 1.2937... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10028... Training loss: 1.4048... 0.0790 sec/batch
Epoch: 11/20... Training Step: 10029... Training loss: 1.3342... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10030... Training loss: 1.3788... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10031... Training loss: 1.3596... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10032... Training loss: 1.3419... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10033... Training loss: 1.4339... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10034... Training loss: 1.3780... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10035... Training loss: 1.3124... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10036... Training loss: 1.4508... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10037... Training loss: 1.4200... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10038... Training loss: 1.4390... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10039... Training loss: 1.3589... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10040... Training loss: 1.3500... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10041... Training loss: 1.4683... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10042... Training loss: 1.3016... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10043... Training loss: 1.3802... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10044... Training loss: 1.3465... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10045... Training loss: 1.4022... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10046... Training loss: 1.4205... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10047... Training loss: 1.4187... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10048... Training loss: 1.3879... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10049... Training loss: 1.2844... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10050... Training loss: 1.3557... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10051... Training loss: 1.3296... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10052... Training loss: 1.3917... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10053... Training loss: 1.3814... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10054... Training loss: 1.3416... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10055... Training loss: 1.3923... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10056... Training loss: 1.3836... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10057... Training loss: 1.3966... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10058... Training loss: 1.4176... 0.0798 sec/batch
Epoch: 11/20... Training Step: 10059... Training loss: 1.4030... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10060... Training loss: 1.3450... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10061... Training loss: 1.3684... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10062... Training loss: 1.3378... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10063... Training loss: 1.3219... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10064... Training loss: 1.3396... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10065... Training loss: 1.2385... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10066... Training loss: 1.3982... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10067... Training loss: 1.3425... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10068... Training loss: 1.2954... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10069... Training loss: 1.3408... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10070... Training loss: 1.2998... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10071... Training loss: 1.2353... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10072... Training loss: 1.3553... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10073... Training loss: 1.2779... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10074... Training loss: 1.3160... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10075... Training loss: 1.3118... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10076... Training loss: 1.2919... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10077... Training loss: 1.3281... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10078... Training loss: 1.3325... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10079... Training loss: 1.3602... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10080... Training loss: 1.3676... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10081... Training loss: 1.3193... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10082... Training loss: 1.3383... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10083... Training loss: 1.3224... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10084... Training loss: 1.3995... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10085... Training loss: 1.3640... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10086... Training loss: 1.4786... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10087... Training loss: 1.3694... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10088... Training loss: 1.3531... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10089... Training loss: 1.3784... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10090... Training loss: 1.3589... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10091... Training loss: 1.3708... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10092... Training loss: 1.3551... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10093... Training loss: 1.3504... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10094... Training loss: 1.3930... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10095... Training loss: 1.3291... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10096... Training loss: 1.3577... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10097... Training loss: 1.3614... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10098... Training loss: 1.4472... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10099... Training loss: 1.3923... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10100... Training loss: 1.3132... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10101... Training loss: 1.3698... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10102... Training loss: 1.4354... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10103... Training loss: 1.3473... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10104... Training loss: 1.3580... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10105... Training loss: 1.3615... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10106... Training loss: 1.4071... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10107... Training loss: 1.3685... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10108... Training loss: 1.3563... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10109... Training loss: 1.3782... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10110... Training loss: 1.4125... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10111... Training loss: 1.4200... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10112... Training loss: 1.3868... 0.0791 sec/batch
Epoch: 11/20... Training Step: 10113... Training loss: 1.4054... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10114... Training loss: 1.4454... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10115... Training loss: 1.3474... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10116... Training loss: 1.3191... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10117... Training loss: 1.4275... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10118... Training loss: 1.3739... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10119... Training loss: 1.4045... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10120... Training loss: 1.3537... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10121... Training loss: 1.3460... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10122... Training loss: 1.3731... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10123... Training loss: 1.3650... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10124... Training loss: 1.3341... 0.0828 sec/batch
Epoch: 11/20... Training Step: 10125... Training loss: 1.3270... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10126... Training loss: 1.4095... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10127... Training loss: 1.3360... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10128... Training loss: 1.3271... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10129... Training loss: 1.3987... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10130... Training loss: 1.3723... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10131... Training loss: 1.3087... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10132... Training loss: 1.4012... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10133... Training loss: 1.3458... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10134... Training loss: 1.4026... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10135... Training loss: 1.3340... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10136... Training loss: 1.3609... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10137... Training loss: 1.3240... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10138... Training loss: 1.3510... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10139... Training loss: 1.3692... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10140... Training loss: 1.3091... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10141... Training loss: 1.2935... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10142... Training loss: 1.4568... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10143... Training loss: 1.3701... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10144... Training loss: 1.3859... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10145... Training loss: 1.3972... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10146... Training loss: 1.3336... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10147... Training loss: 1.3710... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10148... Training loss: 1.3741... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10149... Training loss: 1.2868... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10150... Training loss: 1.3008... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10151... Training loss: 1.3442... 0.0822 sec/batch
Epoch: 11/20... Training Step: 10152... Training loss: 1.3528... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10153... Training loss: 1.3609... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10154... Training loss: 1.4432... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10155... Training loss: 1.3196... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10156... Training loss: 1.3687... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10157... Training loss: 1.3289... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10158... Training loss: 1.4028... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10159... Training loss: 1.4004... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10160... Training loss: 1.3277... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10161... Training loss: 1.3682... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10162... Training loss: 1.3587... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10163... Training loss: 1.4061... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10164... Training loss: 1.3995... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10165... Training loss: 1.3543... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10166... Training loss: 1.3905... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10167... Training loss: 1.3066... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10168... Training loss: 1.2710... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10169... Training loss: 1.3344... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10170... Training loss: 1.3547... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10171... Training loss: 1.3300... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10172... Training loss: 1.3433... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10173... Training loss: 1.3113... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10174... Training loss: 1.3040... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10175... Training loss: 1.2877... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10176... Training loss: 1.3183... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10177... Training loss: 1.4276... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10178... Training loss: 1.3267... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10179... Training loss: 1.3434... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10180... Training loss: 1.3671... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10181... Training loss: 1.3893... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10182... Training loss: 1.3900... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10183... Training loss: 1.3852... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10184... Training loss: 1.3827... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10185... Training loss: 1.3706... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10186... Training loss: 1.2910... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10187... Training loss: 1.3273... 0.0788 sec/batch
Epoch: 11/20... Training Step: 10188... Training loss: 1.3639... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10189... Training loss: 1.3097... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10190... Training loss: 1.3028... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10191... Training loss: 1.3380... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10192... Training loss: 1.3173... 0.0795 sec/batch
Epoch: 11/20... Training Step: 10193... Training loss: 1.2723... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10194... Training loss: 1.2907... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10195... Training loss: 1.3241... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10196... Training loss: 1.3384... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10197... Training loss: 1.3530... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10198... Training loss: 1.2449... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10199... Training loss: 1.3070... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10200... Training loss: 1.3983... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10201... Training loss: 1.3450... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10202... Training loss: 1.3167... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10203... Training loss: 1.3700... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10204... Training loss: 1.3034... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10205... Training loss: 1.2304... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10206... Training loss: 1.2907... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10207... Training loss: 1.3619... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10208... Training loss: 1.3059... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10209... Training loss: 1.3242... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10210... Training loss: 1.3014... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10211... Training loss: 1.3108... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10212... Training loss: 1.3461... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10213... Training loss: 1.2702... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10214... Training loss: 1.2916... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10215... Training loss: 1.3456... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10216... Training loss: 1.3572... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10217... Training loss: 1.2884... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10218... Training loss: 1.4229... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10219... Training loss: 1.3108... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10220... Training loss: 1.3818... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10221... Training loss: 1.4141... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10222... Training loss: 1.3087... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10223... Training loss: 1.3018... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10224... Training loss: 1.3899... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10225... Training loss: 1.3407... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10226... Training loss: 1.4098... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10227... Training loss: 1.3067... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10228... Training loss: 1.2693... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10229... Training loss: 1.4063... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10230... Training loss: 1.3293... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10231... Training loss: 1.4198... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10232... Training loss: 1.3780... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10233... Training loss: 1.3183... 0.0811 sec/batch
Epoch: 11/20... Training Step: 10234... Training loss: 1.3293... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10235... Training loss: 1.3348... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10236... Training loss: 1.3134... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10237... Training loss: 1.3162... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10238... Training loss: 1.2740... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10239... Training loss: 1.3478... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10240... Training loss: 1.3654... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10241... Training loss: 1.3532... 0.0731 sec/batch
Epoch: 11/20... Training Step: 10242... Training loss: 1.2624... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10243... Training loss: 1.3060... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10244... Training loss: 1.3218... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10245... Training loss: 1.3802... 0.0739 sec/batch
Epoch: 11/20... Training Step: 10246... Training loss: 1.3065... 0.0830 sec/batch
Epoch: 11/20... Training Step: 10247... Training loss: 1.3393... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10248... Training loss: 1.3124... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10249... Training loss: 1.3399... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10250... Training loss: 1.3749... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10251... Training loss: 1.3921... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10252... Training loss: 1.2889... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10253... Training loss: 1.4417... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10254... Training loss: 1.4243... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10255... Training loss: 1.4012... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10256... Training loss: 1.3691... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10257... Training loss: 1.3128... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10258... Training loss: 1.3428... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10259... Training loss: 1.3417... 0.0792 sec/batch
Epoch: 11/20... Training Step: 10260... Training loss: 1.3282... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10261... Training loss: 1.3563... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10262... Training loss: 1.2939... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10263... Training loss: 1.2859... 0.0741 sec/batch
Epoch: 11/20... Training Step: 10264... Training loss: 1.3613... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10265... Training loss: 1.3453... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10266... Training loss: 1.3443... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10267... Training loss: 1.3470... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10268... Training loss: 1.3820... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10269... Training loss: 1.3788... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10270... Training loss: 1.4003... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10271... Training loss: 1.3760... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10272... Training loss: 1.3291... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10273... Training loss: 1.3337... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10274... Training loss: 1.3006... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10275... Training loss: 1.3465... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10276... Training loss: 1.3469... 0.0788 sec/batch
Epoch: 11/20... Training Step: 10277... Training loss: 1.3035... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10278... Training loss: 1.3368... 0.0741 sec/batch
Epoch: 11/20... Training Step: 10279... Training loss: 1.3465... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10280... Training loss: 1.3341... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10281... Training loss: 1.3550... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10282... Training loss: 1.4197... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10283... Training loss: 1.3400... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10284... Training loss: 1.3227... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10285... Training loss: 1.3586... 0.0794 sec/batch
Epoch: 11/20... Training Step: 10286... Training loss: 1.3192... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10287... Training loss: 1.3298... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10288... Training loss: 1.2790... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10289... Training loss: 1.3005... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10290... Training loss: 1.2685... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10291... Training loss: 1.3782... 0.0789 sec/batch
Epoch: 11/20... Training Step: 10292... Training loss: 1.2711... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10293... Training loss: 1.3204... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10294... Training loss: 1.2824... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10295... Training loss: 1.3204... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10296... Training loss: 1.3113... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10297... Training loss: 1.3146... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10298... Training loss: 1.3506... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10299... Training loss: 1.3228... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10300... Training loss: 1.3503... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10301... Training loss: 1.3359... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10302... Training loss: 1.3196... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10303... Training loss: 1.3617... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10304... Training loss: 1.3608... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10305... Training loss: 1.3653... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10306... Training loss: 1.3477... 0.0836 sec/batch
Epoch: 11/20... Training Step: 10307... Training loss: 1.3774... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10308... Training loss: 1.3709... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10309... Training loss: 1.3763... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10310... Training loss: 1.3150... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10311... Training loss: 1.3626... 0.0829 sec/batch
Epoch: 11/20... Training Step: 10312... Training loss: 1.3259... 0.0788 sec/batch
Epoch: 11/20... Training Step: 10313... Training loss: 1.3696... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10314... Training loss: 1.3108... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10315... Training loss: 1.3299... 0.0821 sec/batch
Epoch: 11/20... Training Step: 10316... Training loss: 1.3337... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10317... Training loss: 1.3379... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10318... Training loss: 1.3818... 0.0805 sec/batch
Epoch: 11/20... Training Step: 10319... Training loss: 1.3519... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10320... Training loss: 1.3136... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10321... Training loss: 1.3047... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10322... Training loss: 1.3650... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10323... Training loss: 1.3076... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10324... Training loss: 1.3860... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10325... Training loss: 1.3146... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10326... Training loss: 1.3508... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10327... Training loss: 1.2932... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10328... Training loss: 1.4257... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10329... Training loss: 1.3105... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10330... Training loss: 1.2964... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10331... Training loss: 1.2508... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10332... Training loss: 1.2974... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10333... Training loss: 1.3414... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10334... Training loss: 1.2518... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10335... Training loss: 1.2860... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10336... Training loss: 1.3703... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10337... Training loss: 1.2694... 0.0795 sec/batch
Epoch: 11/20... Training Step: 10338... Training loss: 1.2955... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10339... Training loss: 1.3683... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10340... Training loss: 1.4036... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10341... Training loss: 1.3030... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10342... Training loss: 1.3616... 0.0828 sec/batch
Epoch: 11/20... Training Step: 10343... Training loss: 1.3510... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10344... Training loss: 1.3465... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10345... Training loss: 1.3416... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10346... Training loss: 1.3517... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10347... Training loss: 1.4125... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10348... Training loss: 1.4064... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10349... Training loss: 1.4013... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10350... Training loss: 1.4492... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10351... Training loss: 1.3412... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10352... Training loss: 1.3635... 0.0806 sec/batch
Epoch: 11/20... Training Step: 10353... Training loss: 1.3947... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10354... Training loss: 1.3794... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10355... Training loss: 1.3767... 0.0740 sec/batch
Epoch: 11/20... Training Step: 10356... Training loss: 1.3059... 0.0741 sec/batch
Epoch: 11/20... Training Step: 10357... Training loss: 1.3911... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10358... Training loss: 1.4505... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10359... Training loss: 1.3892... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10360... Training loss: 1.4007... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10361... Training loss: 1.3256... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10362... Training loss: 1.3959... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10363... Training loss: 1.3761... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10364... Training loss: 1.3823... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10365... Training loss: 1.3263... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10366... Training loss: 1.4568... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10367... Training loss: 1.3570... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10368... Training loss: 1.3576... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10369... Training loss: 1.4365... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10370... Training loss: 1.4657... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10371... Training loss: 1.3530... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10372... Training loss: 1.3326... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10373... Training loss: 1.4055... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10374... Training loss: 1.3575... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10375... Training loss: 1.3798... 0.0838 sec/batch
Epoch: 11/20... Training Step: 10376... Training loss: 1.3391... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10377... Training loss: 1.3325... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10378... Training loss: 1.3506... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10379... Training loss: 1.3488... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10380... Training loss: 1.3613... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10381... Training loss: 1.3891... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10382... Training loss: 1.3370... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10383... Training loss: 1.3017... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10384... Training loss: 1.3497... 0.0789 sec/batch
Epoch: 11/20... Training Step: 10385... Training loss: 1.3105... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10386... Training loss: 1.3238... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10387... Training loss: 1.2706... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10388... Training loss: 1.3078... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10389... Training loss: 1.3413... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10390... Training loss: 1.4148... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10391... Training loss: 1.3599... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10392... Training loss: 1.3542... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10393... Training loss: 1.3460... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10394... Training loss: 1.3495... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10395... Training loss: 1.2423... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10396... Training loss: 1.3269... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10397... Training loss: 1.3846... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10398... Training loss: 1.3297... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10399... Training loss: 1.2834... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10400... Training loss: 1.3010... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10401... Training loss: 1.2924... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10402... Training loss: 1.3587... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10403... Training loss: 1.3465... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10404... Training loss: 1.3267... 0.0740 sec/batch
Epoch: 11/20... Training Step: 10405... Training loss: 1.2803... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10406... Training loss: 1.2985... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10407... Training loss: 1.3734... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10408... Training loss: 1.2948... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10409... Training loss: 1.2849... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10410... Training loss: 1.2902... 0.0792 sec/batch
Epoch: 11/20... Training Step: 10411... Training loss: 1.3095... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10412... Training loss: 1.2933... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10413... Training loss: 1.3379... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10414... Training loss: 1.3139... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10415... Training loss: 1.2964... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10416... Training loss: 1.3149... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10417... Training loss: 1.3292... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10418... Training loss: 1.3296... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10419... Training loss: 1.2977... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10420... Training loss: 1.3456... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10421... Training loss: 1.2856... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10422... Training loss: 1.3311... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10423... Training loss: 1.3461... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10424... Training loss: 1.3200... 0.0787 sec/batch
Epoch: 11/20... Training Step: 10425... Training loss: 1.3545... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10426... Training loss: 1.2819... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10427... Training loss: 1.3226... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10428... Training loss: 1.3157... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10429... Training loss: 1.2974... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10430... Training loss: 1.3217... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10431... Training loss: 1.3836... 0.0795 sec/batch
Epoch: 11/20... Training Step: 10432... Training loss: 1.3547... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10433... Training loss: 1.3484... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10434... Training loss: 1.3858... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10435... Training loss: 1.3311... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10436... Training loss: 1.3510... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10437... Training loss: 1.3691... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10438... Training loss: 1.3454... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10439... Training loss: 1.3435... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10440... Training loss: 1.2879... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10441... Training loss: 1.3561... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10442... Training loss: 1.3002... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10443... Training loss: 1.4304... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10444... Training loss: 1.3329... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10445... Training loss: 1.3309... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10446... Training loss: 1.3178... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10447... Training loss: 1.3178... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10448... Training loss: 1.3007... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10449... Training loss: 1.3307... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10450... Training loss: 1.3969... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10451... Training loss: 1.4584... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10452... Training loss: 1.3825... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10453... Training loss: 1.3972... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10454... Training loss: 1.3882... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10455... Training loss: 1.3443... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10456... Training loss: 1.3033... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10457... Training loss: 1.2987... 0.0798 sec/batch
Epoch: 11/20... Training Step: 10458... Training loss: 1.3224... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10459... Training loss: 1.3240... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10460... Training loss: 1.3155... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10461... Training loss: 1.4199... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10462... Training loss: 1.3210... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10463... Training loss: 1.3301... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10464... Training loss: 1.3171... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10465... Training loss: 1.2718... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10466... Training loss: 1.3236... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10467... Training loss: 1.4010... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10468... Training loss: 1.3673... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10469... Training loss: 1.2947... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10470... Training loss: 1.3185... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10471... Training loss: 1.3274... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10472... Training loss: 1.2830... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10473... Training loss: 1.2792... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10474... Training loss: 1.3117... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10475... Training loss: 1.3401... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10476... Training loss: 1.3097... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10477... Training loss: 1.2860... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10478... Training loss: 1.3070... 0.0741 sec/batch
Epoch: 11/20... Training Step: 10479... Training loss: 1.3319... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10480... Training loss: 1.2469... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10481... Training loss: 1.3081... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10482... Training loss: 1.3200... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10483... Training loss: 1.2807... 0.0802 sec/batch
Epoch: 11/20... Training Step: 10484... Training loss: 1.3131... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10485... Training loss: 1.3517... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10486... Training loss: 1.3362... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10487... Training loss: 1.2992... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10488... Training loss: 1.2677... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10489... Training loss: 1.2345... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10490... Training loss: 1.2721... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10491... Training loss: 1.2701... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10492... Training loss: 1.2997... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10493... Training loss: 1.3825... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10494... Training loss: 1.3557... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10495... Training loss: 1.3735... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10496... Training loss: 1.3336... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10497... Training loss: 1.3311... 0.0738 sec/batch
Epoch: 11/20... Training Step: 10498... Training loss: 1.4460... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10499... Training loss: 1.3811... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10500... Training loss: 1.2809... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10501... Training loss: 1.2596... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10502... Training loss: 1.3060... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10503... Training loss: 1.3777... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10504... Training loss: 1.2645... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10505... Training loss: 1.3385... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10506... Training loss: 1.3578... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10507... Training loss: 1.3526... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10508... Training loss: 1.3777... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10509... Training loss: 1.3388... 0.0787 sec/batch
Epoch: 11/20... Training Step: 10510... Training loss: 1.3264... 0.0791 sec/batch
Epoch: 11/20... Training Step: 10511... Training loss: 1.3283... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10512... Training loss: 1.3554... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10513... Training loss: 1.2911... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10514... Training loss: 1.2787... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10515... Training loss: 1.2406... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10516... Training loss: 1.3533... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10517... Training loss: 1.2052... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10518... Training loss: 1.2563... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10519... Training loss: 1.2911... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10520... Training loss: 1.3157... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10521... Training loss: 1.3760... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10522... Training loss: 1.3005... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10523... Training loss: 1.2890... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10524... Training loss: 1.3346... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10525... Training loss: 1.2795... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10526... Training loss: 1.3027... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10527... Training loss: 1.3761... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10528... Training loss: 1.3324... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10529... Training loss: 1.3352... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10530... Training loss: 1.3992... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10531... Training loss: 1.2550... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10532... Training loss: 1.3576... 0.0796 sec/batch
Epoch: 11/20... Training Step: 10533... Training loss: 1.3685... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10534... Training loss: 1.3007... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10535... Training loss: 1.3969... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10536... Training loss: 1.3258... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10537... Training loss: 1.3480... 0.0788 sec/batch
Epoch: 11/20... Training Step: 10538... Training loss: 1.3631... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10539... Training loss: 1.2787... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10540... Training loss: 1.3135... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10541... Training loss: 1.3267... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10542... Training loss: 1.2693... 0.0809 sec/batch
Epoch: 11/20... Training Step: 10543... Training loss: 1.3678... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10544... Training loss: 1.3456... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10545... Training loss: 1.2974... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10546... Training loss: 1.3070... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10547... Training loss: 1.3065... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10548... Training loss: 1.3346... 0.0819 sec/batch
Epoch: 11/20... Training Step: 10549... Training loss: 1.1567... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10550... Training loss: 1.3368... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10551... Training loss: 1.3286... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10552... Training loss: 1.2902... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10553... Training loss: 1.2870... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10554... Training loss: 1.3188... 0.0812 sec/batch
Epoch: 11/20... Training Step: 10555... Training loss: 1.3441... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10556... Training loss: 1.3091... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10557... Training loss: 1.3857... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10558... Training loss: 1.2882... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10559... Training loss: 1.3351... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10560... Training loss: 1.4183... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10561... Training loss: 1.4216... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10562... Training loss: 1.3991... 0.0790 sec/batch
Epoch: 11/20... Training Step: 10563... Training loss: 1.3113... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10564... Training loss: 1.3483... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10565... Training loss: 1.2969... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10566... Training loss: 1.3390... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10567... Training loss: 1.3688... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10568... Training loss: 1.2660... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10569... Training loss: 1.2899... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10570... Training loss: 1.3195... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10571... Training loss: 1.2966... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10572... Training loss: 1.3674... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10573... Training loss: 1.3457... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10574... Training loss: 1.3933... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10575... Training loss: 1.3928... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10576... Training loss: 1.3684... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10577... Training loss: 1.2663... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10578... Training loss: 1.2987... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10579... Training loss: 1.3014... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10580... Training loss: 1.4095... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10581... Training loss: 1.3519... 0.0865 sec/batch
Epoch: 11/20... Training Step: 10582... Training loss: 1.4206... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10583... Training loss: 1.3354... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10584... Training loss: 1.4285... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10585... Training loss: 1.4317... 0.0741 sec/batch
Epoch: 11/20... Training Step: 10586... Training loss: 1.2940... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10587... Training loss: 1.4409... 0.0809 sec/batch
Epoch: 11/20... Training Step: 10588... Training loss: 1.3800... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10589... Training loss: 1.3525... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10590... Training loss: 1.3096... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10591... Training loss: 1.3665... 0.0818 sec/batch
Epoch: 11/20... Training Step: 10592... Training loss: 1.4077... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10593... Training loss: 1.3298... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10594... Training loss: 1.3310... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10595... Training loss: 1.2820... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10596... Training loss: 1.2933... 0.0785 sec/batch
Epoch: 11/20... Training Step: 10597... Training loss: 1.4150... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10598... Training loss: 1.3115... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10599... Training loss: 1.2580... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10600... Training loss: 1.3725... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10601... Training loss: 1.3966... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10602... Training loss: 1.3816... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10603... Training loss: 1.3635... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10604... Training loss: 1.3209... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10605... Training loss: 1.3926... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10606... Training loss: 1.3849... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10607... Training loss: 1.3581... 0.0795 sec/batch
Epoch: 11/20... Training Step: 10608... Training loss: 1.3553... 0.0792 sec/batch
Epoch: 11/20... Training Step: 10609... Training loss: 1.3186... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10610... Training loss: 1.3784... 0.0787 sec/batch
Epoch: 11/20... Training Step: 10611... Training loss: 1.3427... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10612... Training loss: 1.3527... 0.0785 sec/batch
Epoch: 11/20... Training Step: 10613... Training loss: 1.3302... 0.0785 sec/batch
Epoch: 11/20... Training Step: 10614... Training loss: 1.2820... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10615... Training loss: 1.3659... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10616... Training loss: 1.3403... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10617... Training loss: 1.4060... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10618... Training loss: 1.3545... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10619... Training loss: 1.3329... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10620... Training loss: 1.3295... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10621... Training loss: 1.3530... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10622... Training loss: 1.3558... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10623... Training loss: 1.3399... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10624... Training loss: 1.2864... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10625... Training loss: 1.2885... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10626... Training loss: 1.3113... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10627... Training loss: 1.3521... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10628... Training loss: 1.3553... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10629... Training loss: 1.3448... 0.0792 sec/batch
Epoch: 11/20... Training Step: 10630... Training loss: 1.3113... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10631... Training loss: 1.3093... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10632... Training loss: 1.3823... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10633... Training loss: 1.2856... 0.0810 sec/batch
Epoch: 11/20... Training Step: 10634... Training loss: 1.3452... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10635... Training loss: 1.3691... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10636... Training loss: 1.3666... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10637... Training loss: 1.3816... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10638... Training loss: 1.3708... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10639... Training loss: 1.3988... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10640... Training loss: 1.4438... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10641... Training loss: 1.3541... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10642... Training loss: 1.3578... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10643... Training loss: 1.3501... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10644... Training loss: 1.3606... 0.0792 sec/batch
Epoch: 11/20... Training Step: 10645... Training loss: 1.3899... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10646... Training loss: 1.3980... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10647... Training loss: 1.3343... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10648... Training loss: 1.3967... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10649... Training loss: 1.4049... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10650... Training loss: 1.3057... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10651... Training loss: 1.3118... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10652... Training loss: 1.3915... 0.0801 sec/batch
Epoch: 11/20... Training Step: 10653... Training loss: 1.3081... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10654... Training loss: 1.3921... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10655... Training loss: 1.3337... 0.0805 sec/batch
Epoch: 11/20... Training Step: 10656... Training loss: 1.3202... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10657... Training loss: 1.3171... 0.0738 sec/batch
Epoch: 11/20... Training Step: 10658... Training loss: 1.4074... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10659... Training loss: 1.3266... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10660... Training loss: 1.3531... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10661... Training loss: 1.3057... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10662... Training loss: 1.3281... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10663... Training loss: 1.3401... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10664... Training loss: 1.3861... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10665... Training loss: 1.3822... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10666... Training loss: 1.3550... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10667... Training loss: 1.3388... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10668... Training loss: 1.3322... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10669... Training loss: 1.2985... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10670... Training loss: 1.2872... 0.0829 sec/batch
Epoch: 11/20... Training Step: 10671... Training loss: 1.4509... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10672... Training loss: 1.2983... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10673... Training loss: 1.3924... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10674... Training loss: 1.3810... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10675... Training loss: 1.2511... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10676... Training loss: 1.3378... 0.0764 sec/batch
Epoch: 11/20... Training Step: 10677... Training loss: 1.3557... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10678... Training loss: 1.3861... 0.0791 sec/batch
Epoch: 11/20... Training Step: 10679... Training loss: 1.3904... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10680... Training loss: 1.4302... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10681... Training loss: 1.3569... 0.0793 sec/batch
Epoch: 11/20... Training Step: 10682... Training loss: 1.2862... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10683... Training loss: 1.3650... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10684... Training loss: 1.3517... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10685... Training loss: 1.3803... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10686... Training loss: 1.3872... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10687... Training loss: 1.3479... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10688... Training loss: 1.3133... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10689... Training loss: 1.3156... 0.0726 sec/batch
Epoch: 11/20... Training Step: 10690... Training loss: 1.3541... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10691... Training loss: 1.3531... 0.0793 sec/batch
Epoch: 11/20... Training Step: 10692... Training loss: 1.4091... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10693... Training loss: 1.4002... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10694... Training loss: 1.3774... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10695... Training loss: 1.3555... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10696... Training loss: 1.2679... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10697... Training loss: 1.3093... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10698... Training loss: 1.3296... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10699... Training loss: 1.3209... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10700... Training loss: 1.3127... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10701... Training loss: 1.3372... 0.0795 sec/batch
Epoch: 11/20... Training Step: 10702... Training loss: 1.3248... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10703... Training loss: 1.3838... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10704... Training loss: 1.3673... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10705... Training loss: 1.3509... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10706... Training loss: 1.3214... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10707... Training loss: 1.3338... 0.0803 sec/batch
Epoch: 11/20... Training Step: 10708... Training loss: 1.3570... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10709... Training loss: 1.3445... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10710... Training loss: 1.3755... 0.0882 sec/batch
Epoch: 11/20... Training Step: 10711... Training loss: 1.3259... 0.0821 sec/batch
Epoch: 11/20... Training Step: 10712... Training loss: 1.3192... 0.0857 sec/batch
Epoch: 11/20... Training Step: 10713... Training loss: 1.3331... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10714... Training loss: 1.3125... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10715... Training loss: 1.2564... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10716... Training loss: 1.3101... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10717... Training loss: 1.2978... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10718... Training loss: 1.3129... 0.0823 sec/batch
Epoch: 11/20... Training Step: 10719... Training loss: 1.3174... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10720... Training loss: 1.2981... 0.0862 sec/batch
Epoch: 11/20... Training Step: 10721... Training loss: 1.3616... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10722... Training loss: 1.4059... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10723... Training loss: 1.4050... 0.0805 sec/batch
Epoch: 11/20... Training Step: 10724... Training loss: 1.3388... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10725... Training loss: 1.2867... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10726... Training loss: 1.3344... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10727... Training loss: 1.3159... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10728... Training loss: 1.3571... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10729... Training loss: 1.3846... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10730... Training loss: 1.4032... 0.0785 sec/batch
Epoch: 11/20... Training Step: 10731... Training loss: 1.3401... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10732... Training loss: 1.4095... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10733... Training loss: 1.3501... 0.0807 sec/batch
Epoch: 11/20... Training Step: 10734... Training loss: 1.4049... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10735... Training loss: 1.3958... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10736... Training loss: 1.3764... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10737... Training loss: 1.4098... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10738... Training loss: 1.4438... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10739... Training loss: 1.3716... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10740... Training loss: 1.3416... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10741... Training loss: 1.3300... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10742... Training loss: 1.2998... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10743... Training loss: 1.3521... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10744... Training loss: 1.3446... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10745... Training loss: 1.3546... 0.0769 sec/batch
Epoch: 11/20... Training Step: 10746... Training loss: 1.3636... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10747... Training loss: 1.3727... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10748... Training loss: 1.3960... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10749... Training loss: 1.3643... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10750... Training loss: 1.3265... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10751... Training loss: 1.3925... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10752... Training loss: 1.3813... 0.0760 sec/batch
Epoch: 11/20... Training Step: 10753... Training loss: 1.4236... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10754... Training loss: 1.3631... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10755... Training loss: 1.2775... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10756... Training loss: 1.3197... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10757... Training loss: 1.3746... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10758... Training loss: 1.3491... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10759... Training loss: 1.3545... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10760... Training loss: 1.3568... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10761... Training loss: 1.3640... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10762... Training loss: 1.4811... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10763... Training loss: 1.3746... 0.0787 sec/batch
Epoch: 11/20... Training Step: 10764... Training loss: 1.3936... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10765... Training loss: 1.3999... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10766... Training loss: 1.2693... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10767... Training loss: 1.3617... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10768... Training loss: 1.3472... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10769... Training loss: 1.3478... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10770... Training loss: 1.3763... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10771... Training loss: 1.3065... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10772... Training loss: 1.4228... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10773... Training loss: 1.4320... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10774... Training loss: 1.3515... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10775... Training loss: 1.3621... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10776... Training loss: 1.3586... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10777... Training loss: 1.3588... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10778... Training loss: 1.3325... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10779... Training loss: 1.3725... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10780... Training loss: 1.3988... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10781... Training loss: 1.3582... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10782... Training loss: 1.3112... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10783... Training loss: 1.3032... 0.0802 sec/batch
Epoch: 11/20... Training Step: 10784... Training loss: 1.3491... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10785... Training loss: 1.3028... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10786... Training loss: 1.3527... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10787... Training loss: 1.3619... 0.0845 sec/batch
Epoch: 11/20... Training Step: 10788... Training loss: 1.3920... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10789... Training loss: 1.3611... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10790... Training loss: 1.4328... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10791... Training loss: 1.3810... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10792... Training loss: 1.3771... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10793... Training loss: 1.3684... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10794... Training loss: 1.3698... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10795... Training loss: 1.3521... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10796... Training loss: 1.3770... 0.0785 sec/batch
Epoch: 11/20... Training Step: 10797... Training loss: 1.3750... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10798... Training loss: 1.3945... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10799... Training loss: 1.3580... 0.0789 sec/batch
Epoch: 11/20... Training Step: 10800... Training loss: 1.3339... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10801... Training loss: 1.3325... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10802... Training loss: 1.2819... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10803... Training loss: 1.3707... 0.0738 sec/batch
Epoch: 11/20... Training Step: 10804... Training loss: 1.3393... 0.0783 sec/batch
Epoch: 11/20... Training Step: 10805... Training loss: 1.3578... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10806... Training loss: 1.4103... 0.0771 sec/batch
Epoch: 11/20... Training Step: 10807... Training loss: 1.3490... 0.0757 sec/batch
Epoch: 11/20... Training Step: 10808... Training loss: 1.3786... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10809... Training loss: 1.3618... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10810... Training loss: 1.4432... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10811... Training loss: 1.4083... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10812... Training loss: 1.2800... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10813... Training loss: 1.3310... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10814... Training loss: 1.2653... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10815... Training loss: 1.4035... 0.0768 sec/batch
Epoch: 11/20... Training Step: 10816... Training loss: 1.2924... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10817... Training loss: 1.3792... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10818... Training loss: 1.3817... 0.0787 sec/batch
Epoch: 11/20... Training Step: 10819... Training loss: 1.3144... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10820... Training loss: 1.2904... 0.0788 sec/batch
Epoch: 11/20... Training Step: 10821... Training loss: 1.3300... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10822... Training loss: 1.3133... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10823... Training loss: 1.2994... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10824... Training loss: 1.2782... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10825... Training loss: 1.2712... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10826... Training loss: 1.3636... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10827... Training loss: 1.2818... 0.0808 sec/batch
Epoch: 11/20... Training Step: 10828... Training loss: 1.3485... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10829... Training loss: 1.2979... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10830... Training loss: 1.2325... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10831... Training loss: 1.2830... 0.0828 sec/batch
Epoch: 11/20... Training Step: 10832... Training loss: 1.3646... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10833... Training loss: 1.4283... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10834... Training loss: 1.3326... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10835... Training loss: 1.3418... 0.0740 sec/batch
Epoch: 11/20... Training Step: 10836... Training loss: 1.3353... 0.0774 sec/batch
Epoch: 11/20... Training Step: 10837... Training loss: 1.3492... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10838... Training loss: 1.3580... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10839... Training loss: 1.3503... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10840... Training loss: 1.3853... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10841... Training loss: 1.4579... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10842... Training loss: 1.3906... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10843... Training loss: 1.3812... 0.0763 sec/batch
Epoch: 11/20... Training Step: 10844... Training loss: 1.3138... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10845... Training loss: 1.3527... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10846... Training loss: 1.3987... 0.0745 sec/batch
Epoch: 11/20... Training Step: 10847... Training loss: 1.3474... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10848... Training loss: 1.3465... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10849... Training loss: 1.4279... 0.0755 sec/batch
Epoch: 11/20... Training Step: 10850... Training loss: 1.4043... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10851... Training loss: 1.3720... 0.0758 sec/batch
Epoch: 11/20... Training Step: 10852... Training loss: 1.4370... 0.0841 sec/batch
Epoch: 11/20... Training Step: 10853... Training loss: 1.3628... 0.0786 sec/batch
Epoch: 11/20... Training Step: 10854... Training loss: 1.3703... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10855... Training loss: 1.3709... 0.0776 sec/batch
Epoch: 11/20... Training Step: 10856... Training loss: 1.3870... 0.0756 sec/batch
Epoch: 11/20... Training Step: 10857... Training loss: 1.3167... 0.0778 sec/batch
Epoch: 11/20... Training Step: 10858... Training loss: 1.3488... 0.0748 sec/batch
Epoch: 11/20... Training Step: 10859... Training loss: 1.3587... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10860... Training loss: 1.4346... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10861... Training loss: 1.3562... 0.0770 sec/batch
Epoch: 11/20... Training Step: 10862... Training loss: 1.3763... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10863... Training loss: 1.3987... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10864... Training loss: 1.3880... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10865... Training loss: 1.4081... 0.0753 sec/batch
Epoch: 11/20... Training Step: 10866... Training loss: 1.3338... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10867... Training loss: 1.3337... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10868... Training loss: 1.3201... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10869... Training loss: 1.3469... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10870... Training loss: 1.2769... 0.0777 sec/batch
Epoch: 11/20... Training Step: 10871... Training loss: 1.3432... 0.0742 sec/batch
Epoch: 11/20... Training Step: 10872... Training loss: 1.2949... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10873... Training loss: 1.3277... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10874... Training loss: 1.3763... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10875... Training loss: 1.3765... 0.0744 sec/batch
Epoch: 11/20... Training Step: 10876... Training loss: 1.3754... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10877... Training loss: 1.3280... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10878... Training loss: 1.3692... 0.0737 sec/batch
Epoch: 11/20... Training Step: 10879... Training loss: 1.3086... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10880... Training loss: 1.3002... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10881... Training loss: 1.3554... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10882... Training loss: 1.3367... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10883... Training loss: 1.3576... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10884... Training loss: 1.3299... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10885... Training loss: 1.4182... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10886... Training loss: 1.4023... 0.0749 sec/batch
Epoch: 11/20... Training Step: 10887... Training loss: 1.4097... 0.0773 sec/batch
Epoch: 11/20... Training Step: 10888... Training loss: 1.4164... 0.0751 sec/batch
Epoch: 11/20... Training Step: 10889... Training loss: 1.3842... 0.0746 sec/batch
Epoch: 11/20... Training Step: 10890... Training loss: 1.5368... 0.0772 sec/batch
Epoch: 11/20... Training Step: 10891... Training loss: 1.3261... 0.0766 sec/batch
Epoch: 11/20... Training Step: 10892... Training loss: 1.4509... 0.0775 sec/batch
Epoch: 11/20... Training Step: 10893... Training loss: 1.4627... 0.0781 sec/batch
Epoch: 11/20... Training Step: 10894... Training loss: 1.3147... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10895... Training loss: 1.3590... 0.0752 sec/batch
Epoch: 11/20... Training Step: 10896... Training loss: 1.4118... 0.0765 sec/batch
Epoch: 11/20... Training Step: 10897... Training loss: 1.4253... 0.0782 sec/batch
Epoch: 11/20... Training Step: 10898... Training loss: 1.3903... 0.0747 sec/batch
Epoch: 11/20... Training Step: 10899... Training loss: 1.3860... 0.0759 sec/batch
Epoch: 11/20... Training Step: 10900... Training loss: 1.3140... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10901... Training loss: 1.3092... 0.0762 sec/batch
Epoch: 11/20... Training Step: 10902... Training loss: 1.4509... 0.0779 sec/batch
Epoch: 11/20... Training Step: 10903... Training loss: 1.3765... 0.0750 sec/batch
Epoch: 11/20... Training Step: 10904... Training loss: 1.3652... 0.0780 sec/batch
Epoch: 11/20... Training Step: 10905... Training loss: 1.3702... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10906... Training loss: 1.3297... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10907... Training loss: 1.3067... 0.0784 sec/batch
Epoch: 11/20... Training Step: 10908... Training loss: 1.3400... 0.0767 sec/batch
Epoch: 11/20... Training Step: 10909... Training loss: 1.3673... 0.0761 sec/batch
Epoch: 11/20... Training Step: 10910... Training loss: 1.2347... 0.0743 sec/batch
Epoch: 11/20... Training Step: 10911... Training loss: 1.3242... 0.0754 sec/batch
Epoch: 11/20... Training Step: 10912... Training loss: 1.3894... 0.0766 sec/batch
Epoch: 12/20... Training Step: 10913... Training loss: 1.4450... 0.0776 sec/batch
Epoch: 12/20... Training Step: 10914... Training loss: 1.3256... 0.0768 sec/batch
Epoch: 12/20... Training Step: 10915... Training loss: 1.4099... 0.0752 sec/batch
Epoch: 12/20... Training Step: 10916... Training loss: 1.3642... 0.0753 sec/batch
Epoch: 12/20... Training Step: 10917... Training loss: 1.2833... 0.0753 sec/batch
Epoch: 12/20... Training Step: 10918... Training loss: 1.2265... 0.0767 sec/batch
Epoch: 12/20... Training Step: 10919... Training loss: 1.3298... 0.0787 sec/batch
Epoch: 12/20... Training Step: 10920... Training loss: 1.3707... 0.0778 sec/batch
Epoch: 12/20... Training Step: 10921... Training loss: 1.4109... 0.0790 sec/batch
Epoch: 12/20... Training Step: 10922... Training loss: 1.2731... 0.0765 sec/batch
Epoch: 12/20... Training Step: 10923... Training loss: 1.4154... 0.0748 sec/batch
Epoch: 12/20... Training Step: 10924... Training loss: 1.2905... 0.0755 sec/batch
Epoch: 12/20... Training Step: 10925... Training loss: 1.3214... 0.0754 sec/batch
Epoch: 12/20... Training Step: 10926... Training loss: 1.3407... 0.0789 sec/batch
Epoch: 12/20... Training Step: 10927... Training loss: 1.3788... 0.0774 sec/batch
Epoch: 12/20... Training Step: 10928... Training loss: 1.2910... 0.0768 sec/batch
Epoch: 12/20... Training Step: 10929... Training loss: 1.3405... 0.0755 sec/batch
Epoch: 12/20... Training Step: 10930... Training loss: 1.3424... 0.0760 sec/batch
Epoch: 12/20... Training Step: 10931... Training loss: 1.3314... 0.0786 sec/batch
Epoch: 12/20... Training Step: 10932... Training loss: 1.3119... 0.0760 sec/batch
Epoch: 12/20... Training Step: 10933... Training loss: 1.3301... 0.0769 sec/batch
Epoch: 12/20... Training Step: 10934... Training loss: 1.2989... 0.0761 sec/batch
Epoch: 12/20... Training Step: 10935... Training loss: 1.3106... 0.0778 sec/batch
Epoch: 12/20... Training Step: 10936... Training loss: 1.2633... 0.0760 sec/batch
Epoch: 12/20... Training Step: 10937... Training loss: 1.3669... 0.0757 sec/batch
Epoch: 12/20... Training Step: 10938... Training loss: 1.3013... 0.0746 sec/batch
Epoch: 12/20... Training Step: 10939... Training loss: 1.3387... 0.0766 sec/batch
Epoch: 12/20... Training Step: 10940... Training loss: 1.2763... 0.0779 sec/batch
Epoch: 12/20... Training Step: 10941... Training loss: 1.3305... 0.0754 sec/batch
Epoch: 12/20... Training Step: 10942... Training loss: 1.3873... 0.0771 sec/batch
Epoch: 12/20... Training Step: 10943... Training loss: 1.3691... 0.0770 sec/batch
Epoch: 12/20... Training Step: 10944... Training loss: 1.3887... 0.0754 sec/batch
Epoch: 12/20... Training Step: 10945... Training loss: 1.3451... 0.0780 sec/batch
Epoch: 12/20... Training Step: 10946... Training loss: 1.3942... 0.0749 sec/batch
Epoch: 12/20... Training Step: 10947... Training loss: 1.4211... 0.0747 sec/batch
Epoch: 12/20... Training Step: 10948... Training loss: 1.4017... 0.0748 sec/batch
Epoch: 12/20... Training Step: 10949... Training loss: 1.3243... 0.0772 sec/batch
Epoch: 12/20... Training Step: 10950... Training loss: 1.4026... 0.0762 sec/batch
Epoch: 12/20... Training Step: 10951... Training loss: 1.3765... 0.0766 sec/batch
Epoch: 12/20... Training Step: 10952... Training loss: 1.3342... 0.0759 sec/batch
Epoch: 12/20... Training Step: 10953... Training loss: 1.3534... 0.0746 sec/batch
Epoch: 12/20... Training Step: 10954... Training loss: 1.3375... 0.0770 sec/batch
Epoch: 12/20... Training Step: 10955... Training loss: 1.2951... 0.0761 sec/batch
Epoch: 12/20... Training Step: 10956... Training loss: 1.3229... 0.0769 sec/batch
Epoch: 12/20... Training Step: 10957... Training loss: 1.3712... 0.0804 sec/batch
Epoch: 12/20... Training Step: 10958... Training loss: 1.3230... 0.0750 sec/batch
Epoch: 12/20... Training Step: 10959... Training loss: 1.3871... 0.0772 sec/batch
Epoch: 12/20... Training Step: 10960... Training loss: 1.2981... 0.0778 sec/batch
Epoch: 12/20... Training Step: 10961... Training loss: 1.3932... 0.0789 sec/batch
Epoch: 12/20... Training Step: 10962... Training loss: 1.3704... 0.0767 sec/batch
Epoch: 12/20... Training Step: 10963... Training loss: 1.2711... 0.0774 sec/batch
Epoch: 12/20... Training Step: 10964... Training loss: 1.3917... 0.0766 sec/batch
Epoch: 12/20... Training Step: 10965... Training loss: 1.2738... 0.0740 sec/batch
Epoch: 12/20... Training Step: 10966... Training loss: 1.2865... 0.0765 sec/batch
Epoch: 12/20... Training Step: 10967... Training loss: 1.3318... 0.0768 sec/batch
Epoch: 12/20... Training Step: 10968... Training loss: 1.2892... 0.0757 sec/batch
Epoch: 12/20... Training Step: 10969... Training loss: 1.3449... 0.0758 sec/batch
Epoch: 12/20... Training Step: 10970... Training loss: 1.4098... 0.0775 sec/batch
Epoch: 12/20... Training Step: 10971... Training loss: 1.3409... 0.0790 sec/batch
Epoch: 12/20... Training Step: 10972... Training loss: 1.3548... 0.0749 sec/batch
Epoch: 12/20... Training Step: 10973... Training loss: 1.2458... 0.0762 sec/batch
Epoch: 12/20... Training Step: 10974... Training loss: 1.2806... 0.0759 sec/batch
Epoch: 12/20... Training Step: 10975... Training loss: 1.3377... 0.0760 sec/batch
Epoch: 12/20... Training Step: 10976... Training loss: 1.2799... 0.0765 sec/batch
Epoch: 12/20... Training Step: 10977... Training loss: 1.3466... 0.0746 sec/batch
Epoch: 12/20... Training Step: 10978... Training loss: 1.3105... 0.0758 sec/batch
Epoch: 12/20... Training Step: 10979... Training loss: 1.2979... 0.0760 sec/batch
Epoch: 12/20... Training Step: 10980... Training loss: 1.3145... 0.0775 sec/batch
Epoch: 12/20... Training Step: 10981... Training loss: 1.3696... 0.0784 sec/batch
Epoch: 12/20... Training Step: 10982... Training loss: 1.3431... 0.0761 sec/batch
Epoch: 12/20... Training Step: 10983... Training loss: 1.3496... 0.0779 sec/batch
Epoch: 12/20... Training Step: 10984... Training loss: 1.3923... 0.0746 sec/batch
Epoch: 12/20... Training Step: 10985... Training loss: 1.3525... 0.0756 sec/batch
Epoch: 12/20... Training Step: 10986... Training loss: 1.2826... 0.0755 sec/batch
Epoch: 12/20... Training Step: 10987... Training loss: 1.2953... 0.0753 sec/batch
Epoch: 12/20... Training Step: 10988... Training loss: 1.2693... 0.0763 sec/batch
Epoch: 12/20... Training Step: 10989... Training loss: 1.3047... 0.0756 sec/batch
Epoch: 12/20... Training Step: 10990... Training loss: 1.3120... 0.0772 sec/batch
Epoch: 12/20... Training Step: 10991... Training loss: 1.3220... 0.0745 sec/batch
Epoch: 12/20... Training Step: 10992... Training loss: 1.3812... 0.0763 sec/batch
Epoch: 12/20... Training Step: 10993... Training loss: 1.3519... 0.0761 sec/batch
Epoch: 12/20... Training Step: 10994... Training loss: 1.3106... 0.0756 sec/batch
Epoch: 12/20... Training Step: 10995... Training loss: 1.3639... 0.0765 sec/batch
Epoch: 12/20... Training Step: 10996... Training loss: 1.3566... 0.0785 sec/batch
Epoch: 12/20... Training Step: 10997... Training loss: 1.3068... 0.0753 sec/batch
Epoch: 12/20... Training Step: 10998... Training loss: 1.3871... 0.0741 sec/batch
Epoch: 12/20... Training Step: 10999... Training loss: 1.3161... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11000... Training loss: 1.3408... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11001... Training loss: 1.3985... 0.0740 sec/batch
Epoch: 12/20... Training Step: 11002... Training loss: 1.3969... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11003... Training loss: 1.3943... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11004... Training loss: 1.2909... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11005... Training loss: 1.2745... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11006... Training loss: 1.2539... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11007... Training loss: 1.2593... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11008... Training loss: 1.3557... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11009... Training loss: 1.3051... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11010... Training loss: 1.2875... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11011... Training loss: 1.3017... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11012... Training loss: 1.2940... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11013... Training loss: 1.2846... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11014... Training loss: 1.3008... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11015... Training loss: 1.3493... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11016... Training loss: 1.3732... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11017... Training loss: 1.3464... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11018... Training loss: 1.3339... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11019... Training loss: 1.2791... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11020... Training loss: 1.3633... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11021... Training loss: 1.3104... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11022... Training loss: 1.3758... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11023... Training loss: 1.3627... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11024... Training loss: 1.3217... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11025... Training loss: 1.4065... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11026... Training loss: 1.3682... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11027... Training loss: 1.2895... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11028... Training loss: 1.4475... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11029... Training loss: 1.3664... 0.0738 sec/batch
Epoch: 12/20... Training Step: 11030... Training loss: 1.4347... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11031... Training loss: 1.3479... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11032... Training loss: 1.3422... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11033... Training loss: 1.4649... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11034... Training loss: 1.3128... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11035... Training loss: 1.3705... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11036... Training loss: 1.3409... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11037... Training loss: 1.3949... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11038... Training loss: 1.4046... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11039... Training loss: 1.3782... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11040... Training loss: 1.3525... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11041... Training loss: 1.2859... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11042... Training loss: 1.3663... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11043... Training loss: 1.3175... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11044... Training loss: 1.3633... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11045... Training loss: 1.3794... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11046... Training loss: 1.3399... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11047... Training loss: 1.3773... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11048... Training loss: 1.3849... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11049... Training loss: 1.3852... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11050... Training loss: 1.4141... 0.0785 sec/batch
Epoch: 12/20... Training Step: 11051... Training loss: 1.4065... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11052... Training loss: 1.3333... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11053... Training loss: 1.3622... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11054... Training loss: 1.3162... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11055... Training loss: 1.3347... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11056... Training loss: 1.3276... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11057... Training loss: 1.2091... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11058... Training loss: 1.3942... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11059... Training loss: 1.3067... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11060... Training loss: 1.2814... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11061... Training loss: 1.3003... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11062... Training loss: 1.3030... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11063... Training loss: 1.2192... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11064... Training loss: 1.3291... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11065... Training loss: 1.2787... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11066... Training loss: 1.3298... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11067... Training loss: 1.3195... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11068... Training loss: 1.2938... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11069... Training loss: 1.3197... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11070... Training loss: 1.3523... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11071... Training loss: 1.3487... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11072... Training loss: 1.3751... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11073... Training loss: 1.3045... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11074... Training loss: 1.3293... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11075... Training loss: 1.3249... 0.0736 sec/batch
Epoch: 12/20... Training Step: 11076... Training loss: 1.3966... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11077... Training loss: 1.3691... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11078... Training loss: 1.4124... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11079... Training loss: 1.3620... 0.0850 sec/batch
Epoch: 12/20... Training Step: 11080... Training loss: 1.3408... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11081... Training loss: 1.3548... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11082... Training loss: 1.3497... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11083... Training loss: 1.3401... 0.0739 sec/batch
Epoch: 12/20... Training Step: 11084... Training loss: 1.3415... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11085... Training loss: 1.3418... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11086... Training loss: 1.3679... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11087... Training loss: 1.3032... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11088... Training loss: 1.3719... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11089... Training loss: 1.3708... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11090... Training loss: 1.4498... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11091... Training loss: 1.3805... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11092... Training loss: 1.3186... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11093... Training loss: 1.3487... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11094... Training loss: 1.3941... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11095... Training loss: 1.3302... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11096... Training loss: 1.3617... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11097... Training loss: 1.3779... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11098... Training loss: 1.3839... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11099... Training loss: 1.3628... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11100... Training loss: 1.3294... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11101... Training loss: 1.3387... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11102... Training loss: 1.3838... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11103... Training loss: 1.4123... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11104... Training loss: 1.3671... 0.0795 sec/batch
Epoch: 12/20... Training Step: 11105... Training loss: 1.4043... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11106... Training loss: 1.4320... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11107... Training loss: 1.3329... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11108... Training loss: 1.3181... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11109... Training loss: 1.4197... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11110... Training loss: 1.3615... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11111... Training loss: 1.3625... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11112... Training loss: 1.3308... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11113... Training loss: 1.3263... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11114... Training loss: 1.3473... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11115... Training loss: 1.3646... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11116... Training loss: 1.3516... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11117... Training loss: 1.3005... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11118... Training loss: 1.3860... 0.0840 sec/batch
Epoch: 12/20... Training Step: 11119... Training loss: 1.3319... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11120... Training loss: 1.3138... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11121... Training loss: 1.3730... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11122... Training loss: 1.3487... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11123... Training loss: 1.3095... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11124... Training loss: 1.3907... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11125... Training loss: 1.3063... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11126... Training loss: 1.3785... 0.0739 sec/batch
Epoch: 12/20... Training Step: 11127... Training loss: 1.3276... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11128... Training loss: 1.3424... 0.0787 sec/batch
Epoch: 12/20... Training Step: 11129... Training loss: 1.3022... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11130... Training loss: 1.3551... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11131... Training loss: 1.3534... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11132... Training loss: 1.2866... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11133... Training loss: 1.2887... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11134... Training loss: 1.4497... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11135... Training loss: 1.3335... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11136... Training loss: 1.3491... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11137... Training loss: 1.3898... 0.0787 sec/batch
Epoch: 12/20... Training Step: 11138... Training loss: 1.3147... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11139... Training loss: 1.3548... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11140... Training loss: 1.3743... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11141... Training loss: 1.2655... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11142... Training loss: 1.2905... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11143... Training loss: 1.3360... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11144... Training loss: 1.3456... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11145... Training loss: 1.3475... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11146... Training loss: 1.4092... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11147... Training loss: 1.3398... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11148... Training loss: 1.3297... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11149... Training loss: 1.3246... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11150... Training loss: 1.3703... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11151... Training loss: 1.3845... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11152... Training loss: 1.3404... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11153... Training loss: 1.3307... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11154... Training loss: 1.3395... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11155... Training loss: 1.3750... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11156... Training loss: 1.3843... 0.0791 sec/batch
Epoch: 12/20... Training Step: 11157... Training loss: 1.3173... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11158... Training loss: 1.3846... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11159... Training loss: 1.2797... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11160... Training loss: 1.2642... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11161... Training loss: 1.3485... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11162... Training loss: 1.3485... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11163... Training loss: 1.3290... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11164... Training loss: 1.3313... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11165... Training loss: 1.3183... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11166... Training loss: 1.2961... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11167... Training loss: 1.2658... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11168... Training loss: 1.3132... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11169... Training loss: 1.4062... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11170... Training loss: 1.3105... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11171... Training loss: 1.3228... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11172... Training loss: 1.3574... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11173... Training loss: 1.3680... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11174... Training loss: 1.3934... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11175... Training loss: 1.3623... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11176... Training loss: 1.3813... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11177... Training loss: 1.3465... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11178... Training loss: 1.2567... 0.0798 sec/batch
Epoch: 12/20... Training Step: 11179... Training loss: 1.3153... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11180... Training loss: 1.3381... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11181... Training loss: 1.2741... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11182... Training loss: 1.2879... 0.0790 sec/batch
Epoch: 12/20... Training Step: 11183... Training loss: 1.3419... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11184... Training loss: 1.3024... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11185... Training loss: 1.2279... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11186... Training loss: 1.2807... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11187... Training loss: 1.3160... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11188... Training loss: 1.3170... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11189... Training loss: 1.3558... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11190... Training loss: 1.2485... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11191... Training loss: 1.3029... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11192... Training loss: 1.3750... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11193... Training loss: 1.3244... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11194... Training loss: 1.2974... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11195... Training loss: 1.3556... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11196... Training loss: 1.2886... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11197... Training loss: 1.2042... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11198... Training loss: 1.2898... 0.0795 sec/batch
Epoch: 12/20... Training Step: 11199... Training loss: 1.3668... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11200... Training loss: 1.3020... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11201... Training loss: 1.3182... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11202... Training loss: 1.2927... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11203... Training loss: 1.2910... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11204... Training loss: 1.3285... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11205... Training loss: 1.2723... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11206... Training loss: 1.2613... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11207... Training loss: 1.3472... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11208... Training loss: 1.3746... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11209... Training loss: 1.3201... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11210... Training loss: 1.4055... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11211... Training loss: 1.3329... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11212... Training loss: 1.3522... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11213... Training loss: 1.3960... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11214... Training loss: 1.2800... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11215... Training loss: 1.2823... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11216... Training loss: 1.3668... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11217... Training loss: 1.3320... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11218... Training loss: 1.3891... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11219... Training loss: 1.2969... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11220... Training loss: 1.2774... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11221... Training loss: 1.3981... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11222... Training loss: 1.3122... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11223... Training loss: 1.4069... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11224... Training loss: 1.3426... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11225... Training loss: 1.3246... 0.0791 sec/batch
Epoch: 12/20... Training Step: 11226... Training loss: 1.3215... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11227... Training loss: 1.3475... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11228... Training loss: 1.2874... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11229... Training loss: 1.3081... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11230... Training loss: 1.2679... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11231... Training loss: 1.2941... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11232... Training loss: 1.3523... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11233... Training loss: 1.3598... 0.0739 sec/batch
Epoch: 12/20... Training Step: 11234... Training loss: 1.2658... 0.0832 sec/batch
Epoch: 12/20... Training Step: 11235... Training loss: 1.2996... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11236... Training loss: 1.3299... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11237... Training loss: 1.3631... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11238... Training loss: 1.3222... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11239... Training loss: 1.3170... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11240... Training loss: 1.3324... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11241... Training loss: 1.3130... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11242... Training loss: 1.3440... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11243... Training loss: 1.3853... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11244... Training loss: 1.2619... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11245... Training loss: 1.4273... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11246... Training loss: 1.4126... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11247... Training loss: 1.3978... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11248... Training loss: 1.3705... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11249... Training loss: 1.3223... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11250... Training loss: 1.3285... 0.0835 sec/batch
Epoch: 12/20... Training Step: 11251... Training loss: 1.3207... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11252... Training loss: 1.2838... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11253... Training loss: 1.3374... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11254... Training loss: 1.2978... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11255... Training loss: 1.2913... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11256... Training loss: 1.3587... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11257... Training loss: 1.3173... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11258... Training loss: 1.3451... 0.0820 sec/batch
Epoch: 12/20... Training Step: 11259... Training loss: 1.3656... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11260... Training loss: 1.3999... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11261... Training loss: 1.3827... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11262... Training loss: 1.4112... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11263... Training loss: 1.3517... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11264... Training loss: 1.3088... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11265... Training loss: 1.3065... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11266... Training loss: 1.3259... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11267... Training loss: 1.3305... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11268... Training loss: 1.3395... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11269... Training loss: 1.3201... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11270... Training loss: 1.3341... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11271... Training loss: 1.3443... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11272... Training loss: 1.3298... 0.0739 sec/batch
Epoch: 12/20... Training Step: 11273... Training loss: 1.3450... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11274... Training loss: 1.4132... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11275... Training loss: 1.3325... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11276... Training loss: 1.3327... 0.0808 sec/batch
Epoch: 12/20... Training Step: 11277... Training loss: 1.3419... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11278... Training loss: 1.3029... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11279... Training loss: 1.3184... 0.0805 sec/batch
Epoch: 12/20... Training Step: 11280... Training loss: 1.2582... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11281... Training loss: 1.2781... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11282... Training loss: 1.2669... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11283... Training loss: 1.3754... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11284... Training loss: 1.2817... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11285... Training loss: 1.3444... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11286... Training loss: 1.2728... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11287... Training loss: 1.3133... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11288... Training loss: 1.2794... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11289... Training loss: 1.3154... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11290... Training loss: 1.3453... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11291... Training loss: 1.2975... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11292... Training loss: 1.3659... 0.0788 sec/batch
Epoch: 12/20... Training Step: 11293... Training loss: 1.3285... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11294... Training loss: 1.3168... 0.0820 sec/batch
Epoch: 12/20... Training Step: 11295... Training loss: 1.3762... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11296... Training loss: 1.3473... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11297... Training loss: 1.3705... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11298... Training loss: 1.3475... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11299... Training loss: 1.3716... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11300... Training loss: 1.3915... 0.0836 sec/batch
Epoch: 12/20... Training Step: 11301... Training loss: 1.3783... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11302... Training loss: 1.3221... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11303... Training loss: 1.3596... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11304... Training loss: 1.3275... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11305... Training loss: 1.3332... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11306... Training loss: 1.3079... 0.0785 sec/batch
Epoch: 12/20... Training Step: 11307... Training loss: 1.3236... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11308... Training loss: 1.3216... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11309... Training loss: 1.3092... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11310... Training loss: 1.3745... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11311... Training loss: 1.3541... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11312... Training loss: 1.3243... 0.0796 sec/batch
Epoch: 12/20... Training Step: 11313... Training loss: 1.2998... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11314... Training loss: 1.3569... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11315... Training loss: 1.2898... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11316... Training loss: 1.3646... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11317... Training loss: 1.3080... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11318... Training loss: 1.3216... 0.0785 sec/batch
Epoch: 12/20... Training Step: 11319... Training loss: 1.2725... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11320... Training loss: 1.3839... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11321... Training loss: 1.3096... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11322... Training loss: 1.2924... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11323... Training loss: 1.2276... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11324... Training loss: 1.2833... 0.0813 sec/batch
Epoch: 12/20... Training Step: 11325... Training loss: 1.3292... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11326... Training loss: 1.2352... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11327... Training loss: 1.2884... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11328... Training loss: 1.3752... 0.0802 sec/batch
Epoch: 12/20... Training Step: 11329... Training loss: 1.2639... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11330... Training loss: 1.2935... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11331... Training loss: 1.3495... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11332... Training loss: 1.4232... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11333... Training loss: 1.2611... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11334... Training loss: 1.3565... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11335... Training loss: 1.3294... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11336... Training loss: 1.3523... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11337... Training loss: 1.3168... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11338... Training loss: 1.3321... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11339... Training loss: 1.4029... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11340... Training loss: 1.4028... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11341... Training loss: 1.4033... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11342... Training loss: 1.4369... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11343... Training loss: 1.3394... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11344... Training loss: 1.3548... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11345... Training loss: 1.3668... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11346... Training loss: 1.3578... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11347... Training loss: 1.3733... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11348... Training loss: 1.2947... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11349... Training loss: 1.3717... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11350... Training loss: 1.4421... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11351... Training loss: 1.3546... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11352... Training loss: 1.3671... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11353... Training loss: 1.3160... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11354... Training loss: 1.3674... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11355... Training loss: 1.3624... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11356... Training loss: 1.3554... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11357... Training loss: 1.3281... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11358... Training loss: 1.4401... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11359... Training loss: 1.3489... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11360... Training loss: 1.3408... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11361... Training loss: 1.4027... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11362... Training loss: 1.4257... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11363... Training loss: 1.3309... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11364... Training loss: 1.3526... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11365... Training loss: 1.3824... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11366... Training loss: 1.3676... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11367... Training loss: 1.3612... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11368... Training loss: 1.3071... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11369... Training loss: 1.3187... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11370... Training loss: 1.3460... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11371... Training loss: 1.3267... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11372... Training loss: 1.3470... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11373... Training loss: 1.3600... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11374... Training loss: 1.3228... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11375... Training loss: 1.3030... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11376... Training loss: 1.3639... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11377... Training loss: 1.2904... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11378... Training loss: 1.2946... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11379... Training loss: 1.2452... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11380... Training loss: 1.3427... 0.0801 sec/batch
Epoch: 12/20... Training Step: 11381... Training loss: 1.3121... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11382... Training loss: 1.4085... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11383... Training loss: 1.3273... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11384... Training loss: 1.3437... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11385... Training loss: 1.3256... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11386... Training loss: 1.3305... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11387... Training loss: 1.2308... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11388... Training loss: 1.3116... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11389... Training loss: 1.3329... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11390... Training loss: 1.3338... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11391... Training loss: 1.2623... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11392... Training loss: 1.2950... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11393... Training loss: 1.2823... 0.0796 sec/batch
Epoch: 12/20... Training Step: 11394... Training loss: 1.3169... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11395... Training loss: 1.3344... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11396... Training loss: 1.3355... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11397... Training loss: 1.2469... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11398... Training loss: 1.3122... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11399... Training loss: 1.3681... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11400... Training loss: 1.2933... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11401... Training loss: 1.2929... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11402... Training loss: 1.2824... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11403... Training loss: 1.3055... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11404... Training loss: 1.3085... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11405... Training loss: 1.3270... 0.0791 sec/batch
Epoch: 12/20... Training Step: 11406... Training loss: 1.3257... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11407... Training loss: 1.2963... 0.0834 sec/batch
Epoch: 12/20... Training Step: 11408... Training loss: 1.3019... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11409... Training loss: 1.3336... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11410... Training loss: 1.3074... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11411... Training loss: 1.2969... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11412... Training loss: 1.3405... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11413... Training loss: 1.2957... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11414... Training loss: 1.3285... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11415... Training loss: 1.3456... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11416... Training loss: 1.3009... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11417... Training loss: 1.3219... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11418... Training loss: 1.2782... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11419... Training loss: 1.3048... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11420... Training loss: 1.2710... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11421... Training loss: 1.2857... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11422... Training loss: 1.2850... 0.0790 sec/batch
Epoch: 12/20... Training Step: 11423... Training loss: 1.3435... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11424... Training loss: 1.3462... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11425... Training loss: 1.3595... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11426... Training loss: 1.3748... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11427... Training loss: 1.3326... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11428... Training loss: 1.3392... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11429... Training loss: 1.3363... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11430... Training loss: 1.3229... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11431... Training loss: 1.3439... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11432... Training loss: 1.2915... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11433... Training loss: 1.3311... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11434... Training loss: 1.2773... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11435... Training loss: 1.4314... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11436... Training loss: 1.2810... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11437... Training loss: 1.3303... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11438... Training loss: 1.3145... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11439... Training loss: 1.3031... 0.0788 sec/batch
Epoch: 12/20... Training Step: 11440... Training loss: 1.3123... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11441... Training loss: 1.3324... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11442... Training loss: 1.3838... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11443... Training loss: 1.4316... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11444... Training loss: 1.3507... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11445... Training loss: 1.3841... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11446... Training loss: 1.3738... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11447... Training loss: 1.3213... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11448... Training loss: 1.2908... 0.0855 sec/batch
Epoch: 12/20... Training Step: 11449... Training loss: 1.3176... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11450... Training loss: 1.3045... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11451... Training loss: 1.3221... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11452... Training loss: 1.3115... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11453... Training loss: 1.4293... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11454... Training loss: 1.2988... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11455... Training loss: 1.3068... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11456... Training loss: 1.3049... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11457... Training loss: 1.2721... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11458... Training loss: 1.3130... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11459... Training loss: 1.3986... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11460... Training loss: 1.3435... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11461... Training loss: 1.2751... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11462... Training loss: 1.3078... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11463... Training loss: 1.3287... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11464... Training loss: 1.2652... 0.0787 sec/batch
Epoch: 12/20... Training Step: 11465... Training loss: 1.2657... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11466... Training loss: 1.2770... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11467... Training loss: 1.3533... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11468... Training loss: 1.2884... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11469... Training loss: 1.2630... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11470... Training loss: 1.3067... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11471... Training loss: 1.3280... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11472... Training loss: 1.2475... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11473... Training loss: 1.2793... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11474... Training loss: 1.3232... 0.0796 sec/batch
Epoch: 12/20... Training Step: 11475... Training loss: 1.2740... 0.0814 sec/batch
Epoch: 12/20... Training Step: 11476... Training loss: 1.3104... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11477... Training loss: 1.3385... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11478... Training loss: 1.3480... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11479... Training loss: 1.2939... 0.0813 sec/batch
Epoch: 12/20... Training Step: 11480... Training loss: 1.2727... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11481... Training loss: 1.2295... 0.0853 sec/batch
Epoch: 12/20... Training Step: 11482... Training loss: 1.2544... 0.0788 sec/batch
Epoch: 12/20... Training Step: 11483... Training loss: 1.2693... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11484... Training loss: 1.2961... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11485... Training loss: 1.3789... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11486... Training loss: 1.3339... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11487... Training loss: 1.3553... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11488... Training loss: 1.3520... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11489... Training loss: 1.3065... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11490... Training loss: 1.4372... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11491... Training loss: 1.3632... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11492... Training loss: 1.2952... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11493... Training loss: 1.2652... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11494... Training loss: 1.2831... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11495... Training loss: 1.3707... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11496... Training loss: 1.2536... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11497... Training loss: 1.3066... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11498... Training loss: 1.3612... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11499... Training loss: 1.3333... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11500... Training loss: 1.3597... 0.0797 sec/batch
Epoch: 12/20... Training Step: 11501... Training loss: 1.2990... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11502... Training loss: 1.3019... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11503... Training loss: 1.3481... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11504... Training loss: 1.3599... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11505... Training loss: 1.2800... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11506... Training loss: 1.2888... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11507... Training loss: 1.2476... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11508... Training loss: 1.3442... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11509... Training loss: 1.2064... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11510... Training loss: 1.2206... 0.0740 sec/batch
Epoch: 12/20... Training Step: 11511... Training loss: 1.2838... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11512... Training loss: 1.2952... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11513... Training loss: 1.3504... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11514... Training loss: 1.3075... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11515... Training loss: 1.2799... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11516... Training loss: 1.3242... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11517... Training loss: 1.2843... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11518... Training loss: 1.3035... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11519... Training loss: 1.3749... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11520... Training loss: 1.3331... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11521... Training loss: 1.3147... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11522... Training loss: 1.3539... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11523... Training loss: 1.2306... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11524... Training loss: 1.3349... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11525... Training loss: 1.3517... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11526... Training loss: 1.2995... 0.0816 sec/batch
Epoch: 12/20... Training Step: 11527... Training loss: 1.3509... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11528... Training loss: 1.3303... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11529... Training loss: 1.3329... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11530... Training loss: 1.3631... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11531... Training loss: 1.2950... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11532... Training loss: 1.3140... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11533... Training loss: 1.2919... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11534... Training loss: 1.2648... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11535... Training loss: 1.3243... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11536... Training loss: 1.3146... 0.0834 sec/batch
Epoch: 12/20... Training Step: 11537... Training loss: 1.2673... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11538... Training loss: 1.2871... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11539... Training loss: 1.2892... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11540... Training loss: 1.3353... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11541... Training loss: 1.1318... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11542... Training loss: 1.3309... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11543... Training loss: 1.3204... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11544... Training loss: 1.2919... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11545... Training loss: 1.2553... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11546... Training loss: 1.3076... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11547... Training loss: 1.3443... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11548... Training loss: 1.2915... 0.0830 sec/batch
Epoch: 12/20... Training Step: 11549... Training loss: 1.3858... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11550... Training loss: 1.3159... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11551... Training loss: 1.3377... 0.0790 sec/batch
Epoch: 12/20... Training Step: 11552... Training loss: 1.4055... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11553... Training loss: 1.4076... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11554... Training loss: 1.3775... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11555... Training loss: 1.3119... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11556... Training loss: 1.3288... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11557... Training loss: 1.2923... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11558... Training loss: 1.3442... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11559... Training loss: 1.3456... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11560... Training loss: 1.2569... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11561... Training loss: 1.2781... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11562... Training loss: 1.3239... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11563... Training loss: 1.3141... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11564... Training loss: 1.3562... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11565... Training loss: 1.3043... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11566... Training loss: 1.3748... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11567... Training loss: 1.3697... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11568... Training loss: 1.3385... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11569... Training loss: 1.2475... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11570... Training loss: 1.2966... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11571... Training loss: 1.2832... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11572... Training loss: 1.4013... 0.0825 sec/batch
Epoch: 12/20... Training Step: 11573... Training loss: 1.3552... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11574... Training loss: 1.4099... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11575... Training loss: 1.3336... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11576... Training loss: 1.4171... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11577... Training loss: 1.4235... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11578... Training loss: 1.3038... 0.0816 sec/batch
Epoch: 12/20... Training Step: 11579... Training loss: 1.4618... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11580... Training loss: 1.3621... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11581... Training loss: 1.3346... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11582... Training loss: 1.3163... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11583... Training loss: 1.3489... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11584... Training loss: 1.3881... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11585... Training loss: 1.3293... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11586... Training loss: 1.3374... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11587... Training loss: 1.2543... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11588... Training loss: 1.3028... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11589... Training loss: 1.4060... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11590... Training loss: 1.3039... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11591... Training loss: 1.2618... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11592... Training loss: 1.3706... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11593... Training loss: 1.3858... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11594... Training loss: 1.3627... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11595... Training loss: 1.3619... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11596... Training loss: 1.3267... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11597... Training loss: 1.3846... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11598... Training loss: 1.3926... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11599... Training loss: 1.3156... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11600... Training loss: 1.3290... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11601... Training loss: 1.3118... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11602... Training loss: 1.3719... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11603... Training loss: 1.3407... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11604... Training loss: 1.3250... 0.0793 sec/batch
Epoch: 12/20... Training Step: 11605... Training loss: 1.3327... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11606... Training loss: 1.2629... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11607... Training loss: 1.3433... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11608... Training loss: 1.3193... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11609... Training loss: 1.4045... 0.0733 sec/batch
Epoch: 12/20... Training Step: 11610... Training loss: 1.3373... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11611... Training loss: 1.3217... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11612... Training loss: 1.3401... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11613... Training loss: 1.3503... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11614... Training loss: 1.3313... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11615... Training loss: 1.3377... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11616... Training loss: 1.2501... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11617... Training loss: 1.2531... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11618... Training loss: 1.2771... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11619... Training loss: 1.3399... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11620... Training loss: 1.3394... 0.0792 sec/batch
Epoch: 12/20... Training Step: 11621... Training loss: 1.3462... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11622... Training loss: 1.2967... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11623... Training loss: 1.3016... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11624... Training loss: 1.3812... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11625... Training loss: 1.2830... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11626... Training loss: 1.3227... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11627... Training loss: 1.3452... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11628... Training loss: 1.3487... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11629... Training loss: 1.3686... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11630... Training loss: 1.3734... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11631... Training loss: 1.3626... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11632... Training loss: 1.4486... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11633... Training loss: 1.3599... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11634... Training loss: 1.3451... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11635... Training loss: 1.3348... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11636... Training loss: 1.3616... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11637... Training loss: 1.3886... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11638... Training loss: 1.3943... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11639... Training loss: 1.3326... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11640... Training loss: 1.3779... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11641... Training loss: 1.3837... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11642... Training loss: 1.3130... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11643... Training loss: 1.3251... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11644... Training loss: 1.3589... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11645... Training loss: 1.3041... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11646... Training loss: 1.3845... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11647... Training loss: 1.3319... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11648... Training loss: 1.3302... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11649... Training loss: 1.2980... 0.0825 sec/batch
Epoch: 12/20... Training Step: 11650... Training loss: 1.4249... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11651... Training loss: 1.3019... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11652... Training loss: 1.3330... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11653... Training loss: 1.2951... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11654... Training loss: 1.3110... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11655... Training loss: 1.3112... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11656... Training loss: 1.3811... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11657... Training loss: 1.3686... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11658... Training loss: 1.3289... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11659... Training loss: 1.3211... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11660... Training loss: 1.3350... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11661... Training loss: 1.2651... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11662... Training loss: 1.2736... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11663... Training loss: 1.4231... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11664... Training loss: 1.3108... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11665... Training loss: 1.3828... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11666... Training loss: 1.3652... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11667... Training loss: 1.2514... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11668... Training loss: 1.3607... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11669... Training loss: 1.3465... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11670... Training loss: 1.3935... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11671... Training loss: 1.3789... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11672... Training loss: 1.4021... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11673... Training loss: 1.3432... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11674... Training loss: 1.2884... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11675... Training loss: 1.3474... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11676... Training loss: 1.3285... 0.0791 sec/batch
Epoch: 12/20... Training Step: 11677... Training loss: 1.3573... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11678... Training loss: 1.3761... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11679... Training loss: 1.3329... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11680... Training loss: 1.3009... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11681... Training loss: 1.3122... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11682... Training loss: 1.3616... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11683... Training loss: 1.3430... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11684... Training loss: 1.3940... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11685... Training loss: 1.3817... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11686... Training loss: 1.3572... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11687... Training loss: 1.3252... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11688... Training loss: 1.2635... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11689... Training loss: 1.2888... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11690... Training loss: 1.3347... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11691... Training loss: 1.2905... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11692... Training loss: 1.3018... 0.0738 sec/batch
Epoch: 12/20... Training Step: 11693... Training loss: 1.3175... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11694... Training loss: 1.3324... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11695... Training loss: 1.3752... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11696... Training loss: 1.3502... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11697... Training loss: 1.3494... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11698... Training loss: 1.3438... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11699... Training loss: 1.3347... 0.0805 sec/batch
Epoch: 12/20... Training Step: 11700... Training loss: 1.3535... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11701... Training loss: 1.3317... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11702... Training loss: 1.3451... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11703... Training loss: 1.3365... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11704... Training loss: 1.3150... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11705... Training loss: 1.3300... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11706... Training loss: 1.2958... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11707... Training loss: 1.2451... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11708... Training loss: 1.3114... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11709... Training loss: 1.2899... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11710... Training loss: 1.3278... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11711... Training loss: 1.2936... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11712... Training loss: 1.3022... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11713... Training loss: 1.3448... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11714... Training loss: 1.3812... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11715... Training loss: 1.3835... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11716... Training loss: 1.3380... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11717... Training loss: 1.2846... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11718... Training loss: 1.3273... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11719... Training loss: 1.3122... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11720... Training loss: 1.3740... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11721... Training loss: 1.3875... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11722... Training loss: 1.3741... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11723... Training loss: 1.3270... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11724... Training loss: 1.3912... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11725... Training loss: 1.3490... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11726... Training loss: 1.3856... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11727... Training loss: 1.3681... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11728... Training loss: 1.3772... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11729... Training loss: 1.3896... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11730... Training loss: 1.4533... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11731... Training loss: 1.3598... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11732... Training loss: 1.3484... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11733... Training loss: 1.3182... 0.0837 sec/batch
Epoch: 12/20... Training Step: 11734... Training loss: 1.2677... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11735... Training loss: 1.3511... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11736... Training loss: 1.3144... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11737... Training loss: 1.3335... 0.0743 sec/batch
Epoch: 12/20... Training Step: 11738... Training loss: 1.3448... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11739... Training loss: 1.3713... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11740... Training loss: 1.3944... 0.0778 sec/batch
Epoch: 12/20... Training Step: 11741... Training loss: 1.3658... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11742... Training loss: 1.2878... 0.0751 sec/batch
Epoch: 12/20... Training Step: 11743... Training loss: 1.3744... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11744... Training loss: 1.3548... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11745... Training loss: 1.4062... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11746... Training loss: 1.3482... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11747... Training loss: 1.2768... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11748... Training loss: 1.3157... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11749... Training loss: 1.3528... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11750... Training loss: 1.3589... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11751... Training loss: 1.3607... 0.0805 sec/batch
Epoch: 12/20... Training Step: 11752... Training loss: 1.3547... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11753... Training loss: 1.3648... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11754... Training loss: 1.4903... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11755... Training loss: 1.3826... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11756... Training loss: 1.3924... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11757... Training loss: 1.3820... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11758... Training loss: 1.2838... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11759... Training loss: 1.3426... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11760... Training loss: 1.3144... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11761... Training loss: 1.3536... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11762... Training loss: 1.3700... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11763... Training loss: 1.2973... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11764... Training loss: 1.4311... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11765... Training loss: 1.4243... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11766... Training loss: 1.3456... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11767... Training loss: 1.3338... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11768... Training loss: 1.3461... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11769... Training loss: 1.3632... 0.0840 sec/batch
Epoch: 12/20... Training Step: 11770... Training loss: 1.3349... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11771... Training loss: 1.3370... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11772... Training loss: 1.3865... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11773... Training loss: 1.3472... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11774... Training loss: 1.2896... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11775... Training loss: 1.2893... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11776... Training loss: 1.3223... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11777... Training loss: 1.2855... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11778... Training loss: 1.3238... 0.0793 sec/batch
Epoch: 12/20... Training Step: 11779... Training loss: 1.3549... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11780... Training loss: 1.3784... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11781... Training loss: 1.3418... 0.0789 sec/batch
Epoch: 12/20... Training Step: 11782... Training loss: 1.4223... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11783... Training loss: 1.3847... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11784... Training loss: 1.3486... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11785... Training loss: 1.3555... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11786... Training loss: 1.3427... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11787... Training loss: 1.3341... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11788... Training loss: 1.3802... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11789... Training loss: 1.3363... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11790... Training loss: 1.3960... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11791... Training loss: 1.3511... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11792... Training loss: 1.3499... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11793... Training loss: 1.3546... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11794... Training loss: 1.2740... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11795... Training loss: 1.3546... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11796... Training loss: 1.3355... 0.0794 sec/batch
Epoch: 12/20... Training Step: 11797... Training loss: 1.3386... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11798... Training loss: 1.3909... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11799... Training loss: 1.3462... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11800... Training loss: 1.3868... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11801... Training loss: 1.3680... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11802... Training loss: 1.4189... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11803... Training loss: 1.4117... 0.0746 sec/batch
Epoch: 12/20... Training Step: 11804... Training loss: 1.2612... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11805... Training loss: 1.3145... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11806... Training loss: 1.2774... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11807... Training loss: 1.3786... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11808... Training loss: 1.2772... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11809... Training loss: 1.3659... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11810... Training loss: 1.3591... 0.0749 sec/batch
Epoch: 12/20... Training Step: 11811... Training loss: 1.2877... 0.0740 sec/batch
Epoch: 12/20... Training Step: 11812... Training loss: 1.2831... 0.0741 sec/batch
Epoch: 12/20... Training Step: 11813... Training loss: 1.2983... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11814... Training loss: 1.3322... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11815... Training loss: 1.3001... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11816... Training loss: 1.2746... 0.0783 sec/batch
Epoch: 12/20... Training Step: 11817... Training loss: 1.2733... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11818... Training loss: 1.3296... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11819... Training loss: 1.2671... 0.0782 sec/batch
Epoch: 12/20... Training Step: 11820... Training loss: 1.3290... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11821... Training loss: 1.2738... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11822... Training loss: 1.2473... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11823... Training loss: 1.2700... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11824... Training loss: 1.3466... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11825... Training loss: 1.4291... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11826... Training loss: 1.3433... 0.0760 sec/batch
Epoch: 12/20... Training Step: 11827... Training loss: 1.3295... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11828... Training loss: 1.3137... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11829... Training loss: 1.3351... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11830... Training loss: 1.3533... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11831... Training loss: 1.3428... 0.0813 sec/batch
Epoch: 12/20... Training Step: 11832... Training loss: 1.3791... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11833... Training loss: 1.4486... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11834... Training loss: 1.3677... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11835... Training loss: 1.3343... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11836... Training loss: 1.3147... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11837... Training loss: 1.3493... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11838... Training loss: 1.3752... 0.0744 sec/batch
Epoch: 12/20... Training Step: 11839... Training loss: 1.3260... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11840... Training loss: 1.3316... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11841... Training loss: 1.4271... 0.0787 sec/batch
Epoch: 12/20... Training Step: 11842... Training loss: 1.3952... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11843... Training loss: 1.3607... 0.0768 sec/batch
Epoch: 12/20... Training Step: 11844... Training loss: 1.4295... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11845... Training loss: 1.3148... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11846... Training loss: 1.3585... 0.0739 sec/batch
Epoch: 12/20... Training Step: 11847... Training loss: 1.3654... 0.0774 sec/batch
Epoch: 12/20... Training Step: 11848... Training loss: 1.3893... 0.0757 sec/batch
Epoch: 12/20... Training Step: 11849... Training loss: 1.3154... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11850... Training loss: 1.3374... 0.0779 sec/batch
Epoch: 12/20... Training Step: 11851... Training loss: 1.3289... 0.0758 sec/batch
Epoch: 12/20... Training Step: 11852... Training loss: 1.4143... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11853... Training loss: 1.3529... 0.0801 sec/batch
Epoch: 12/20... Training Step: 11854... Training loss: 1.3701... 0.0796 sec/batch
Epoch: 12/20... Training Step: 11855... Training loss: 1.4049... 0.0787 sec/batch
Epoch: 12/20... Training Step: 11856... Training loss: 1.3747... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11857... Training loss: 1.3924... 0.0771 sec/batch
Epoch: 12/20... Training Step: 11858... Training loss: 1.3172... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11859... Training loss: 1.3278... 0.0780 sec/batch
Epoch: 12/20... Training Step: 11860... Training loss: 1.3131... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11861... Training loss: 1.3539... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11862... Training loss: 1.2699... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11863... Training loss: 1.3394... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11864... Training loss: 1.3044... 0.0797 sec/batch
Epoch: 12/20... Training Step: 11865... Training loss: 1.3104... 0.0762 sec/batch
Epoch: 12/20... Training Step: 11866... Training loss: 1.3819... 0.0765 sec/batch
Epoch: 12/20... Training Step: 11867... Training loss: 1.3554... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11868... Training loss: 1.3595... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11869... Training loss: 1.3446... 0.0781 sec/batch
Epoch: 12/20... Training Step: 11870... Training loss: 1.3671... 0.0755 sec/batch
Epoch: 12/20... Training Step: 11871... Training loss: 1.2820... 0.0809 sec/batch
Epoch: 12/20... Training Step: 11872... Training loss: 1.3084... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11873... Training loss: 1.3465... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11874... Training loss: 1.3056... 0.0763 sec/batch
Epoch: 12/20... Training Step: 11875... Training loss: 1.3540... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11876... Training loss: 1.3337... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11877... Training loss: 1.3827... 0.0742 sec/batch
Epoch: 12/20... Training Step: 11878... Training loss: 1.3763... 0.0747 sec/batch
Epoch: 12/20... Training Step: 11879... Training loss: 1.3997... 0.0759 sec/batch
Epoch: 12/20... Training Step: 11880... Training loss: 1.3691... 0.0750 sec/batch
Epoch: 12/20... Training Step: 11881... Training loss: 1.4099... 0.0764 sec/batch
Epoch: 12/20... Training Step: 11882... Training loss: 1.5299... 0.0756 sec/batch
Epoch: 12/20... Training Step: 11883... Training loss: 1.2982... 0.0769 sec/batch
Epoch: 12/20... Training Step: 11884... Training loss: 1.4470... 0.0745 sec/batch
Epoch: 12/20... Training Step: 11885... Training loss: 1.4418... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11886... Training loss: 1.3122... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11887... Training loss: 1.3843... 0.0775 sec/batch
Epoch: 12/20... Training Step: 11888... Training loss: 1.4058... 0.0770 sec/batch
Epoch: 12/20... Training Step: 11889... Training loss: 1.4104... 0.0767 sec/batch
Epoch: 12/20... Training Step: 11890... Training loss: 1.3815... 0.0784 sec/batch
Epoch: 12/20... Training Step: 11891... Training loss: 1.3897... 0.0785 sec/batch
Epoch: 12/20... Training Step: 11892... Training loss: 1.2849... 0.0753 sec/batch
Epoch: 12/20... Training Step: 11893... Training loss: 1.2711... 0.0766 sec/batch
Epoch: 12/20... Training Step: 11894... Training loss: 1.4565... 0.0790 sec/batch
Epoch: 12/20... Training Step: 11895... Training loss: 1.3443... 0.0777 sec/batch
Epoch: 12/20... Training Step: 11896... Training loss: 1.3574... 0.0761 sec/batch
Epoch: 12/20... Training Step: 11897... Training loss: 1.3380... 0.0773 sec/batch
Epoch: 12/20... Training Step: 11898... Training loss: 1.3122... 0.0754 sec/batch
Epoch: 12/20... Training Step: 11899... Training loss: 1.2930... 0.0772 sec/batch
Epoch: 12/20... Training Step: 11900... Training loss: 1.3321... 0.0752 sec/batch
Epoch: 12/20... Training Step: 11901... Training loss: 1.3506... 0.0786 sec/batch
Epoch: 12/20... Training Step: 11902... Training loss: 1.2275... 0.0748 sec/batch
Epoch: 12/20... Training Step: 11903... Training loss: 1.3163... 0.0776 sec/batch
Epoch: 12/20... Training Step: 11904... Training loss: 1.4000... 0.0757 sec/batch
Epoch: 13/20... Training Step: 11905... Training loss: 1.4124... 0.0747 sec/batch
Epoch: 13/20... Training Step: 11906... Training loss: 1.2908... 0.0780 sec/batch
Epoch: 13/20... Training Step: 11907... Training loss: 1.4155... 0.0762 sec/batch
Epoch: 13/20... Training Step: 11908... Training loss: 1.3517... 0.0745 sec/batch
Epoch: 13/20... Training Step: 11909... Training loss: 1.2445... 0.0753 sec/batch
Epoch: 13/20... Training Step: 11910... Training loss: 1.2024... 0.0775 sec/batch
Epoch: 13/20... Training Step: 11911... Training loss: 1.3040... 0.0778 sec/batch
Epoch: 13/20... Training Step: 11912... Training loss: 1.3603... 0.0765 sec/batch
Epoch: 13/20... Training Step: 11913... Training loss: 1.3611... 0.0762 sec/batch
Epoch: 13/20... Training Step: 11914... Training loss: 1.2671... 0.0758 sec/batch
Epoch: 13/20... Training Step: 11915... Training loss: 1.3922... 0.0832 sec/batch
Epoch: 13/20... Training Step: 11916... Training loss: 1.2885... 0.0768 sec/batch
Epoch: 13/20... Training Step: 11917... Training loss: 1.3112... 0.0748 sec/batch
Epoch: 13/20... Training Step: 11918... Training loss: 1.3352... 0.0757 sec/batch
Epoch: 13/20... Training Step: 11919... Training loss: 1.3608... 0.0744 sec/batch
Epoch: 13/20... Training Step: 11920... Training loss: 1.2811... 0.0741 sec/batch
Epoch: 13/20... Training Step: 11921... Training loss: 1.3199... 0.0751 sec/batch
Epoch: 13/20... Training Step: 11922... Training loss: 1.3490... 0.0768 sec/batch
Epoch: 13/20... Training Step: 11923... Training loss: 1.3114... 0.0766 sec/batch
Epoch: 13/20... Training Step: 11924... Training loss: 1.3089... 0.0828 sec/batch
Epoch: 13/20... Training Step: 11925... Training loss: 1.3105... 0.0759 sec/batch
Epoch: 13/20... Training Step: 11926... Training loss: 1.2928... 0.0752 sec/batch
Epoch: 13/20... Training Step: 11927... Training loss: 1.3125... 0.0749 sec/batch
Epoch: 13/20... Training Step: 11928... Training loss: 1.2545... 0.0772 sec/batch
Epoch: 13/20... Training Step: 11929... Training loss: 1.3506... 0.0752 sec/batch
Epoch: 13/20... Training Step: 11930... Training loss: 1.2942... 0.0761 sec/batch
Epoch: 13/20... Training Step: 11931... Training loss: 1.3085... 0.0772 sec/batch
Epoch: 13/20... Training Step: 11932... Training loss: 1.2958... 0.0745 sec/batch
Epoch: 13/20... Training Step: 11933... Training loss: 1.3321... 0.0766 sec/batch
Epoch: 13/20... Training Step: 11934... Training loss: 1.3564... 0.0771 sec/batch
Epoch: 13/20... Training Step: 11935... Training loss: 1.3774... 0.0759 sec/batch
Epoch: 13/20... Training Step: 11936... Training loss: 1.3720... 0.0764 sec/batch
Epoch: 13/20... Training Step: 11937... Training loss: 1.3245... 0.0748 sec/batch
Epoch: 13/20... Training Step: 11938... Training loss: 1.3715... 0.0756 sec/batch
Epoch: 13/20... Training Step: 11939... Training loss: 1.4121... 0.0774 sec/batch
Epoch: 13/20... Training Step: 11940... Training loss: 1.3869... 0.0780 sec/batch
Epoch: 13/20... Training Step: 11941... Training loss: 1.3106... 0.0756 sec/batch
Epoch: 13/20... Training Step: 11942... Training loss: 1.4016... 0.0772 sec/batch
Epoch: 13/20... Training Step: 11943... Training loss: 1.3413... 0.0761 sec/batch
Epoch: 13/20... Training Step: 11944... Training loss: 1.3292... 0.0754 sec/batch
Epoch: 13/20... Training Step: 11945... Training loss: 1.3382... 0.0752 sec/batch
Epoch: 13/20... Training Step: 11946... Training loss: 1.3127... 0.0745 sec/batch
Epoch: 13/20... Training Step: 11947... Training loss: 1.3071... 0.0750 sec/batch
Epoch: 13/20... Training Step: 11948... Training loss: 1.2984... 0.0753 sec/batch
Epoch: 13/20... Training Step: 11949... Training loss: 1.3479... 0.0784 sec/batch
Epoch: 13/20... Training Step: 11950... Training loss: 1.3149... 0.0771 sec/batch
Epoch: 13/20... Training Step: 11951... Training loss: 1.3608... 0.0772 sec/batch
Epoch: 13/20... Training Step: 11952... Training loss: 1.2953... 0.0768 sec/batch
Epoch: 13/20... Training Step: 11953... Training loss: 1.3601... 0.0754 sec/batch
Epoch: 13/20... Training Step: 11954... Training loss: 1.3489... 0.0765 sec/batch
Epoch: 13/20... Training Step: 11955... Training loss: 1.2409... 0.0749 sec/batch
Epoch: 13/20... Training Step: 11956... Training loss: 1.3598... 0.0770 sec/batch
Epoch: 13/20... Training Step: 11957... Training loss: 1.2667... 0.0808 sec/batch
Epoch: 13/20... Training Step: 11958... Training loss: 1.2729... 0.0760 sec/batch
Epoch: 13/20... Training Step: 11959... Training loss: 1.3203... 0.0743 sec/batch
Epoch: 13/20... Training Step: 11960... Training loss: 1.2733... 0.0789 sec/batch
Epoch: 13/20... Training Step: 11961... Training loss: 1.3323... 0.0773 sec/batch
Epoch: 13/20... Training Step: 11962... Training loss: 1.3684... 0.0758 sec/batch
Epoch: 13/20... Training Step: 11963... Training loss: 1.3395... 0.0771 sec/batch
Epoch: 13/20... Training Step: 11964... Training loss: 1.3609... 0.0782 sec/batch
Epoch: 13/20... Training Step: 11965... Training loss: 1.2411... 0.0773 sec/batch
Epoch: 13/20... Training Step: 11966... Training loss: 1.2508... 0.0794 sec/batch
Epoch: 13/20... Training Step: 11967... Training loss: 1.3297... 0.0764 sec/batch
Epoch: 13/20... Training Step: 11968... Training loss: 1.2612... 0.0778 sec/batch
Epoch: 13/20... Training Step: 11969... Training loss: 1.3474... 0.0768 sec/batch
Epoch: 13/20... Training Step: 11970... Training loss: 1.3125... 0.0747 sec/batch
Epoch: 13/20... Training Step: 11971... Training loss: 1.2915... 0.0753 sec/batch
Epoch: 13/20... Training Step: 11972... Training loss: 1.3109... 0.0781 sec/batch
Epoch: 13/20... Training Step: 11973... Training loss: 1.3717... 0.0766 sec/batch
Epoch: 13/20... Training Step: 11974... Training loss: 1.3275... 0.0765 sec/batch
Epoch: 13/20... Training Step: 11975... Training loss: 1.3503... 0.0771 sec/batch
Epoch: 13/20... Training Step: 11976... Training loss: 1.3983... 0.0772 sec/batch
Epoch: 13/20... Training Step: 11977... Training loss: 1.3414... 0.0747 sec/batch
Epoch: 13/20... Training Step: 11978... Training loss: 1.2892... 0.0823 sec/batch
Epoch: 13/20... Training Step: 11979... Training loss: 1.2972... 0.0768 sec/batch
Epoch: 13/20... Training Step: 11980... Training loss: 1.2669... 0.0755 sec/batch
Epoch: 13/20... Training Step: 11981... Training loss: 1.3119... 0.0836 sec/batch
Epoch: 13/20... Training Step: 11982... Training loss: 1.3226... 0.0769 sec/batch
Epoch: 13/20... Training Step: 11983... Training loss: 1.3277... 0.0752 sec/batch
Epoch: 13/20... Training Step: 11984... Training loss: 1.3791... 0.0760 sec/batch
Epoch: 13/20... Training Step: 11985... Training loss: 1.3396... 0.0752 sec/batch
Epoch: 13/20... Training Step: 11986... Training loss: 1.2899... 0.0769 sec/batch
Epoch: 13/20... Training Step: 11987... Training loss: 1.3341... 0.0776 sec/batch
Epoch: 13/20... Training Step: 11988... Training loss: 1.3411... 0.0755 sec/batch
Epoch: 13/20... Training Step: 11989... Training loss: 1.3123... 0.0765 sec/batch
Epoch: 13/20... Training Step: 11990... Training loss: 1.3698... 0.0778 sec/batch
Epoch: 13/20... Training Step: 11991... Training loss: 1.3160... 0.0767 sec/batch
Epoch: 13/20... Training Step: 11992... Training loss: 1.3163... 0.0744 sec/batch
Epoch: 13/20... Training Step: 11993... Training loss: 1.3653... 0.0754 sec/batch
Epoch: 13/20... Training Step: 11994... Training loss: 1.3720... 0.0750 sec/batch
Epoch: 13/20... Training Step: 11995... Training loss: 1.3703... 0.0748 sec/batch
Epoch: 13/20... Training Step: 11996... Training loss: 1.2996... 0.0771 sec/batch
Epoch: 13/20... Training Step: 11997... Training loss: 1.2757... 0.0749 sec/batch
Epoch: 13/20... Training Step: 11998... Training loss: 1.2366... 0.0746 sec/batch
Epoch: 13/20... Training Step: 11999... Training loss: 1.2447... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12000... Training loss: 1.3410... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12001... Training loss: 1.3075... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12002... Training loss: 1.2778... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12003... Training loss: 1.2868... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12004... Training loss: 1.2843... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12005... Training loss: 1.2760... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12006... Training loss: 1.3221... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12007... Training loss: 1.3221... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12008... Training loss: 1.3619... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12009... Training loss: 1.3378... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12010... Training loss: 1.3247... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12011... Training loss: 1.2712... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12012... Training loss: 1.3688... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12013... Training loss: 1.3127... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12014... Training loss: 1.3771... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12015... Training loss: 1.3224... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12016... Training loss: 1.3031... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12017... Training loss: 1.3963... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12018... Training loss: 1.3444... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12019... Training loss: 1.2757... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12020... Training loss: 1.4568... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12021... Training loss: 1.3840... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12022... Training loss: 1.4337... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12023... Training loss: 1.3314... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12024... Training loss: 1.3190... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12025... Training loss: 1.4432... 0.0798 sec/batch
Epoch: 13/20... Training Step: 12026... Training loss: 1.3036... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12027... Training loss: 1.3583... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12028... Training loss: 1.3464... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12029... Training loss: 1.3958... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12030... Training loss: 1.3805... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12031... Training loss: 1.3698... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12032... Training loss: 1.3400... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12033... Training loss: 1.2699... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12034... Training loss: 1.3630... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12035... Training loss: 1.3033... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12036... Training loss: 1.3675... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12037... Training loss: 1.3210... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12038... Training loss: 1.3403... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12039... Training loss: 1.3872... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12040... Training loss: 1.3458... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12041... Training loss: 1.3793... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12042... Training loss: 1.3867... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12043... Training loss: 1.3856... 0.0795 sec/batch
Epoch: 13/20... Training Step: 12044... Training loss: 1.3430... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12045... Training loss: 1.3613... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12046... Training loss: 1.3041... 0.0797 sec/batch
Epoch: 13/20... Training Step: 12047... Training loss: 1.3030... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12048... Training loss: 1.3188... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12049... Training loss: 1.1931... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12050... Training loss: 1.3621... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12051... Training loss: 1.3266... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12052... Training loss: 1.2536... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12053... Training loss: 1.3147... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12054... Training loss: 1.2880... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12055... Training loss: 1.2101... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12056... Training loss: 1.3306... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12057... Training loss: 1.2533... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12058... Training loss: 1.2815... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12059... Training loss: 1.2964... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12060... Training loss: 1.2742... 0.0738 sec/batch
Epoch: 13/20... Training Step: 12061... Training loss: 1.3124... 0.0795 sec/batch
Epoch: 13/20... Training Step: 12062... Training loss: 1.3271... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12063... Training loss: 1.3359... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12064... Training loss: 1.3494... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12065... Training loss: 1.3142... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12066... Training loss: 1.3289... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12067... Training loss: 1.3258... 0.0786 sec/batch
Epoch: 13/20... Training Step: 12068... Training loss: 1.3929... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12069... Training loss: 1.3437... 0.0808 sec/batch
Epoch: 13/20... Training Step: 12070... Training loss: 1.4234... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12071... Training loss: 1.3513... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12072... Training loss: 1.3267... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12073... Training loss: 1.3531... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12074... Training loss: 1.3231... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12075... Training loss: 1.3363... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12076... Training loss: 1.3241... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12077... Training loss: 1.3258... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12078... Training loss: 1.3814... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12079... Training loss: 1.2802... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12080... Training loss: 1.3609... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12081... Training loss: 1.3476... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12082... Training loss: 1.4456... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12083... Training loss: 1.3814... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12084... Training loss: 1.3128... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12085... Training loss: 1.3400... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12086... Training loss: 1.4032... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12087... Training loss: 1.3115... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12088... Training loss: 1.3449... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12089... Training loss: 1.3578... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12090... Training loss: 1.3882... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12091... Training loss: 1.3151... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12092... Training loss: 1.3121... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12093... Training loss: 1.3335... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12094... Training loss: 1.3972... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12095... Training loss: 1.4055... 0.0826 sec/batch
Epoch: 13/20... Training Step: 12096... Training loss: 1.3626... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12097... Training loss: 1.4029... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12098... Training loss: 1.4100... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12099... Training loss: 1.3486... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12100... Training loss: 1.2630... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12101... Training loss: 1.3942... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12102... Training loss: 1.3380... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12103... Training loss: 1.3504... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12104... Training loss: 1.3228... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12105... Training loss: 1.3360... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12106... Training loss: 1.3509... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12107... Training loss: 1.3188... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12108... Training loss: 1.3109... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12109... Training loss: 1.3085... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12110... Training loss: 1.3850... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12111... Training loss: 1.3297... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12112... Training loss: 1.3285... 0.0801 sec/batch
Epoch: 13/20... Training Step: 12113... Training loss: 1.3686... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12114... Training loss: 1.3426... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12115... Training loss: 1.2881... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12116... Training loss: 1.3778... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12117... Training loss: 1.3140... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12118... Training loss: 1.3763... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12119... Training loss: 1.3292... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12120... Training loss: 1.3296... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12121... Training loss: 1.3040... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12122... Training loss: 1.3496... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12123... Training loss: 1.3274... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12124... Training loss: 1.2845... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12125... Training loss: 1.2969... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12126... Training loss: 1.4305... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12127... Training loss: 1.3289... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12128... Training loss: 1.3680... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12129... Training loss: 1.3939... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12130... Training loss: 1.2942... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12131... Training loss: 1.3502... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12132... Training loss: 1.3682... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12133... Training loss: 1.2523... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12134... Training loss: 1.2868... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12135... Training loss: 1.3292... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12136... Training loss: 1.3152... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12137... Training loss: 1.3221... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12138... Training loss: 1.3956... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12139... Training loss: 1.3122... 0.0831 sec/batch
Epoch: 13/20... Training Step: 12140... Training loss: 1.3241... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12141... Training loss: 1.2981... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12142... Training loss: 1.3667... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12143... Training loss: 1.3570... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12144... Training loss: 1.3014... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12145... Training loss: 1.3423... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12146... Training loss: 1.3339... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12147... Training loss: 1.3616... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12148... Training loss: 1.3733... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12149... Training loss: 1.3051... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12150... Training loss: 1.3589... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12151... Training loss: 1.2825... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12152... Training loss: 1.2481... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12153... Training loss: 1.3229... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12154... Training loss: 1.3386... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12155... Training loss: 1.3212... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12156... Training loss: 1.3106... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12157... Training loss: 1.3054... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12158... Training loss: 1.2769... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12159... Training loss: 1.2604... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12160... Training loss: 1.3057... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12161... Training loss: 1.4178... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12162... Training loss: 1.3118... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12163... Training loss: 1.3162... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12164... Training loss: 1.3322... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12165... Training loss: 1.3536... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12166... Training loss: 1.3980... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12167... Training loss: 1.3580... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12168... Training loss: 1.3549... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12169... Training loss: 1.3551... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12170... Training loss: 1.2415... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12171... Training loss: 1.3049... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12172... Training loss: 1.3394... 0.0798 sec/batch
Epoch: 13/20... Training Step: 12173... Training loss: 1.2695... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12174... Training loss: 1.2775... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12175... Training loss: 1.3348... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12176... Training loss: 1.2831... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12177... Training loss: 1.2484... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12178... Training loss: 1.2701... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12179... Training loss: 1.3035... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12180... Training loss: 1.3169... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12181... Training loss: 1.3316... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12182... Training loss: 1.2125... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12183... Training loss: 1.2775... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12184... Training loss: 1.3948... 0.0827 sec/batch
Epoch: 13/20... Training Step: 12185... Training loss: 1.3085... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12186... Training loss: 1.3059... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12187... Training loss: 1.3474... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12188... Training loss: 1.2768... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12189... Training loss: 1.2013... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12190... Training loss: 1.2645... 0.0827 sec/batch
Epoch: 13/20... Training Step: 12191... Training loss: 1.3488... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12192... Training loss: 1.2984... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12193... Training loss: 1.2946... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12194... Training loss: 1.2825... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12195... Training loss: 1.2931... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12196... Training loss: 1.3310... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12197... Training loss: 1.2482... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12198... Training loss: 1.2807... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12199... Training loss: 1.3299... 0.0866 sec/batch
Epoch: 13/20... Training Step: 12200... Training loss: 1.3391... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12201... Training loss: 1.2738... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12202... Training loss: 1.4055... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12203... Training loss: 1.2982... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12204... Training loss: 1.3488... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12205... Training loss: 1.3934... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12206... Training loss: 1.2682... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12207... Training loss: 1.2784... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12208... Training loss: 1.3420... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12209... Training loss: 1.3198... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12210... Training loss: 1.3949... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12211... Training loss: 1.2754... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12212... Training loss: 1.2472... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12213... Training loss: 1.3879... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12214... Training loss: 1.2964... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12215... Training loss: 1.3864... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12216... Training loss: 1.3186... 0.0738 sec/batch
Epoch: 13/20... Training Step: 12217... Training loss: 1.2975... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12218... Training loss: 1.3193... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12219... Training loss: 1.3353... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12220... Training loss: 1.2763... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12221... Training loss: 1.2966... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12222... Training loss: 1.2568... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12223... Training loss: 1.3291... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12224... Training loss: 1.3495... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12225... Training loss: 1.3327... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12226... Training loss: 1.2545... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12227... Training loss: 1.3056... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12228... Training loss: 1.2894... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12229... Training loss: 1.3759... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12230... Training loss: 1.3030... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12231... Training loss: 1.3236... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12232... Training loss: 1.2894... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12233... Training loss: 1.3205... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12234... Training loss: 1.3381... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12235... Training loss: 1.3934... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12236... Training loss: 1.2653... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12237... Training loss: 1.4262... 0.0836 sec/batch
Epoch: 13/20... Training Step: 12238... Training loss: 1.3989... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12239... Training loss: 1.3964... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12240... Training loss: 1.3566... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12241... Training loss: 1.3183... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12242... Training loss: 1.3092... 0.0786 sec/batch
Epoch: 13/20... Training Step: 12243... Training loss: 1.2924... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12244... Training loss: 1.2943... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12245... Training loss: 1.3162... 0.0740 sec/batch
Epoch: 13/20... Training Step: 12246... Training loss: 1.2798... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12247... Training loss: 1.2699... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12248... Training loss: 1.3399... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12249... Training loss: 1.3061... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12250... Training loss: 1.3232... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12251... Training loss: 1.3438... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12252... Training loss: 1.3824... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12253... Training loss: 1.3783... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12254... Training loss: 1.3812... 0.0793 sec/batch
Epoch: 13/20... Training Step: 12255... Training loss: 1.3445... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12256... Training loss: 1.2972... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12257... Training loss: 1.3246... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12258... Training loss: 1.2885... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12259... Training loss: 1.3297... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12260... Training loss: 1.3470... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12261... Training loss: 1.2839... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12262... Training loss: 1.3098... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12263... Training loss: 1.3350... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12264... Training loss: 1.3076... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12265... Training loss: 1.3634... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12266... Training loss: 1.3779... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12267... Training loss: 1.3104... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12268... Training loss: 1.3307... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12269... Training loss: 1.3428... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12270... Training loss: 1.2901... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12271... Training loss: 1.3235... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12272... Training loss: 1.2439... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12273... Training loss: 1.2948... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12274... Training loss: 1.2764... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12275... Training loss: 1.3808... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12276... Training loss: 1.2641... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12277... Training loss: 1.3131... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12278... Training loss: 1.2360... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12279... Training loss: 1.3099... 0.0791 sec/batch
Epoch: 13/20... Training Step: 12280... Training loss: 1.2657... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12281... Training loss: 1.3187... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12282... Training loss: 1.3288... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12283... Training loss: 1.2882... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12284... Training loss: 1.3278... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12285... Training loss: 1.3215... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12286... Training loss: 1.2894... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12287... Training loss: 1.3598... 0.0793 sec/batch
Epoch: 13/20... Training Step: 12288... Training loss: 1.3336... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12289... Training loss: 1.3427... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12290... Training loss: 1.3326... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12291... Training loss: 1.3542... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12292... Training loss: 1.3465... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12293... Training loss: 1.3646... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12294... Training loss: 1.3076... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12295... Training loss: 1.3581... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12296... Training loss: 1.3169... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12297... Training loss: 1.3301... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12298... Training loss: 1.2872... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12299... Training loss: 1.3048... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12300... Training loss: 1.3051... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12301... Training loss: 1.2862... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12302... Training loss: 1.3748... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12303... Training loss: 1.3400... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12304... Training loss: 1.2863... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12305... Training loss: 1.2822... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12306... Training loss: 1.3403... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12307... Training loss: 1.2726... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12308... Training loss: 1.3774... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12309... Training loss: 1.3180... 0.0786 sec/batch
Epoch: 13/20... Training Step: 12310... Training loss: 1.3217... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12311... Training loss: 1.2511... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12312... Training loss: 1.3856... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12313... Training loss: 1.2703... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12314... Training loss: 1.3029... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12315... Training loss: 1.2434... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12316... Training loss: 1.2786... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12317... Training loss: 1.3173... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12318... Training loss: 1.2450... 0.0839 sec/batch
Epoch: 13/20... Training Step: 12319... Training loss: 1.2649... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12320... Training loss: 1.3540... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12321... Training loss: 1.2435... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12322... Training loss: 1.3014... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12323... Training loss: 1.3331... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12324... Training loss: 1.3962... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12325... Training loss: 1.2600... 0.0732 sec/batch
Epoch: 13/20... Training Step: 12326... Training loss: 1.3418... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12327... Training loss: 1.3153... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12328... Training loss: 1.3259... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12329... Training loss: 1.2963... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12330... Training loss: 1.3354... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12331... Training loss: 1.3908... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12332... Training loss: 1.4051... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12333... Training loss: 1.3581... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12334... Training loss: 1.4164... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12335... Training loss: 1.3309... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12336... Training loss: 1.3326... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12337... Training loss: 1.3726... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12338... Training loss: 1.3642... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12339... Training loss: 1.3545... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12340... Training loss: 1.2973... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12341... Training loss: 1.3717... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12342... Training loss: 1.4096... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12343... Training loss: 1.3388... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12344... Training loss: 1.3641... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12345... Training loss: 1.3174... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12346... Training loss: 1.3789... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12347... Training loss: 1.3418... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12348... Training loss: 1.3701... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12349... Training loss: 1.3124... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12350... Training loss: 1.4268... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12351... Training loss: 1.3312... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12352... Training loss: 1.3346... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12353... Training loss: 1.4096... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12354... Training loss: 1.4114... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12355... Training loss: 1.3396... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12356... Training loss: 1.3235... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12357... Training loss: 1.3889... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12358... Training loss: 1.3379... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12359... Training loss: 1.3487... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12360... Training loss: 1.3198... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12361... Training loss: 1.3019... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12362... Training loss: 1.3198... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12363... Training loss: 1.3168... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12364... Training loss: 1.3505... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12365... Training loss: 1.3601... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12366... Training loss: 1.3075... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12367... Training loss: 1.2961... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12368... Training loss: 1.3528... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12369... Training loss: 1.3138... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12370... Training loss: 1.3002... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12371... Training loss: 1.2441... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12372... Training loss: 1.3177... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12373... Training loss: 1.3121... 0.0738 sec/batch
Epoch: 13/20... Training Step: 12374... Training loss: 1.3831... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12375... Training loss: 1.3529... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12376... Training loss: 1.3384... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12377... Training loss: 1.3522... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12378... Training loss: 1.3335... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12379... Training loss: 1.2166... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12380... Training loss: 1.2891... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12381... Training loss: 1.3361... 0.0831 sec/batch
Epoch: 13/20... Training Step: 12382... Training loss: 1.3181... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12383... Training loss: 1.2566... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12384... Training loss: 1.2866... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12385... Training loss: 1.2737... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12386... Training loss: 1.3254... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12387... Training loss: 1.3052... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12388... Training loss: 1.3187... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12389... Training loss: 1.2414... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12390... Training loss: 1.2978... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12391... Training loss: 1.3464... 0.0808 sec/batch
Epoch: 13/20... Training Step: 12392... Training loss: 1.3052... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12393... Training loss: 1.2780... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12394... Training loss: 1.2678... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12395... Training loss: 1.2744... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12396... Training loss: 1.2936... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12397... Training loss: 1.3087... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12398... Training loss: 1.3053... 0.0789 sec/batch
Epoch: 13/20... Training Step: 12399... Training loss: 1.2838... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12400... Training loss: 1.2716... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12401... Training loss: 1.3052... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12402... Training loss: 1.2990... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12403... Training loss: 1.3048... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12404... Training loss: 1.3301... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12405... Training loss: 1.2662... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12406... Training loss: 1.3045... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12407... Training loss: 1.3142... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12408... Training loss: 1.2919... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12409... Training loss: 1.3014... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12410... Training loss: 1.2694... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12411... Training loss: 1.3022... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12412... Training loss: 1.2757... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12413... Training loss: 1.2714... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12414... Training loss: 1.2870... 0.0790 sec/batch
Epoch: 13/20... Training Step: 12415... Training loss: 1.3541... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12416... Training loss: 1.3215... 0.0789 sec/batch
Epoch: 13/20... Training Step: 12417... Training loss: 1.3618... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12418... Training loss: 1.3640... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12419... Training loss: 1.3007... 0.0817 sec/batch
Epoch: 13/20... Training Step: 12420... Training loss: 1.3197... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12421... Training loss: 1.3243... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12422... Training loss: 1.3277... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12423... Training loss: 1.3459... 0.0739 sec/batch
Epoch: 13/20... Training Step: 12424... Training loss: 1.2762... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12425... Training loss: 1.3179... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12426... Training loss: 1.2835... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12427... Training loss: 1.4004... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12428... Training loss: 1.2897... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12429... Training loss: 1.3121... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12430... Training loss: 1.2904... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12431... Training loss: 1.2988... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12432... Training loss: 1.2741... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12433... Training loss: 1.3032... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12434... Training loss: 1.3783... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12435... Training loss: 1.4121... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12436... Training loss: 1.3444... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12437... Training loss: 1.3860... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12438... Training loss: 1.3659... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12439... Training loss: 1.3174... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12440... Training loss: 1.2843... 0.0800 sec/batch
Epoch: 13/20... Training Step: 12441... Training loss: 1.2817... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12442... Training loss: 1.2916... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12443... Training loss: 1.2937... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12444... Training loss: 1.3124... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12445... Training loss: 1.3937... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12446... Training loss: 1.2881... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12447... Training loss: 1.2884... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12448... Training loss: 1.3046... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12449... Training loss: 1.2634... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12450... Training loss: 1.2805... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12451... Training loss: 1.3527... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12452... Training loss: 1.3270... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12453... Training loss: 1.2916... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12454... Training loss: 1.2936... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12455... Training loss: 1.3049... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12456... Training loss: 1.2590... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12457... Training loss: 1.2692... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12458... Training loss: 1.2934... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12459... Training loss: 1.3220... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12460... Training loss: 1.2825... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12461... Training loss: 1.2685... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12462... Training loss: 1.2810... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12463... Training loss: 1.3313... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12464... Training loss: 1.2234... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12465... Training loss: 1.2773... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12466... Training loss: 1.3106... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12467... Training loss: 1.2807... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12468... Training loss: 1.3028... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12469... Training loss: 1.3186... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12470... Training loss: 1.3362... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12471... Training loss: 1.2851... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12472... Training loss: 1.2255... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12473... Training loss: 1.2346... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12474... Training loss: 1.2655... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12475... Training loss: 1.2537... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12476... Training loss: 1.2757... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12477... Training loss: 1.3686... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12478... Training loss: 1.3471... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12479... Training loss: 1.3373... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12480... Training loss: 1.3286... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12481... Training loss: 1.3010... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12482... Training loss: 1.4285... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12483... Training loss: 1.3662... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12484... Training loss: 1.2835... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12485... Training loss: 1.2435... 0.0803 sec/batch
Epoch: 13/20... Training Step: 12486... Training loss: 1.2546... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12487... Training loss: 1.3688... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12488... Training loss: 1.2337... 0.0838 sec/batch
Epoch: 13/20... Training Step: 12489... Training loss: 1.2814... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12490... Training loss: 1.3521... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12491... Training loss: 1.3347... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12492... Training loss: 1.3459... 0.0812 sec/batch
Epoch: 13/20... Training Step: 12493... Training loss: 1.2981... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12494... Training loss: 1.2793... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12495... Training loss: 1.3181... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12496... Training loss: 1.3305... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12497... Training loss: 1.2506... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12498... Training loss: 1.2867... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12499... Training loss: 1.2132... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12500... Training loss: 1.3369... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12501... Training loss: 1.2001... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12502... Training loss: 1.2222... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12503... Training loss: 1.2745... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12504... Training loss: 1.2855... 0.0736 sec/batch
Epoch: 13/20... Training Step: 12505... Training loss: 1.3350... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12506... Training loss: 1.2909... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12507... Training loss: 1.2732... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12508... Training loss: 1.3358... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12509... Training loss: 1.2575... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12510... Training loss: 1.2852... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12511... Training loss: 1.3490... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12512... Training loss: 1.3419... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12513... Training loss: 1.3105... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12514... Training loss: 1.3376... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12515... Training loss: 1.2495... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12516... Training loss: 1.3231... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12517... Training loss: 1.3429... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12518... Training loss: 1.2773... 0.0805 sec/batch
Epoch: 13/20... Training Step: 12519... Training loss: 1.3751... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12520... Training loss: 1.3060... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12521... Training loss: 1.3328... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12522... Training loss: 1.3631... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12523... Training loss: 1.2760... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12524... Training loss: 1.3044... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12525... Training loss: 1.2785... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12526... Training loss: 1.2476... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12527... Training loss: 1.3678... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12528... Training loss: 1.3348... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12529... Training loss: 1.2662... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12530... Training loss: 1.2874... 0.0786 sec/batch
Epoch: 13/20... Training Step: 12531... Training loss: 1.3060... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12532... Training loss: 1.3134... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12533... Training loss: 1.1448... 0.0792 sec/batch
Epoch: 13/20... Training Step: 12534... Training loss: 1.3124... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12535... Training loss: 1.3191... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12536... Training loss: 1.2843... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12537... Training loss: 1.2673... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12538... Training loss: 1.3029... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12539... Training loss: 1.3515... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12540... Training loss: 1.3000... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12541... Training loss: 1.3762... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12542... Training loss: 1.2781... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12543... Training loss: 1.3134... 0.0740 sec/batch
Epoch: 13/20... Training Step: 12544... Training loss: 1.3955... 0.0802 sec/batch
Epoch: 13/20... Training Step: 12545... Training loss: 1.4093... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12546... Training loss: 1.3671... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12547... Training loss: 1.2734... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12548... Training loss: 1.3357... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12549... Training loss: 1.2767... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12550... Training loss: 1.3138... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12551... Training loss: 1.3385... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12552... Training loss: 1.2334... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12553... Training loss: 1.2627... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12554... Training loss: 1.3043... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12555... Training loss: 1.2952... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12556... Training loss: 1.3638... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12557... Training loss: 1.2920... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12558... Training loss: 1.3863... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12559... Training loss: 1.3888... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12560... Training loss: 1.3447... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12561... Training loss: 1.2563... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12562... Training loss: 1.2964... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12563... Training loss: 1.2960... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12564... Training loss: 1.3906... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12565... Training loss: 1.3213... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12566... Training loss: 1.4125... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12567... Training loss: 1.3092... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12568... Training loss: 1.4103... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12569... Training loss: 1.4078... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12570... Training loss: 1.2852... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12571... Training loss: 1.4295... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12572... Training loss: 1.3705... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12573... Training loss: 1.3288... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12574... Training loss: 1.2841... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12575... Training loss: 1.3643... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12576... Training loss: 1.3975... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12577... Training loss: 1.3165... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12578... Training loss: 1.3062... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12579... Training loss: 1.2685... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12580... Training loss: 1.2739... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12581... Training loss: 1.3955... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12582... Training loss: 1.3050... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12583... Training loss: 1.2518... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12584... Training loss: 1.3882... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12585... Training loss: 1.3824... 0.0793 sec/batch
Epoch: 13/20... Training Step: 12586... Training loss: 1.3464... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12587... Training loss: 1.3563... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12588... Training loss: 1.3003... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12589... Training loss: 1.3672... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12590... Training loss: 1.3886... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12591... Training loss: 1.3041... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12592... Training loss: 1.3102... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12593... Training loss: 1.2867... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12594... Training loss: 1.3625... 0.0786 sec/batch
Epoch: 13/20... Training Step: 12595... Training loss: 1.3090... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12596... Training loss: 1.3369... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12597... Training loss: 1.3131... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12598... Training loss: 1.2579... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12599... Training loss: 1.3136... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12600... Training loss: 1.3028... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12601... Training loss: 1.4160... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12602... Training loss: 1.3525... 0.0737 sec/batch
Epoch: 13/20... Training Step: 12603... Training loss: 1.3117... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12604... Training loss: 1.3103... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12605... Training loss: 1.3348... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12606... Training loss: 1.3036... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12607... Training loss: 1.3160... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12608... Training loss: 1.2491... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12609... Training loss: 1.2458... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12610... Training loss: 1.2827... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12611... Training loss: 1.3087... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12612... Training loss: 1.3365... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12613... Training loss: 1.3213... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12614... Training loss: 1.2959... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12615... Training loss: 1.2736... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12616... Training loss: 1.3801... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12617... Training loss: 1.2845... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12618... Training loss: 1.3207... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12619... Training loss: 1.3141... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12620... Training loss: 1.3409... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12621... Training loss: 1.3673... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12622... Training loss: 1.3454... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12623... Training loss: 1.3478... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12624... Training loss: 1.4000... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12625... Training loss: 1.3616... 0.0797 sec/batch
Epoch: 13/20... Training Step: 12626... Training loss: 1.3164... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12627... Training loss: 1.3284... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12628... Training loss: 1.3496... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12629... Training loss: 1.3796... 0.0789 sec/batch
Epoch: 13/20... Training Step: 12630... Training loss: 1.4114... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12631... Training loss: 1.3010... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12632... Training loss: 1.3675... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12633... Training loss: 1.3681... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12634... Training loss: 1.3092... 0.0829 sec/batch
Epoch: 13/20... Training Step: 12635... Training loss: 1.3089... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12636... Training loss: 1.3283... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12637... Training loss: 1.3050... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12638... Training loss: 1.3658... 0.0741 sec/batch
Epoch: 13/20... Training Step: 12639... Training loss: 1.3311... 0.0807 sec/batch
Epoch: 13/20... Training Step: 12640... Training loss: 1.3161... 0.0826 sec/batch
Epoch: 13/20... Training Step: 12641... Training loss: 1.2966... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12642... Training loss: 1.3986... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12643... Training loss: 1.2937... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12644... Training loss: 1.3109... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12645... Training loss: 1.2952... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12646... Training loss: 1.3106... 0.0782 sec/batch
Epoch: 13/20... Training Step: 12647... Training loss: 1.3346... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12648... Training loss: 1.3665... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12649... Training loss: 1.3401... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12650... Training loss: 1.2851... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12651... Training loss: 1.3307... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12652... Training loss: 1.3317... 0.0839 sec/batch
Epoch: 13/20... Training Step: 12653... Training loss: 1.2563... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12654... Training loss: 1.2818... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12655... Training loss: 1.4235... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12656... Training loss: 1.2981... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12657... Training loss: 1.3591... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12658... Training loss: 1.3567... 0.0818 sec/batch
Epoch: 13/20... Training Step: 12659... Training loss: 1.2628... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12660... Training loss: 1.3476... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12661... Training loss: 1.3293... 0.0809 sec/batch
Epoch: 13/20... Training Step: 12662... Training loss: 1.3783... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12663... Training loss: 1.3675... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12664... Training loss: 1.3994... 0.0791 sec/batch
Epoch: 13/20... Training Step: 12665... Training loss: 1.3155... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12666... Training loss: 1.2643... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12667... Training loss: 1.3501... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12668... Training loss: 1.3238... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12669... Training loss: 1.3530... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12670... Training loss: 1.3823... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12671... Training loss: 1.3353... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12672... Training loss: 1.2774... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12673... Training loss: 1.2967... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12674... Training loss: 1.3484... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12675... Training loss: 1.3317... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12676... Training loss: 1.4098... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12677... Training loss: 1.3647... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12678... Training loss: 1.3580... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12679... Training loss: 1.3145... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12680... Training loss: 1.2373... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12681... Training loss: 1.2806... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12682... Training loss: 1.3298... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12683... Training loss: 1.2791... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12684... Training loss: 1.3051... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12685... Training loss: 1.3316... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12686... Training loss: 1.3289... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12687... Training loss: 1.3515... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12688... Training loss: 1.3352... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12689... Training loss: 1.3215... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12690... Training loss: 1.3351... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12691... Training loss: 1.3401... 0.0789 sec/batch
Epoch: 13/20... Training Step: 12692... Training loss: 1.3145... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12693... Training loss: 1.3493... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12694... Training loss: 1.3417... 0.0790 sec/batch
Epoch: 13/20... Training Step: 12695... Training loss: 1.2966... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12696... Training loss: 1.3118... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12697... Training loss: 1.3223... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12698... Training loss: 1.2852... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12699... Training loss: 1.2383... 0.0740 sec/batch
Epoch: 13/20... Training Step: 12700... Training loss: 1.2904... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12701... Training loss: 1.2636... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12702... Training loss: 1.3108... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12703... Training loss: 1.2878... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12704... Training loss: 1.2935... 0.0743 sec/batch
Epoch: 13/20... Training Step: 12705... Training loss: 1.3489... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12706... Training loss: 1.3799... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12707... Training loss: 1.3777... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12708... Training loss: 1.3329... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12709... Training loss: 1.2668... 0.0803 sec/batch
Epoch: 13/20... Training Step: 12710... Training loss: 1.3587... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12711... Training loss: 1.3005... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12712... Training loss: 1.3651... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12713... Training loss: 1.3601... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12714... Training loss: 1.3968... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12715... Training loss: 1.3287... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12716... Training loss: 1.3750... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12717... Training loss: 1.3096... 0.0806 sec/batch
Epoch: 13/20... Training Step: 12718... Training loss: 1.3722... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12719... Training loss: 1.3564... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12720... Training loss: 1.3570... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12721... Training loss: 1.3846... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12722... Training loss: 1.4501... 0.0811 sec/batch
Epoch: 13/20... Training Step: 12723... Training loss: 1.3594... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12724... Training loss: 1.3254... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12725... Training loss: 1.2964... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12726... Training loss: 1.2406... 0.0740 sec/batch
Epoch: 13/20... Training Step: 12727... Training loss: 1.3319... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12728... Training loss: 1.3172... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12729... Training loss: 1.3478... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12730... Training loss: 1.3503... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12731... Training loss: 1.3617... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12732... Training loss: 1.3916... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12733... Training loss: 1.3744... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12734... Training loss: 1.2973... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12735... Training loss: 1.3901... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12736... Training loss: 1.3401... 0.0788 sec/batch
Epoch: 13/20... Training Step: 12737... Training loss: 1.3823... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12738... Training loss: 1.3397... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12739... Training loss: 1.2618... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12740... Training loss: 1.3038... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12741... Training loss: 1.3457... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12742... Training loss: 1.3586... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12743... Training loss: 1.3569... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12744... Training loss: 1.3480... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12745... Training loss: 1.3343... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12746... Training loss: 1.4693... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12747... Training loss: 1.3355... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12748... Training loss: 1.3884... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12749... Training loss: 1.3791... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12750... Training loss: 1.2767... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12751... Training loss: 1.3489... 0.0779 sec/batch
Epoch: 13/20... Training Step: 12752... Training loss: 1.3103... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12753... Training loss: 1.3291... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12754... Training loss: 1.3665... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12755... Training loss: 1.3131... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12756... Training loss: 1.4215... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12757... Training loss: 1.3977... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12758... Training loss: 1.3322... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12759... Training loss: 1.3353... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12760... Training loss: 1.3389... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12761... Training loss: 1.3515... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12762... Training loss: 1.3454... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12763... Training loss: 1.3479... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12764... Training loss: 1.3567... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12765... Training loss: 1.3446... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12766... Training loss: 1.2781... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12767... Training loss: 1.2940... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12768... Training loss: 1.3124... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12769... Training loss: 1.2853... 0.0797 sec/batch
Epoch: 13/20... Training Step: 12770... Training loss: 1.3525... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12771... Training loss: 1.3524... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12772... Training loss: 1.3755... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12773... Training loss: 1.3423... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12774... Training loss: 1.4130... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12775... Training loss: 1.3596... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12776... Training loss: 1.3328... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12777... Training loss: 1.3413... 0.0737 sec/batch
Epoch: 13/20... Training Step: 12778... Training loss: 1.3638... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12779... Training loss: 1.3348... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12780... Training loss: 1.3513... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12781... Training loss: 1.3261... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12782... Training loss: 1.3823... 0.0739 sec/batch
Epoch: 13/20... Training Step: 12783... Training loss: 1.3385... 0.0747 sec/batch
Epoch: 13/20... Training Step: 12784... Training loss: 1.3178... 0.0769 sec/batch
Epoch: 13/20... Training Step: 12785... Training loss: 1.3281... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12786... Training loss: 1.2714... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12787... Training loss: 1.3554... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12788... Training loss: 1.3203... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12789... Training loss: 1.3462... 0.0771 sec/batch
Epoch: 13/20... Training Step: 12790... Training loss: 1.3914... 0.0787 sec/batch
Epoch: 13/20... Training Step: 12791... Training loss: 1.3483... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12792... Training loss: 1.3659... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12793... Training loss: 1.3642... 0.0770 sec/batch
Epoch: 13/20... Training Step: 12794... Training loss: 1.4049... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12795... Training loss: 1.3683... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12796... Training loss: 1.2633... 0.0783 sec/batch
Epoch: 13/20... Training Step: 12797... Training loss: 1.3289... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12798... Training loss: 1.2558... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12799... Training loss: 1.3833... 0.0795 sec/batch
Epoch: 13/20... Training Step: 12800... Training loss: 1.2711... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12801... Training loss: 1.3538... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12802... Training loss: 1.3643... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12803... Training loss: 1.2975... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12804... Training loss: 1.2690... 0.0759 sec/batch
Epoch: 13/20... Training Step: 12805... Training loss: 1.2779... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12806... Training loss: 1.2934... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12807... Training loss: 1.2675... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12808... Training loss: 1.2804... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12809... Training loss: 1.2436... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12810... Training loss: 1.3458... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12811... Training loss: 1.2696... 0.0798 sec/batch
Epoch: 13/20... Training Step: 12812... Training loss: 1.3290... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12813... Training loss: 1.2704... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12814... Training loss: 1.2357... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12815... Training loss: 1.2560... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12816... Training loss: 1.3269... 0.0810 sec/batch
Epoch: 13/20... Training Step: 12817... Training loss: 1.4193... 0.0784 sec/batch
Epoch: 13/20... Training Step: 12818... Training loss: 1.3418... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12819... Training loss: 1.3374... 0.0751 sec/batch
Epoch: 13/20... Training Step: 12820... Training loss: 1.3340... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12821... Training loss: 1.2917... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12822... Training loss: 1.3695... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12823... Training loss: 1.3207... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12824... Training loss: 1.3556... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12825... Training loss: 1.4219... 0.0789 sec/batch
Epoch: 13/20... Training Step: 12826... Training loss: 1.3589... 0.0750 sec/batch
Epoch: 13/20... Training Step: 12827... Training loss: 1.3298... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12828... Training loss: 1.2847... 0.0772 sec/batch
Epoch: 13/20... Training Step: 12829... Training loss: 1.3419... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12830... Training loss: 1.3597... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12831... Training loss: 1.3377... 0.0757 sec/batch
Epoch: 13/20... Training Step: 12832... Training loss: 1.3240... 0.0746 sec/batch
Epoch: 13/20... Training Step: 12833... Training loss: 1.4350... 0.0758 sec/batch
Epoch: 13/20... Training Step: 12834... Training loss: 1.3919... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12835... Training loss: 1.3545... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12836... Training loss: 1.4180... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12837... Training loss: 1.3022... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12838... Training loss: 1.3460... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12839... Training loss: 1.3524... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12840... Training loss: 1.3729... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12841... Training loss: 1.3138... 0.0781 sec/batch
Epoch: 13/20... Training Step: 12842... Training loss: 1.3416... 0.0739 sec/batch
Epoch: 13/20... Training Step: 12843... Training loss: 1.3235... 0.0740 sec/batch
Epoch: 13/20... Training Step: 12844... Training loss: 1.4209... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12845... Training loss: 1.3556... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12846... Training loss: 1.3440... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12847... Training loss: 1.3906... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12848... Training loss: 1.3764... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12849... Training loss: 1.3712... 0.0805 sec/batch
Epoch: 13/20... Training Step: 12850... Training loss: 1.3270... 0.0776 sec/batch
Epoch: 13/20... Training Step: 12851... Training loss: 1.3272... 0.0785 sec/batch
Epoch: 13/20... Training Step: 12852... Training loss: 1.2742... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12853... Training loss: 1.3387... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12854... Training loss: 1.2536... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12855... Training loss: 1.3251... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12856... Training loss: 1.2682... 0.0744 sec/batch
Epoch: 13/20... Training Step: 12857... Training loss: 1.3176... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12858... Training loss: 1.3412... 0.0773 sec/batch
Epoch: 13/20... Training Step: 12859... Training loss: 1.3565... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12860... Training loss: 1.3431... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12861... Training loss: 1.3342... 0.0756 sec/batch
Epoch: 13/20... Training Step: 12862... Training loss: 1.3447... 0.0764 sec/batch
Epoch: 13/20... Training Step: 12863... Training loss: 1.2965... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12864... Training loss: 1.2914... 0.0766 sec/batch
Epoch: 13/20... Training Step: 12865... Training loss: 1.3278... 0.0742 sec/batch
Epoch: 13/20... Training Step: 12866... Training loss: 1.3104... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12867... Training loss: 1.3335... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12868... Training loss: 1.3430... 0.0749 sec/batch
Epoch: 13/20... Training Step: 12869... Training loss: 1.4159... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12870... Training loss: 1.3823... 0.0827 sec/batch
Epoch: 13/20... Training Step: 12871... Training loss: 1.3714... 0.0755 sec/batch
Epoch: 13/20... Training Step: 12872... Training loss: 1.3909... 0.0748 sec/batch
Epoch: 13/20... Training Step: 12873... Training loss: 1.4083... 0.0780 sec/batch
Epoch: 13/20... Training Step: 12874... Training loss: 1.5347... 0.0763 sec/batch
Epoch: 13/20... Training Step: 12875... Training loss: 1.3136... 0.0768 sec/batch
Epoch: 13/20... Training Step: 12876... Training loss: 1.4273... 0.0793 sec/batch
Epoch: 13/20... Training Step: 12877... Training loss: 1.4235... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12878... Training loss: 1.2917... 0.0760 sec/batch
Epoch: 13/20... Training Step: 12879... Training loss: 1.3725... 0.0791 sec/batch
Epoch: 13/20... Training Step: 12880... Training loss: 1.3752... 0.0778 sec/batch
Epoch: 13/20... Training Step: 12881... Training loss: 1.3892... 0.0752 sec/batch
Epoch: 13/20... Training Step: 12882... Training loss: 1.3832... 0.0774 sec/batch
Epoch: 13/20... Training Step: 12883... Training loss: 1.3609... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12884... Training loss: 1.2986... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12885... Training loss: 1.2571... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12886... Training loss: 1.4239... 0.0761 sec/batch
Epoch: 13/20... Training Step: 12887... Training loss: 1.3526... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12888... Training loss: 1.3424... 0.0775 sec/batch
Epoch: 13/20... Training Step: 12889... Training loss: 1.3389... 0.0777 sec/batch
Epoch: 13/20... Training Step: 12890... Training loss: 1.3174... 0.0765 sec/batch
Epoch: 13/20... Training Step: 12891... Training loss: 1.2981... 0.0762 sec/batch
Epoch: 13/20... Training Step: 12892... Training loss: 1.3243... 0.0754 sec/batch
Epoch: 13/20... Training Step: 12893... Training loss: 1.3211... 0.0753 sec/batch
Epoch: 13/20... Training Step: 12894... Training loss: 1.2318... 0.0767 sec/batch
Epoch: 13/20... Training Step: 12895... Training loss: 1.3131... 0.0745 sec/batch
Epoch: 13/20... Training Step: 12896... Training loss: 1.3686... 0.0750 sec/batch
Epoch: 14/20... Training Step: 12897... Training loss: 1.4121... 0.0772 sec/batch
Epoch: 14/20... Training Step: 12898... Training loss: 1.2843... 0.0744 sec/batch
Epoch: 14/20... Training Step: 12899... Training loss: 1.3872... 0.0777 sec/batch
Epoch: 14/20... Training Step: 12900... Training loss: 1.3294... 0.0765 sec/batch
Epoch: 14/20... Training Step: 12901... Training loss: 1.2491... 0.0769 sec/batch
Epoch: 14/20... Training Step: 12902... Training loss: 1.2125... 0.0761 sec/batch
Epoch: 14/20... Training Step: 12903... Training loss: 1.3060... 0.0789 sec/batch
Epoch: 14/20... Training Step: 12904... Training loss: 1.3413... 0.0765 sec/batch
Epoch: 14/20... Training Step: 12905... Training loss: 1.3686... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12906... Training loss: 1.2396... 0.0757 sec/batch
Epoch: 14/20... Training Step: 12907... Training loss: 1.4016... 0.0774 sec/batch
Epoch: 14/20... Training Step: 12908... Training loss: 1.2935... 0.0749 sec/batch
Epoch: 14/20... Training Step: 12909... Training loss: 1.3062... 0.0752 sec/batch
Epoch: 14/20... Training Step: 12910... Training loss: 1.3229... 0.0758 sec/batch
Epoch: 14/20... Training Step: 12911... Training loss: 1.3519... 0.0760 sec/batch
Epoch: 14/20... Training Step: 12912... Training loss: 1.2792... 0.0761 sec/batch
Epoch: 14/20... Training Step: 12913... Training loss: 1.3161... 0.0769 sec/batch
Epoch: 14/20... Training Step: 12914... Training loss: 1.3449... 0.0760 sec/batch
Epoch: 14/20... Training Step: 12915... Training loss: 1.3307... 0.0761 sec/batch
Epoch: 14/20... Training Step: 12916... Training loss: 1.2859... 0.0763 sec/batch
Epoch: 14/20... Training Step: 12917... Training loss: 1.3169... 0.0750 sec/batch
Epoch: 14/20... Training Step: 12918... Training loss: 1.2842... 0.0764 sec/batch
Epoch: 14/20... Training Step: 12919... Training loss: 1.3220... 0.0757 sec/batch
Epoch: 14/20... Training Step: 12920... Training loss: 1.2724... 0.0753 sec/batch
Epoch: 14/20... Training Step: 12921... Training loss: 1.3594... 0.0778 sec/batch
Epoch: 14/20... Training Step: 12922... Training loss: 1.2627... 0.0760 sec/batch
Epoch: 14/20... Training Step: 12923... Training loss: 1.2925... 0.0755 sec/batch
Epoch: 14/20... Training Step: 12924... Training loss: 1.2617... 0.0745 sec/batch
Epoch: 14/20... Training Step: 12925... Training loss: 1.2995... 0.0754 sec/batch
Epoch: 14/20... Training Step: 12926... Training loss: 1.3668... 0.0763 sec/batch
Epoch: 14/20... Training Step: 12927... Training loss: 1.3587... 0.0771 sec/batch
Epoch: 14/20... Training Step: 12928... Training loss: 1.3654... 0.0766 sec/batch
Epoch: 14/20... Training Step: 12929... Training loss: 1.2968... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12930... Training loss: 1.3623... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12931... Training loss: 1.4120... 0.0775 sec/batch
Epoch: 14/20... Training Step: 12932... Training loss: 1.3740... 0.0746 sec/batch
Epoch: 14/20... Training Step: 12933... Training loss: 1.3149... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12934... Training loss: 1.3777... 0.0782 sec/batch
Epoch: 14/20... Training Step: 12935... Training loss: 1.3295... 0.0738 sec/batch
Epoch: 14/20... Training Step: 12936... Training loss: 1.3068... 0.0805 sec/batch
Epoch: 14/20... Training Step: 12937... Training loss: 1.3233... 0.0797 sec/batch
Epoch: 14/20... Training Step: 12938... Training loss: 1.3135... 0.0762 sec/batch
Epoch: 14/20... Training Step: 12939... Training loss: 1.2651... 0.0799 sec/batch
Epoch: 14/20... Training Step: 12940... Training loss: 1.2892... 0.0739 sec/batch
Epoch: 14/20... Training Step: 12941... Training loss: 1.3538... 0.0745 sec/batch
Epoch: 14/20... Training Step: 12942... Training loss: 1.3034... 0.0788 sec/batch
Epoch: 14/20... Training Step: 12943... Training loss: 1.3794... 0.0738 sec/batch
Epoch: 14/20... Training Step: 12944... Training loss: 1.2839... 0.0739 sec/batch
Epoch: 14/20... Training Step: 12945... Training loss: 1.3501... 0.0779 sec/batch
Epoch: 14/20... Training Step: 12946... Training loss: 1.3613... 0.0773 sec/batch
Epoch: 14/20... Training Step: 12947... Training loss: 1.2381... 0.0749 sec/batch
Epoch: 14/20... Training Step: 12948... Training loss: 1.3679... 0.0762 sec/batch
Epoch: 14/20... Training Step: 12949... Training loss: 1.2635... 0.0744 sec/batch
Epoch: 14/20... Training Step: 12950... Training loss: 1.2605... 0.0767 sec/batch
Epoch: 14/20... Training Step: 12951... Training loss: 1.2991... 0.0761 sec/batch
Epoch: 14/20... Training Step: 12952... Training loss: 1.2828... 0.0743 sec/batch
Epoch: 14/20... Training Step: 12953... Training loss: 1.3288... 0.0751 sec/batch
Epoch: 14/20... Training Step: 12954... Training loss: 1.3535... 0.0780 sec/batch
Epoch: 14/20... Training Step: 12955... Training loss: 1.3235... 0.0765 sec/batch
Epoch: 14/20... Training Step: 12956... Training loss: 1.3286... 0.0739 sec/batch
Epoch: 14/20... Training Step: 12957... Training loss: 1.2340... 0.0770 sec/batch
Epoch: 14/20... Training Step: 12958... Training loss: 1.2649... 0.0745 sec/batch
Epoch: 14/20... Training Step: 12959... Training loss: 1.2975... 0.0750 sec/batch
Epoch: 14/20... Training Step: 12960... Training loss: 1.2423... 0.0758 sec/batch
Epoch: 14/20... Training Step: 12961... Training loss: 1.3334... 0.0748 sec/batch
Epoch: 14/20... Training Step: 12962... Training loss: 1.3183... 0.0763 sec/batch
Epoch: 14/20... Training Step: 12963... Training loss: 1.2746... 0.0750 sec/batch
Epoch: 14/20... Training Step: 12964... Training loss: 1.3067... 0.0757 sec/batch
Epoch: 14/20... Training Step: 12965... Training loss: 1.3389... 0.0744 sec/batch
Epoch: 14/20... Training Step: 12966... Training loss: 1.3302... 0.0773 sec/batch
Epoch: 14/20... Training Step: 12967... Training loss: 1.3388... 0.0782 sec/batch
Epoch: 14/20... Training Step: 12968... Training loss: 1.3948... 0.0782 sec/batch
Epoch: 14/20... Training Step: 12969... Training loss: 1.3363... 0.0744 sec/batch
Epoch: 14/20... Training Step: 12970... Training loss: 1.2527... 0.0740 sec/batch
Epoch: 14/20... Training Step: 12971... Training loss: 1.2880... 0.0755 sec/batch
Epoch: 14/20... Training Step: 12972... Training loss: 1.2645... 0.0751 sec/batch
Epoch: 14/20... Training Step: 12973... Training loss: 1.2694... 0.0762 sec/batch
Epoch: 14/20... Training Step: 12974... Training loss: 1.2973... 0.0745 sec/batch
Epoch: 14/20... Training Step: 12975... Training loss: 1.2989... 0.0786 sec/batch
Epoch: 14/20... Training Step: 12976... Training loss: 1.3847... 0.0789 sec/batch
Epoch: 14/20... Training Step: 12977... Training loss: 1.3180... 0.0758 sec/batch
Epoch: 14/20... Training Step: 12978... Training loss: 1.2741... 0.0768 sec/batch
Epoch: 14/20... Training Step: 12979... Training loss: 1.3320... 0.0773 sec/batch
Epoch: 14/20... Training Step: 12980... Training loss: 1.3454... 0.0771 sec/batch
Epoch: 14/20... Training Step: 12981... Training loss: 1.3069... 0.0783 sec/batch
Epoch: 14/20... Training Step: 12982... Training loss: 1.3511... 0.0755 sec/batch
Epoch: 14/20... Training Step: 12983... Training loss: 1.3186... 0.0746 sec/batch
Epoch: 14/20... Training Step: 12984... Training loss: 1.3152... 0.0751 sec/batch
Epoch: 14/20... Training Step: 12985... Training loss: 1.3595... 0.0768 sec/batch
Epoch: 14/20... Training Step: 12986... Training loss: 1.3380... 0.0747 sec/batch
Epoch: 14/20... Training Step: 12987... Training loss: 1.3625... 0.0765 sec/batch
Epoch: 14/20... Training Step: 12988... Training loss: 1.2989... 0.0768 sec/batch
Epoch: 14/20... Training Step: 12989... Training loss: 1.2637... 0.0739 sec/batch
Epoch: 14/20... Training Step: 12990... Training loss: 1.2344... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12991... Training loss: 1.2347... 0.0753 sec/batch
Epoch: 14/20... Training Step: 12992... Training loss: 1.3207... 0.0741 sec/batch
Epoch: 14/20... Training Step: 12993... Training loss: 1.2980... 0.0777 sec/batch
Epoch: 14/20... Training Step: 12994... Training loss: 1.2813... 0.0779 sec/batch
Epoch: 14/20... Training Step: 12995... Training loss: 1.2996... 0.0775 sec/batch
Epoch: 14/20... Training Step: 12996... Training loss: 1.2670... 0.0783 sec/batch
Epoch: 14/20... Training Step: 12997... Training loss: 1.2951... 0.0756 sec/batch
Epoch: 14/20... Training Step: 12998... Training loss: 1.3026... 0.0751 sec/batch
Epoch: 14/20... Training Step: 12999... Training loss: 1.3213... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13000... Training loss: 1.3224... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13001... Training loss: 1.3465... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13002... Training loss: 1.3051... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13003... Training loss: 1.2656... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13004... Training loss: 1.3492... 0.0787 sec/batch
Epoch: 14/20... Training Step: 13005... Training loss: 1.3007... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13006... Training loss: 1.3697... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13007... Training loss: 1.3122... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13008... Training loss: 1.3112... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13009... Training loss: 1.3872... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13010... Training loss: 1.3375... 0.0853 sec/batch
Epoch: 14/20... Training Step: 13011... Training loss: 1.2688... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13012... Training loss: 1.4129... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13013... Training loss: 1.3772... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13014... Training loss: 1.4184... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13015... Training loss: 1.3154... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13016... Training loss: 1.3355... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13017... Training loss: 1.4445... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13018... Training loss: 1.2967... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13019... Training loss: 1.3523... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13020... Training loss: 1.3325... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13021... Training loss: 1.3980... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13022... Training loss: 1.3918... 0.0791 sec/batch
Epoch: 14/20... Training Step: 13023... Training loss: 1.3687... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13024... Training loss: 1.3341... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13025... Training loss: 1.2681... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13026... Training loss: 1.3402... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13027... Training loss: 1.2952... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13028... Training loss: 1.3666... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13029... Training loss: 1.3503... 0.0738 sec/batch
Epoch: 14/20... Training Step: 13030... Training loss: 1.3299... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13031... Training loss: 1.3704... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13032... Training loss: 1.3650... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13033... Training loss: 1.3726... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13034... Training loss: 1.3965... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13035... Training loss: 1.3617... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13036... Training loss: 1.3047... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13037... Training loss: 1.3245... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13038... Training loss: 1.2840... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13039... Training loss: 1.3001... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13040... Training loss: 1.3282... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13041... Training loss: 1.1923... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13042... Training loss: 1.3752... 0.0726 sec/batch
Epoch: 14/20... Training Step: 13043... Training loss: 1.3105... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13044... Training loss: 1.2508... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13045... Training loss: 1.2831... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13046... Training loss: 1.2870... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13047... Training loss: 1.2042... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13048... Training loss: 1.2990... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13049... Training loss: 1.2490... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13050... Training loss: 1.3174... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13051... Training loss: 1.3039... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13052... Training loss: 1.2643... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13053... Training loss: 1.2928... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13054... Training loss: 1.3089... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13055... Training loss: 1.3252... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13056... Training loss: 1.3511... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13057... Training loss: 1.2985... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13058... Training loss: 1.3275... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13059... Training loss: 1.2896... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13060... Training loss: 1.3606... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13061... Training loss: 1.3307... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13062... Training loss: 1.3955... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13063... Training loss: 1.3515... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13064... Training loss: 1.2953... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13065... Training loss: 1.3473... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13066... Training loss: 1.3257... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13067... Training loss: 1.3465... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13068... Training loss: 1.3115... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13069... Training loss: 1.3306... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13070... Training loss: 1.3649... 0.0818 sec/batch
Epoch: 14/20... Training Step: 13071... Training loss: 1.2955... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13072... Training loss: 1.3368... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13073... Training loss: 1.3380... 0.0805 sec/batch
Epoch: 14/20... Training Step: 13074... Training loss: 1.4151... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13075... Training loss: 1.3621... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13076... Training loss: 1.2988... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13077... Training loss: 1.3347... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13078... Training loss: 1.3931... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13079... Training loss: 1.3293... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13080... Training loss: 1.3504... 0.0740 sec/batch
Epoch: 14/20... Training Step: 13081... Training loss: 1.3529... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13082... Training loss: 1.3566... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13083... Training loss: 1.3204... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13084... Training loss: 1.3066... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13085... Training loss: 1.3254... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13086... Training loss: 1.3731... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13087... Training loss: 1.4000... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13088... Training loss: 1.3387... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13089... Training loss: 1.3742... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13090... Training loss: 1.4153... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13091... Training loss: 1.3316... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13092... Training loss: 1.2852... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13093... Training loss: 1.3794... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13094... Training loss: 1.3535... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13095... Training loss: 1.3450... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13096... Training loss: 1.3210... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13097... Training loss: 1.3301... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13098... Training loss: 1.3287... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13099... Training loss: 1.3183... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13100... Training loss: 1.3265... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13101... Training loss: 1.3083... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13102... Training loss: 1.3650... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13103... Training loss: 1.3216... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13104... Training loss: 1.3072... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13105... Training loss: 1.3592... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13106... Training loss: 1.3244... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13107... Training loss: 1.2789... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13108... Training loss: 1.3690... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13109... Training loss: 1.3064... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13110... Training loss: 1.3605... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13111... Training loss: 1.3108... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13112... Training loss: 1.3365... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13113... Training loss: 1.2933... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13114... Training loss: 1.3318... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13115... Training loss: 1.3378... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13116... Training loss: 1.2613... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13117... Training loss: 1.2753... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13118... Training loss: 1.4276... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13119... Training loss: 1.3506... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13120... Training loss: 1.3364... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13121... Training loss: 1.3685... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13122... Training loss: 1.2997... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13123... Training loss: 1.3443... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13124... Training loss: 1.3453... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13125... Training loss: 1.2313... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13126... Training loss: 1.2586... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13127... Training loss: 1.3160... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13128... Training loss: 1.3188... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13129... Training loss: 1.3054... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13130... Training loss: 1.4067... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13131... Training loss: 1.2992... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13132... Training loss: 1.3163... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13133... Training loss: 1.2795... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13134... Training loss: 1.3611... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13135... Training loss: 1.3539... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13136... Training loss: 1.3028... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13137... Training loss: 1.3129... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13138... Training loss: 1.3455... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13139... Training loss: 1.3604... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13140... Training loss: 1.3593... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13141... Training loss: 1.3052... 0.0797 sec/batch
Epoch: 14/20... Training Step: 13142... Training loss: 1.3525... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13143... Training loss: 1.2778... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13144... Training loss: 1.2385... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13145... Training loss: 1.3087... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13146... Training loss: 1.3481... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13147... Training loss: 1.3013... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13148... Training loss: 1.3146... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13149... Training loss: 1.2650... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13150... Training loss: 1.2804... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13151... Training loss: 1.2566... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13152... Training loss: 1.2837... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13153... Training loss: 1.4046... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13154... Training loss: 1.2978... 0.0810 sec/batch
Epoch: 14/20... Training Step: 13155... Training loss: 1.3103... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13156... Training loss: 1.3164... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13157... Training loss: 1.3494... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13158... Training loss: 1.3705... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13159... Training loss: 1.3476... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13160... Training loss: 1.3530... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13161... Training loss: 1.3243... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13162... Training loss: 1.2371... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13163... Training loss: 1.2869... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13164... Training loss: 1.3188... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13165... Training loss: 1.2638... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13166... Training loss: 1.2696... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13167... Training loss: 1.3087... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13168... Training loss: 1.2672... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13169... Training loss: 1.2260... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13170... Training loss: 1.2543... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13171... Training loss: 1.3115... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13172... Training loss: 1.3104... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13173... Training loss: 1.3298... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13174... Training loss: 1.2318... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13175... Training loss: 1.2808... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13176... Training loss: 1.3678... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13177... Training loss: 1.3036... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13178... Training loss: 1.2926... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13179... Training loss: 1.3382... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13180... Training loss: 1.2749... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13181... Training loss: 1.2046... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13182... Training loss: 1.2782... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13183... Training loss: 1.3389... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13184... Training loss: 1.2834... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13185... Training loss: 1.2901... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13186... Training loss: 1.2605... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13187... Training loss: 1.2812... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13188... Training loss: 1.3051... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13189... Training loss: 1.2694... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13190... Training loss: 1.2706... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13191... Training loss: 1.3109... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13192... Training loss: 1.3448... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13193... Training loss: 1.2898... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13194... Training loss: 1.4000... 0.0791 sec/batch
Epoch: 14/20... Training Step: 13195... Training loss: 1.3038... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13196... Training loss: 1.3302... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13197... Training loss: 1.3885... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13198... Training loss: 1.2653... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13199... Training loss: 1.2541... 0.0814 sec/batch
Epoch: 14/20... Training Step: 13200... Training loss: 1.3437... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13201... Training loss: 1.3012... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13202... Training loss: 1.3727... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13203... Training loss: 1.2699... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13204... Training loss: 1.2456... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13205... Training loss: 1.3896... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13206... Training loss: 1.3049... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13207... Training loss: 1.3752... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13208... Training loss: 1.2941... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13209... Training loss: 1.2878... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13210... Training loss: 1.3240... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13211... Training loss: 1.3249... 0.0785 sec/batch
Epoch: 14/20... Training Step: 13212... Training loss: 1.2744... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13213... Training loss: 1.2903... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13214... Training loss: 1.2620... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13215... Training loss: 1.3023... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13216... Training loss: 1.3123... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13217... Training loss: 1.3266... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13218... Training loss: 1.2539... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13219... Training loss: 1.2856... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13220... Training loss: 1.2885... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13221... Training loss: 1.3422... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13222... Training loss: 1.2939... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13223... Training loss: 1.3200... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13224... Training loss: 1.2862... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13225... Training loss: 1.3121... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13226... Training loss: 1.3394... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13227... Training loss: 1.3454... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13228... Training loss: 1.2549... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13229... Training loss: 1.3983... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13230... Training loss: 1.3828... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13231... Training loss: 1.3900... 0.0798 sec/batch
Epoch: 14/20... Training Step: 13232... Training loss: 1.3358... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13233... Training loss: 1.2874... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13234... Training loss: 1.3117... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13235... Training loss: 1.2929... 0.0801 sec/batch
Epoch: 14/20... Training Step: 13236... Training loss: 1.2762... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13237... Training loss: 1.3035... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13238... Training loss: 1.2899... 0.0794 sec/batch
Epoch: 14/20... Training Step: 13239... Training loss: 1.2749... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13240... Training loss: 1.3522... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13241... Training loss: 1.3146... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13242... Training loss: 1.3164... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13243... Training loss: 1.3317... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13244... Training loss: 1.3727... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13245... Training loss: 1.3672... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13246... Training loss: 1.3701... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13247... Training loss: 1.3475... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13248... Training loss: 1.2787... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13249... Training loss: 1.3249... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13250... Training loss: 1.3005... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13251... Training loss: 1.3215... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13252... Training loss: 1.3305... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13253... Training loss: 1.3018... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13254... Training loss: 1.3215... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13255... Training loss: 1.3114... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13256... Training loss: 1.3200... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13257... Training loss: 1.3298... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13258... Training loss: 1.3787... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13259... Training loss: 1.3004... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13260... Training loss: 1.3273... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13261... Training loss: 1.3085... 0.0802 sec/batch
Epoch: 14/20... Training Step: 13262... Training loss: 1.2773... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13263... Training loss: 1.2858... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13264... Training loss: 1.2478... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13265... Training loss: 1.2786... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13266... Training loss: 1.2517... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13267... Training loss: 1.3722... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13268... Training loss: 1.2567... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13269... Training loss: 1.3068... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13270... Training loss: 1.2286... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13271... Training loss: 1.2876... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13272... Training loss: 1.2749... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13273... Training loss: 1.3065... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13274... Training loss: 1.3228... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13275... Training loss: 1.2870... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13276... Training loss: 1.3249... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13277... Training loss: 1.3274... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13278... Training loss: 1.2892... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13279... Training loss: 1.3494... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13280... Training loss: 1.3205... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13281... Training loss: 1.3535... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13282... Training loss: 1.3291... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13283... Training loss: 1.3513... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13284... Training loss: 1.3496... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13285... Training loss: 1.3570... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13286... Training loss: 1.2963... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13287... Training loss: 1.3268... 0.0815 sec/batch
Epoch: 14/20... Training Step: 13288... Training loss: 1.3111... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13289... Training loss: 1.3402... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13290... Training loss: 1.2810... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13291... Training loss: 1.3179... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13292... Training loss: 1.3013... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13293... Training loss: 1.2942... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13294... Training loss: 1.3581... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13295... Training loss: 1.3243... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13296... Training loss: 1.2918... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13297... Training loss: 1.2676... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13298... Training loss: 1.3229... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13299... Training loss: 1.2895... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13300... Training loss: 1.3508... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13301... Training loss: 1.3053... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13302... Training loss: 1.3104... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13303... Training loss: 1.2436... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13304... Training loss: 1.3817... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13305... Training loss: 1.2606... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13306... Training loss: 1.2468... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13307... Training loss: 1.2106... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13308... Training loss: 1.2668... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13309... Training loss: 1.3059... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13310... Training loss: 1.2227... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13311... Training loss: 1.2587... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13312... Training loss: 1.3369... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13313... Training loss: 1.2467... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13314... Training loss: 1.2756... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13315... Training loss: 1.3580... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13316... Training loss: 1.3823... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13317... Training loss: 1.2310... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13318... Training loss: 1.3275... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13319... Training loss: 1.3026... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13320... Training loss: 1.3306... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13321... Training loss: 1.3116... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13322... Training loss: 1.3082... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13323... Training loss: 1.3768... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13324... Training loss: 1.4133... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13325... Training loss: 1.3694... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13326... Training loss: 1.4071... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13327... Training loss: 1.3067... 0.0826 sec/batch
Epoch: 14/20... Training Step: 13328... Training loss: 1.3209... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13329... Training loss: 1.3592... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13330... Training loss: 1.3483... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13331... Training loss: 1.3482... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13332... Training loss: 1.2825... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13333... Training loss: 1.3642... 0.0785 sec/batch
Epoch: 14/20... Training Step: 13334... Training loss: 1.4179... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13335... Training loss: 1.3364... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13336... Training loss: 1.3683... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13337... Training loss: 1.2992... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13338... Training loss: 1.3646... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13339... Training loss: 1.3321... 0.0794 sec/batch
Epoch: 14/20... Training Step: 13340... Training loss: 1.3434... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13341... Training loss: 1.3009... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13342... Training loss: 1.4105... 0.0821 sec/batch
Epoch: 14/20... Training Step: 13343... Training loss: 1.3213... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13344... Training loss: 1.3160... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13345... Training loss: 1.3924... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13346... Training loss: 1.4197... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13347... Training loss: 1.3143... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13348... Training loss: 1.3001... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13349... Training loss: 1.3601... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13350... Training loss: 1.3344... 0.0787 sec/batch
Epoch: 14/20... Training Step: 13351... Training loss: 1.3581... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13352... Training loss: 1.2903... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13353... Training loss: 1.2972... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13354... Training loss: 1.2984... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13355... Training loss: 1.3003... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13356... Training loss: 1.3283... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13357... Training loss: 1.3309... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13358... Training loss: 1.3160... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13359... Training loss: 1.2918... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13360... Training loss: 1.3380... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13361... Training loss: 1.3008... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13362... Training loss: 1.2839... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13363... Training loss: 1.2284... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13364... Training loss: 1.2906... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13365... Training loss: 1.2967... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13366... Training loss: 1.3947... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13367... Training loss: 1.3291... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13368... Training loss: 1.3147... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13369... Training loss: 1.3295... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13370... Training loss: 1.3011... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13371... Training loss: 1.2034... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13372... Training loss: 1.2767... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13373... Training loss: 1.3315... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13374... Training loss: 1.3107... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13375... Training loss: 1.2569... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13376... Training loss: 1.2572... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13377... Training loss: 1.2862... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13378... Training loss: 1.3374... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13379... Training loss: 1.2971... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13380... Training loss: 1.3131... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13381... Training loss: 1.2459... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13382... Training loss: 1.2924... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13383... Training loss: 1.3188... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13384... Training loss: 1.2881... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13385... Training loss: 1.2799... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13386... Training loss: 1.2665... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13387... Training loss: 1.2948... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13388... Training loss: 1.2783... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13389... Training loss: 1.3020... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13390... Training loss: 1.3055... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13391... Training loss: 1.2670... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13392... Training loss: 1.2559... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13393... Training loss: 1.2973... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13394... Training loss: 1.2754... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13395... Training loss: 1.2634... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13396... Training loss: 1.3261... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13397... Training loss: 1.2601... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13398... Training loss: 1.3149... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13399... Training loss: 1.3065... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13400... Training loss: 1.2838... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13401... Training loss: 1.3106... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13402... Training loss: 1.2282... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13403... Training loss: 1.2969... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13404... Training loss: 1.2550... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13405... Training loss: 1.2643... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13406... Training loss: 1.2946... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13407... Training loss: 1.3384... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13408... Training loss: 1.3286... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13409... Training loss: 1.3438... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13410... Training loss: 1.3618... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13411... Training loss: 1.2914... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13412... Training loss: 1.3198... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13413... Training loss: 1.3155... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13414... Training loss: 1.3049... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13415... Training loss: 1.3339... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13416... Training loss: 1.2745... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13417... Training loss: 1.3143... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13418... Training loss: 1.2738... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13419... Training loss: 1.3947... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13420... Training loss: 1.2870... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13421... Training loss: 1.3072... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13422... Training loss: 1.2963... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13423... Training loss: 1.3113... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13424... Training loss: 1.2675... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13425... Training loss: 1.2986... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13426... Training loss: 1.3608... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13427... Training loss: 1.4271... 0.0738 sec/batch
Epoch: 14/20... Training Step: 13428... Training loss: 1.3254... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13429... Training loss: 1.3498... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13430... Training loss: 1.3424... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13431... Training loss: 1.3033... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13432... Training loss: 1.2693... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13433... Training loss: 1.2883... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13434... Training loss: 1.3099... 0.0802 sec/batch
Epoch: 14/20... Training Step: 13435... Training loss: 1.2772... 0.0739 sec/batch
Epoch: 14/20... Training Step: 13436... Training loss: 1.2883... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13437... Training loss: 1.3810... 0.0792 sec/batch
Epoch: 14/20... Training Step: 13438... Training loss: 1.2869... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13439... Training loss: 1.2899... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13440... Training loss: 1.3097... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13441... Training loss: 1.2459... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13442... Training loss: 1.3009... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13443... Training loss: 1.3706... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13444... Training loss: 1.3186... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13445... Training loss: 1.2705... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13446... Training loss: 1.2740... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13447... Training loss: 1.2956... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13448... Training loss: 1.2430... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13449... Training loss: 1.2491... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13450... Training loss: 1.2816... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13451... Training loss: 1.3107... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13452... Training loss: 1.2612... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13453... Training loss: 1.2531... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13454... Training loss: 1.2795... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13455... Training loss: 1.3166... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13456... Training loss: 1.2204... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13457... Training loss: 1.2663... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13458... Training loss: 1.2956... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13459... Training loss: 1.2589... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13460... Training loss: 1.2983... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13461... Training loss: 1.2910... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13462... Training loss: 1.3073... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13463... Training loss: 1.2721... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13464... Training loss: 1.2355... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13465... Training loss: 1.2000... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13466... Training loss: 1.2330... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13467... Training loss: 1.2603... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13468... Training loss: 1.2840... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13469... Training loss: 1.3600... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13470... Training loss: 1.3179... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13471... Training loss: 1.3349... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13472... Training loss: 1.3140... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13473... Training loss: 1.3054... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13474... Training loss: 1.4220... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13475... Training loss: 1.3634... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13476... Training loss: 1.2858... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13477... Training loss: 1.2277... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13478... Training loss: 1.2591... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13479... Training loss: 1.3588... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13480... Training loss: 1.2512... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13481... Training loss: 1.2899... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13482... Training loss: 1.3238... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13483... Training loss: 1.3152... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13484... Training loss: 1.3315... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13485... Training loss: 1.2903... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13486... Training loss: 1.2849... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13487... Training loss: 1.3163... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13488... Training loss: 1.3276... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13489... Training loss: 1.2609... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13490... Training loss: 1.2602... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13491... Training loss: 1.2281... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13492... Training loss: 1.3167... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13493... Training loss: 1.1740... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13494... Training loss: 1.2261... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13495... Training loss: 1.2761... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13496... Training loss: 1.2840... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13497... Training loss: 1.3293... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13498... Training loss: 1.2957... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13499... Training loss: 1.2861... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13500... Training loss: 1.3181... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13501... Training loss: 1.2402... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13502... Training loss: 1.2640... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13503... Training loss: 1.3377... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13504... Training loss: 1.3392... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13505... Training loss: 1.3182... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13506... Training loss: 1.3547... 0.0786 sec/batch
Epoch: 14/20... Training Step: 13507... Training loss: 1.2277... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13508... Training loss: 1.3364... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13509... Training loss: 1.3306... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13510... Training loss: 1.2846... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13511... Training loss: 1.3375... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13512... Training loss: 1.3093... 0.0807 sec/batch
Epoch: 14/20... Training Step: 13513... Training loss: 1.3372... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13514... Training loss: 1.3491... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13515... Training loss: 1.2460... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13516... Training loss: 1.2851... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13517... Training loss: 1.2658... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13518... Training loss: 1.2517... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13519... Training loss: 1.3512... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13520... Training loss: 1.3266... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13521... Training loss: 1.2642... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13522... Training loss: 1.2875... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13523... Training loss: 1.2758... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13524... Training loss: 1.3146... 0.0832 sec/batch
Epoch: 14/20... Training Step: 13525... Training loss: 1.1378... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13526... Training loss: 1.2843... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13527... Training loss: 1.3187... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13528... Training loss: 1.2328... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13529... Training loss: 1.2530... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13530... Training loss: 1.2933... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13531... Training loss: 1.3153... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13532... Training loss: 1.2786... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13533... Training loss: 1.3535... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13534... Training loss: 1.2811... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13535... Training loss: 1.3293... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13536... Training loss: 1.3994... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13537... Training loss: 1.3799... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13538... Training loss: 1.3722... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13539... Training loss: 1.2665... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13540... Training loss: 1.3339... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13541... Training loss: 1.2576... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13542... Training loss: 1.3353... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13543... Training loss: 1.3122... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13544... Training loss: 1.2343... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13545... Training loss: 1.2710... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13546... Training loss: 1.2961... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13547... Training loss: 1.2690... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13548... Training loss: 1.3427... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13549... Training loss: 1.2864... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13550... Training loss: 1.3531... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13551... Training loss: 1.3688... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13552... Training loss: 1.3378... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13553... Training loss: 1.2222... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13554... Training loss: 1.2772... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13555... Training loss: 1.2734... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13556... Training loss: 1.3920... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13557... Training loss: 1.3480... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13558... Training loss: 1.3697... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13559... Training loss: 1.3035... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13560... Training loss: 1.3821... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13561... Training loss: 1.4116... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13562... Training loss: 1.2736... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13563... Training loss: 1.4139... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13564... Training loss: 1.3344... 0.0810 sec/batch
Epoch: 14/20... Training Step: 13565... Training loss: 1.3328... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13566... Training loss: 1.2880... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13567... Training loss: 1.3448... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13568... Training loss: 1.3788... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13569... Training loss: 1.2979... 0.0845 sec/batch
Epoch: 14/20... Training Step: 13570... Training loss: 1.3164... 0.0742 sec/batch
Epoch: 14/20... Training Step: 13571... Training loss: 1.2545... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13572... Training loss: 1.2803... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13573... Training loss: 1.3821... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13574... Training loss: 1.2911... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13575... Training loss: 1.2550... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13576... Training loss: 1.3464... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13577... Training loss: 1.3542... 0.0787 sec/batch
Epoch: 14/20... Training Step: 13578... Training loss: 1.3560... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13579... Training loss: 1.3663... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13580... Training loss: 1.2874... 0.0783 sec/batch
Epoch: 14/20... Training Step: 13581... Training loss: 1.3611... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13582... Training loss: 1.3595... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13583... Training loss: 1.3010... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13584... Training loss: 1.3120... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13585... Training loss: 1.2618... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13586... Training loss: 1.3590... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13587... Training loss: 1.3012... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13588... Training loss: 1.3082... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13589... Training loss: 1.2940... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13590... Training loss: 1.2414... 0.0792 sec/batch
Epoch: 14/20... Training Step: 13591... Training loss: 1.3202... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13592... Training loss: 1.3068... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13593... Training loss: 1.3636... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13594... Training loss: 1.3427... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13595... Training loss: 1.2984... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13596... Training loss: 1.3139... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13597... Training loss: 1.3125... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13598... Training loss: 1.3070... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13599... Training loss: 1.3059... 0.0793 sec/batch
Epoch: 14/20... Training Step: 13600... Training loss: 1.2583... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13601... Training loss: 1.2592... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13602... Training loss: 1.2438... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13603... Training loss: 1.3009... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13604... Training loss: 1.3277... 0.0826 sec/batch
Epoch: 14/20... Training Step: 13605... Training loss: 1.3457... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13606... Training loss: 1.2817... 0.0826 sec/batch
Epoch: 14/20... Training Step: 13607... Training loss: 1.2983... 0.0792 sec/batch
Epoch: 14/20... Training Step: 13608... Training loss: 1.3530... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13609... Training loss: 1.2729... 0.0791 sec/batch
Epoch: 14/20... Training Step: 13610... Training loss: 1.3026... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13611... Training loss: 1.3092... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13612... Training loss: 1.3516... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13613... Training loss: 1.3472... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13614... Training loss: 1.3356... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13615... Training loss: 1.3532... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13616... Training loss: 1.4221... 0.0824 sec/batch
Epoch: 14/20... Training Step: 13617... Training loss: 1.3404... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13618... Training loss: 1.3196... 0.0789 sec/batch
Epoch: 14/20... Training Step: 13619... Training loss: 1.2903... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13620... Training loss: 1.3493... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13621... Training loss: 1.3608... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13622... Training loss: 1.4013... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13623... Training loss: 1.3077... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13624... Training loss: 1.3694... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13625... Training loss: 1.3358... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13626... Training loss: 1.2746... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13627... Training loss: 1.2986... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13628... Training loss: 1.3387... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13629... Training loss: 1.2970... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13630... Training loss: 1.3312... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13631... Training loss: 1.3078... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13632... Training loss: 1.3076... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13633... Training loss: 1.2952... 0.0738 sec/batch
Epoch: 14/20... Training Step: 13634... Training loss: 1.3911... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13635... Training loss: 1.2932... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13636... Training loss: 1.3298... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13637... Training loss: 1.2603... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13638... Training loss: 1.3044... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13639... Training loss: 1.2967... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13640... Training loss: 1.3567... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13641... Training loss: 1.3340... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13642... Training loss: 1.3156... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13643... Training loss: 1.3311... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13644... Training loss: 1.3175... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13645... Training loss: 1.2499... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13646... Training loss: 1.2688... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13647... Training loss: 1.3904... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13648... Training loss: 1.2753... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13649... Training loss: 1.3463... 0.0798 sec/batch
Epoch: 14/20... Training Step: 13650... Training loss: 1.3350... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13651... Training loss: 1.2410... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13652... Training loss: 1.3221... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13653... Training loss: 1.3125... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13654... Training loss: 1.3584... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13655... Training loss: 1.3658... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13656... Training loss: 1.3979... 0.0805 sec/batch
Epoch: 14/20... Training Step: 13657... Training loss: 1.3192... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13658... Training loss: 1.2692... 0.0875 sec/batch
Epoch: 14/20... Training Step: 13659... Training loss: 1.3362... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13660... Training loss: 1.3018... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13661... Training loss: 1.3502... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13662... Training loss: 1.3652... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13663... Training loss: 1.3139... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13664... Training loss: 1.2837... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13665... Training loss: 1.2740... 0.0746 sec/batch
Epoch: 14/20... Training Step: 13666... Training loss: 1.3345... 0.0787 sec/batch
Epoch: 14/20... Training Step: 13667... Training loss: 1.3346... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13668... Training loss: 1.3882... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13669... Training loss: 1.3623... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13670... Training loss: 1.3428... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13671... Training loss: 1.3178... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13672... Training loss: 1.2282... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13673... Training loss: 1.2550... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13674... Training loss: 1.3162... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13675... Training loss: 1.2786... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13676... Training loss: 1.3148... 0.0791 sec/batch
Epoch: 14/20... Training Step: 13677... Training loss: 1.3011... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13678... Training loss: 1.2995... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13679... Training loss: 1.3486... 0.0845 sec/batch
Epoch: 14/20... Training Step: 13680... Training loss: 1.3349... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13681... Training loss: 1.3109... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13682... Training loss: 1.3092... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13683... Training loss: 1.3080... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13684... Training loss: 1.3321... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13685... Training loss: 1.3222... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13686... Training loss: 1.3444... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13687... Training loss: 1.2960... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13688... Training loss: 1.2940... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13689... Training loss: 1.3034... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13690... Training loss: 1.2695... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13691... Training loss: 1.2393... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13692... Training loss: 1.3104... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13693... Training loss: 1.2699... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13694... Training loss: 1.3045... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13695... Training loss: 1.2828... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13696... Training loss: 1.3012... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13697... Training loss: 1.3512... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13698... Training loss: 1.3513... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13699... Training loss: 1.3777... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13700... Training loss: 1.2967... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13701... Training loss: 1.2688... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13702... Training loss: 1.3093... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13703... Training loss: 1.2753... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13704... Training loss: 1.3342... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13705... Training loss: 1.3518... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13706... Training loss: 1.3671... 0.0827 sec/batch
Epoch: 14/20... Training Step: 13707... Training loss: 1.3213... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13708... Training loss: 1.3427... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13709... Training loss: 1.3061... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13710... Training loss: 1.3686... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13711... Training loss: 1.3418... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13712... Training loss: 1.3599... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13713... Training loss: 1.3868... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13714... Training loss: 1.4288... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13715... Training loss: 1.3702... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13716... Training loss: 1.2946... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13717... Training loss: 1.3053... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13718... Training loss: 1.2552... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13719... Training loss: 1.3119... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13720... Training loss: 1.2962... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13721... Training loss: 1.3191... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13722... Training loss: 1.3298... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13723... Training loss: 1.3432... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13724... Training loss: 1.3644... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13725... Training loss: 1.3814... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13726... Training loss: 1.2856... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13727... Training loss: 1.3538... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13728... Training loss: 1.3203... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13729... Training loss: 1.4031... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13730... Training loss: 1.3350... 0.0797 sec/batch
Epoch: 14/20... Training Step: 13731... Training loss: 1.2557... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13732... Training loss: 1.2942... 0.0745 sec/batch
Epoch: 14/20... Training Step: 13733... Training loss: 1.3179... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13734... Training loss: 1.3339... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13735... Training loss: 1.3485... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13736... Training loss: 1.3169... 0.0799 sec/batch
Epoch: 14/20... Training Step: 13737... Training loss: 1.3357... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13738... Training loss: 1.4525... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13739... Training loss: 1.3706... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13740... Training loss: 1.3781... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13741... Training loss: 1.3601... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13742... Training loss: 1.2467... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13743... Training loss: 1.3532... 0.0735 sec/batch
Epoch: 14/20... Training Step: 13744... Training loss: 1.3096... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13745... Training loss: 1.3181... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13746... Training loss: 1.3485... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13747... Training loss: 1.2945... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13748... Training loss: 1.4139... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13749... Training loss: 1.4026... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13750... Training loss: 1.3117... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13751... Training loss: 1.3175... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13752... Training loss: 1.3268... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13753... Training loss: 1.3136... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13754... Training loss: 1.3094... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13755... Training loss: 1.3272... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13756... Training loss: 1.3581... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13757... Training loss: 1.3310... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13758... Training loss: 1.2809... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13759... Training loss: 1.2592... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13760... Training loss: 1.3091... 0.0788 sec/batch
Epoch: 14/20... Training Step: 13761... Training loss: 1.2820... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13762... Training loss: 1.3415... 0.0806 sec/batch
Epoch: 14/20... Training Step: 13763... Training loss: 1.3280... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13764... Training loss: 1.3472... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13765... Training loss: 1.3384... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13766... Training loss: 1.3876... 0.0781 sec/batch
Epoch: 14/20... Training Step: 13767... Training loss: 1.3530... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13768... Training loss: 1.3274... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13769... Training loss: 1.3113... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13770... Training loss: 1.3216... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13771... Training loss: 1.3405... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13772... Training loss: 1.3345... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13773... Training loss: 1.3230... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13774... Training loss: 1.3643... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13775... Training loss: 1.3074... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13776... Training loss: 1.2980... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13777... Training loss: 1.3235... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13778... Training loss: 1.2783... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13779... Training loss: 1.3421... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13780... Training loss: 1.3204... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13781... Training loss: 1.3142... 0.0809 sec/batch
Epoch: 14/20... Training Step: 13782... Training loss: 1.3665... 0.0774 sec/batch
Epoch: 14/20... Training Step: 13783... Training loss: 1.3128... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13784... Training loss: 1.3395... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13785... Training loss: 1.3396... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13786... Training loss: 1.4004... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13787... Training loss: 1.3895... 0.0807 sec/batch
Epoch: 14/20... Training Step: 13788... Training loss: 1.2671... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13789... Training loss: 1.2996... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13790... Training loss: 1.2606... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13791... Training loss: 1.3511... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13792... Training loss: 1.2692... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13793... Training loss: 1.3439... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13794... Training loss: 1.3572... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13795... Training loss: 1.2973... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13796... Training loss: 1.2458... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13797... Training loss: 1.2823... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13798... Training loss: 1.3038... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13799... Training loss: 1.2757... 0.0821 sec/batch
Epoch: 14/20... Training Step: 13800... Training loss: 1.2791... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13801... Training loss: 1.2407... 0.0758 sec/batch
Epoch: 14/20... Training Step: 13802... Training loss: 1.3257... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13803... Training loss: 1.2422... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13804... Training loss: 1.3121... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13805... Training loss: 1.2456... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13806... Training loss: 1.2244... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13807... Training loss: 1.2653... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13808... Training loss: 1.3197... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13809... Training loss: 1.4189... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13810... Training loss: 1.3192... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13811... Training loss: 1.2898... 0.0749 sec/batch
Epoch: 14/20... Training Step: 13812... Training loss: 1.3058... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13813... Training loss: 1.2997... 0.0820 sec/batch
Epoch: 14/20... Training Step: 13814... Training loss: 1.3506... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13815... Training loss: 1.3065... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13816... Training loss: 1.3520... 0.0827 sec/batch
Epoch: 14/20... Training Step: 13817... Training loss: 1.4410... 0.0750 sec/batch
Epoch: 14/20... Training Step: 13818... Training loss: 1.3352... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13819... Training loss: 1.3153... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13820... Training loss: 1.2923... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13821... Training loss: 1.3379... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13822... Training loss: 1.3746... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13823... Training loss: 1.3073... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13824... Training loss: 1.3317... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13825... Training loss: 1.3781... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13826... Training loss: 1.3724... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13827... Training loss: 1.3595... 0.0756 sec/batch
Epoch: 14/20... Training Step: 13828... Training loss: 1.4204... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13829... Training loss: 1.2972... 0.0759 sec/batch
Epoch: 14/20... Training Step: 13830... Training loss: 1.3493... 0.0771 sec/batch
Epoch: 14/20... Training Step: 13831... Training loss: 1.3369... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13832... Training loss: 1.3430... 0.0744 sec/batch
Epoch: 14/20... Training Step: 13833... Training loss: 1.3042... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13834... Training loss: 1.3359... 0.0790 sec/batch
Epoch: 14/20... Training Step: 13835... Training loss: 1.3181... 0.0741 sec/batch
Epoch: 14/20... Training Step: 13836... Training loss: 1.3975... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13837... Training loss: 1.3297... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13838... Training loss: 1.3565... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13839... Training loss: 1.3823... 0.0748 sec/batch
Epoch: 14/20... Training Step: 13840... Training loss: 1.3551... 0.0847 sec/batch
Epoch: 14/20... Training Step: 13841... Training loss: 1.3666... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13842... Training loss: 1.3154... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13843... Training loss: 1.2907... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13844... Training loss: 1.2861... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13845... Training loss: 1.3296... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13846... Training loss: 1.2522... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13847... Training loss: 1.3153... 0.0757 sec/batch
Epoch: 14/20... Training Step: 13848... Training loss: 1.2660... 0.0778 sec/batch
Epoch: 14/20... Training Step: 13849... Training loss: 1.3010... 0.0769 sec/batch
Epoch: 14/20... Training Step: 13850... Training loss: 1.3452... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13851... Training loss: 1.3595... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13852... Training loss: 1.3012... 0.0772 sec/batch
Epoch: 14/20... Training Step: 13853... Training loss: 1.3215... 0.0768 sec/batch
Epoch: 14/20... Training Step: 13854... Training loss: 1.3433... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13855... Training loss: 1.2823... 0.0789 sec/batch
Epoch: 14/20... Training Step: 13856... Training loss: 1.2787... 0.0779 sec/batch
Epoch: 14/20... Training Step: 13857... Training loss: 1.3301... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13858... Training loss: 1.2923... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13859... Training loss: 1.3266... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13860... Training loss: 1.3113... 0.0743 sec/batch
Epoch: 14/20... Training Step: 13861... Training loss: 1.3776... 0.0770 sec/batch
Epoch: 14/20... Training Step: 13862... Training loss: 1.3685... 0.0777 sec/batch
Epoch: 14/20... Training Step: 13863... Training loss: 1.3698... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13864... Training loss: 1.3655... 0.0753 sec/batch
Epoch: 14/20... Training Step: 13865... Training loss: 1.3581... 0.0755 sec/batch
Epoch: 14/20... Training Step: 13866... Training loss: 1.4923... 0.0784 sec/batch
Epoch: 14/20... Training Step: 13867... Training loss: 1.3118... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13868... Training loss: 1.4000... 0.0767 sec/batch
Epoch: 14/20... Training Step: 13869... Training loss: 1.4226... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13870... Training loss: 1.2935... 0.0765 sec/batch
Epoch: 14/20... Training Step: 13871... Training loss: 1.3423... 0.0762 sec/batch
Epoch: 14/20... Training Step: 13872... Training loss: 1.3689... 0.0764 sec/batch
Epoch: 14/20... Training Step: 13873... Training loss: 1.3828... 0.0760 sec/batch
Epoch: 14/20... Training Step: 13874... Training loss: 1.3748... 0.0754 sec/batch
Epoch: 14/20... Training Step: 13875... Training loss: 1.3672... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13876... Training loss: 1.2898... 0.0780 sec/batch
Epoch: 14/20... Training Step: 13877... Training loss: 1.2549... 0.0747 sec/batch
Epoch: 14/20... Training Step: 13878... Training loss: 1.4043... 0.0766 sec/batch
Epoch: 14/20... Training Step: 13879... Training loss: 1.3393... 0.0776 sec/batch
Epoch: 14/20... Training Step: 13880... Training loss: 1.3585... 0.0752 sec/batch
Epoch: 14/20... Training Step: 13881... Training loss: 1.3287... 0.0761 sec/batch
Epoch: 14/20... Training Step: 13882... Training loss: 1.2809... 0.0782 sec/batch
Epoch: 14/20... Training Step: 13883... Training loss: 1.2872... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13884... Training loss: 1.3349... 0.0775 sec/batch
Epoch: 14/20... Training Step: 13885... Training loss: 1.3519... 0.0773 sec/batch
Epoch: 14/20... Training Step: 13886... Training loss: 1.2172... 0.0763 sec/batch
Epoch: 14/20... Training Step: 13887... Training loss: 1.3093... 0.0751 sec/batch
Epoch: 14/20... Training Step: 13888... Training loss: 1.3571... 0.0779 sec/batch
Epoch: 15/20... Training Step: 13889... Training loss: 1.4089... 0.0756 sec/batch
Epoch: 15/20... Training Step: 13890... Training loss: 1.2703... 0.0744 sec/batch
Epoch: 15/20... Training Step: 13891... Training loss: 1.3753... 0.0781 sec/batch
Epoch: 15/20... Training Step: 13892... Training loss: 1.3418... 0.0777 sec/batch
Epoch: 15/20... Training Step: 13893... Training loss: 1.2599... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13894... Training loss: 1.1975... 0.0778 sec/batch
Epoch: 15/20... Training Step: 13895... Training loss: 1.2829... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13896... Training loss: 1.3253... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13897... Training loss: 1.3479... 0.0758 sec/batch
Epoch: 15/20... Training Step: 13898... Training loss: 1.2328... 0.0749 sec/batch
Epoch: 15/20... Training Step: 13899... Training loss: 1.3891... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13900... Training loss: 1.2847... 0.0754 sec/batch
Epoch: 15/20... Training Step: 13901... Training loss: 1.3017... 0.0749 sec/batch
Epoch: 15/20... Training Step: 13902... Training loss: 1.3072... 0.0748 sec/batch
Epoch: 15/20... Training Step: 13903... Training loss: 1.3615... 0.0792 sec/batch
Epoch: 15/20... Training Step: 13904... Training loss: 1.2713... 0.0755 sec/batch
Epoch: 15/20... Training Step: 13905... Training loss: 1.2887... 0.0772 sec/batch
Epoch: 15/20... Training Step: 13906... Training loss: 1.3393... 0.0794 sec/batch
Epoch: 15/20... Training Step: 13907... Training loss: 1.3089... 0.0773 sec/batch
Epoch: 15/20... Training Step: 13908... Training loss: 1.2921... 0.0770 sec/batch
Epoch: 15/20... Training Step: 13909... Training loss: 1.3090... 0.0828 sec/batch
Epoch: 15/20... Training Step: 13910... Training loss: 1.2809... 0.0752 sec/batch
Epoch: 15/20... Training Step: 13911... Training loss: 1.2910... 0.0756 sec/batch
Epoch: 15/20... Training Step: 13912... Training loss: 1.2442... 0.0776 sec/batch
Epoch: 15/20... Training Step: 13913... Training loss: 1.3205... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13914... Training loss: 1.2663... 0.0743 sec/batch
Epoch: 15/20... Training Step: 13915... Training loss: 1.2969... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13916... Training loss: 1.2443... 0.0767 sec/batch
Epoch: 15/20... Training Step: 13917... Training loss: 1.3141... 0.0752 sec/batch
Epoch: 15/20... Training Step: 13918... Training loss: 1.3600... 0.0767 sec/batch
Epoch: 15/20... Training Step: 13919... Training loss: 1.3353... 0.0742 sec/batch
Epoch: 15/20... Training Step: 13920... Training loss: 1.3347... 0.0750 sec/batch
Epoch: 15/20... Training Step: 13921... Training loss: 1.2892... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13922... Training loss: 1.3460... 0.0758 sec/batch
Epoch: 15/20... Training Step: 13923... Training loss: 1.4234... 0.0753 sec/batch
Epoch: 15/20... Training Step: 13924... Training loss: 1.3584... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13925... Training loss: 1.3155... 0.0770 sec/batch
Epoch: 15/20... Training Step: 13926... Training loss: 1.3982... 0.0766 sec/batch
Epoch: 15/20... Training Step: 13927... Training loss: 1.3191... 0.0766 sec/batch
Epoch: 15/20... Training Step: 13928... Training loss: 1.2811... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13929... Training loss: 1.3179... 0.0762 sec/batch
Epoch: 15/20... Training Step: 13930... Training loss: 1.3124... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13931... Training loss: 1.2770... 0.0775 sec/batch
Epoch: 15/20... Training Step: 13932... Training loss: 1.2756... 0.0754 sec/batch
Epoch: 15/20... Training Step: 13933... Training loss: 1.3482... 0.0747 sec/batch
Epoch: 15/20... Training Step: 13934... Training loss: 1.2863... 0.0791 sec/batch
Epoch: 15/20... Training Step: 13935... Training loss: 1.3700... 0.0762 sec/batch
Epoch: 15/20... Training Step: 13936... Training loss: 1.2815... 0.0777 sec/batch
Epoch: 15/20... Training Step: 13937... Training loss: 1.3487... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13938... Training loss: 1.3220... 0.0830 sec/batch
Epoch: 15/20... Training Step: 13939... Training loss: 1.2374... 0.0749 sec/batch
Epoch: 15/20... Training Step: 13940... Training loss: 1.3723... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13941... Training loss: 1.2420... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13942... Training loss: 1.2661... 0.0775 sec/batch
Epoch: 15/20... Training Step: 13943... Training loss: 1.2875... 0.0748 sec/batch
Epoch: 15/20... Training Step: 13944... Training loss: 1.2636... 0.0755 sec/batch
Epoch: 15/20... Training Step: 13945... Training loss: 1.3152... 0.0793 sec/batch
Epoch: 15/20... Training Step: 13946... Training loss: 1.3650... 0.0749 sec/batch
Epoch: 15/20... Training Step: 13947... Training loss: 1.3158... 0.0755 sec/batch
Epoch: 15/20... Training Step: 13948... Training loss: 1.3265... 0.0755 sec/batch
Epoch: 15/20... Training Step: 13949... Training loss: 1.2281... 0.0783 sec/batch
Epoch: 15/20... Training Step: 13950... Training loss: 1.2486... 0.0781 sec/batch
Epoch: 15/20... Training Step: 13951... Training loss: 1.3199... 0.0764 sec/batch
Epoch: 15/20... Training Step: 13952... Training loss: 1.2624... 0.0748 sec/batch
Epoch: 15/20... Training Step: 13953... Training loss: 1.3041... 0.0740 sec/batch
Epoch: 15/20... Training Step: 13954... Training loss: 1.3005... 0.0768 sec/batch
Epoch: 15/20... Training Step: 13955... Training loss: 1.2773... 0.0748 sec/batch
Epoch: 15/20... Training Step: 13956... Training loss: 1.3030... 0.0757 sec/batch
Epoch: 15/20... Training Step: 13957... Training loss: 1.3392... 0.0771 sec/batch
Epoch: 15/20... Training Step: 13958... Training loss: 1.3238... 0.0766 sec/batch
Epoch: 15/20... Training Step: 13959... Training loss: 1.3200... 0.0777 sec/batch
Epoch: 15/20... Training Step: 13960... Training loss: 1.3650... 0.0794 sec/batch
Epoch: 15/20... Training Step: 13961... Training loss: 1.3205... 0.0758 sec/batch
Epoch: 15/20... Training Step: 13962... Training loss: 1.2535... 0.0767 sec/batch
Epoch: 15/20... Training Step: 13963... Training loss: 1.2729... 0.0753 sec/batch
Epoch: 15/20... Training Step: 13964... Training loss: 1.2744... 0.0754 sec/batch
Epoch: 15/20... Training Step: 13965... Training loss: 1.2784... 0.0761 sec/batch
Epoch: 15/20... Training Step: 13966... Training loss: 1.2803... 0.0775 sec/batch
Epoch: 15/20... Training Step: 13967... Training loss: 1.3127... 0.0747 sec/batch
Epoch: 15/20... Training Step: 13968... Training loss: 1.3505... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13969... Training loss: 1.3207... 0.0778 sec/batch
Epoch: 15/20... Training Step: 13970... Training loss: 1.2815... 0.0754 sec/batch
Epoch: 15/20... Training Step: 13971... Training loss: 1.3083... 0.0785 sec/batch
Epoch: 15/20... Training Step: 13972... Training loss: 1.3205... 0.0770 sec/batch
Epoch: 15/20... Training Step: 13973... Training loss: 1.2868... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13974... Training loss: 1.3569... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13975... Training loss: 1.2966... 0.0799 sec/batch
Epoch: 15/20... Training Step: 13976... Training loss: 1.3152... 0.0749 sec/batch
Epoch: 15/20... Training Step: 13977... Training loss: 1.3533... 0.0770 sec/batch
Epoch: 15/20... Training Step: 13978... Training loss: 1.3424... 0.0760 sec/batch
Epoch: 15/20... Training Step: 13979... Training loss: 1.3600... 0.0782 sec/batch
Epoch: 15/20... Training Step: 13980... Training loss: 1.2889... 0.0780 sec/batch
Epoch: 15/20... Training Step: 13981... Training loss: 1.2777... 0.0770 sec/batch
Epoch: 15/20... Training Step: 13982... Training loss: 1.2354... 0.0774 sec/batch
Epoch: 15/20... Training Step: 13983... Training loss: 1.2218... 0.0780 sec/batch
Epoch: 15/20... Training Step: 13984... Training loss: 1.3357... 0.0768 sec/batch
Epoch: 15/20... Training Step: 13985... Training loss: 1.2936... 0.0761 sec/batch
Epoch: 15/20... Training Step: 13986... Training loss: 1.2535... 0.0835 sec/batch
Epoch: 15/20... Training Step: 13987... Training loss: 1.2721... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13988... Training loss: 1.2663... 0.0763 sec/batch
Epoch: 15/20... Training Step: 13989... Training loss: 1.2737... 0.0790 sec/batch
Epoch: 15/20... Training Step: 13990... Training loss: 1.2716... 0.0773 sec/batch
Epoch: 15/20... Training Step: 13991... Training loss: 1.3082... 0.0762 sec/batch
Epoch: 15/20... Training Step: 13992... Training loss: 1.3301... 0.0742 sec/batch
Epoch: 15/20... Training Step: 13993... Training loss: 1.3194... 0.0788 sec/batch
Epoch: 15/20... Training Step: 13994... Training loss: 1.2862... 0.0777 sec/batch
Epoch: 15/20... Training Step: 13995... Training loss: 1.2423... 0.0767 sec/batch
Epoch: 15/20... Training Step: 13996... Training loss: 1.3371... 0.0768 sec/batch
Epoch: 15/20... Training Step: 13997... Training loss: 1.3198... 0.0757 sec/batch
Epoch: 15/20... Training Step: 13998... Training loss: 1.3506... 0.0764 sec/batch
Epoch: 15/20... Training Step: 13999... Training loss: 1.2962... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14000... Training loss: 1.2776... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14001... Training loss: 1.3887... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14002... Training loss: 1.3286... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14003... Training loss: 1.2803... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14004... Training loss: 1.3953... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14005... Training loss: 1.3503... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14006... Training loss: 1.4130... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14007... Training loss: 1.3232... 0.0865 sec/batch
Epoch: 15/20... Training Step: 14008... Training loss: 1.3106... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14009... Training loss: 1.4242... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14010... Training loss: 1.2967... 0.0787 sec/batch
Epoch: 15/20... Training Step: 14011... Training loss: 1.3436... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14012... Training loss: 1.3038... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14013... Training loss: 1.3652... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14014... Training loss: 1.3588... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14015... Training loss: 1.3576... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14016... Training loss: 1.3241... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14017... Training loss: 1.2534... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14018... Training loss: 1.3423... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14019... Training loss: 1.2901... 0.0784 sec/batch
Epoch: 15/20... Training Step: 14020... Training loss: 1.3601... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14021... Training loss: 1.3305... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14022... Training loss: 1.3081... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14023... Training loss: 1.3528... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14024... Training loss: 1.3436... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14025... Training loss: 1.3524... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14026... Training loss: 1.3910... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14027... Training loss: 1.3697... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14028... Training loss: 1.3270... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14029... Training loss: 1.3201... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14030... Training loss: 1.2817... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14031... Training loss: 1.3066... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14032... Training loss: 1.3122... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14033... Training loss: 1.1946... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14034... Training loss: 1.3492... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14035... Training loss: 1.3126... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14036... Training loss: 1.2544... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14037... Training loss: 1.2811... 0.0780 sec/batch
Epoch: 15/20... Training Step: 14038... Training loss: 1.2675... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14039... Training loss: 1.2097... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14040... Training loss: 1.3147... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14041... Training loss: 1.2414... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14042... Training loss: 1.2802... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14043... Training loss: 1.2669... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14044... Training loss: 1.2675... 0.0791 sec/batch
Epoch: 15/20... Training Step: 14045... Training loss: 1.2978... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14046... Training loss: 1.2919... 0.0828 sec/batch
Epoch: 15/20... Training Step: 14047... Training loss: 1.3151... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14048... Training loss: 1.3270... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14049... Training loss: 1.2912... 0.0842 sec/batch
Epoch: 15/20... Training Step: 14050... Training loss: 1.3115... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14051... Training loss: 1.3117... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14052... Training loss: 1.3614... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14053... Training loss: 1.3256... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14054... Training loss: 1.4070... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14055... Training loss: 1.3265... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14056... Training loss: 1.3123... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14057... Training loss: 1.3227... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14058... Training loss: 1.3221... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14059... Training loss: 1.2953... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14060... Training loss: 1.2974... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14061... Training loss: 1.3248... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14062... Training loss: 1.3627... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14063... Training loss: 1.2723... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14064... Training loss: 1.3458... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14065... Training loss: 1.3140... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14066... Training loss: 1.4419... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14067... Training loss: 1.3554... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14068... Training loss: 1.2932... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14069... Training loss: 1.3273... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14070... Training loss: 1.3853... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14071... Training loss: 1.3052... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14072... Training loss: 1.3295... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14073... Training loss: 1.3383... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14074... Training loss: 1.3645... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14075... Training loss: 1.3115... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14076... Training loss: 1.2995... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14077... Training loss: 1.3079... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14078... Training loss: 1.3552... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14079... Training loss: 1.3819... 0.0841 sec/batch
Epoch: 15/20... Training Step: 14080... Training loss: 1.3380... 0.0790 sec/batch
Epoch: 15/20... Training Step: 14081... Training loss: 1.3679... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14082... Training loss: 1.3923... 0.0829 sec/batch
Epoch: 15/20... Training Step: 14083... Training loss: 1.3263... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14084... Training loss: 1.2719... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14085... Training loss: 1.3665... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14086... Training loss: 1.3259... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14087... Training loss: 1.3552... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14088... Training loss: 1.3079... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14089... Training loss: 1.3079... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14090... Training loss: 1.3105... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14091... Training loss: 1.3078... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14092... Training loss: 1.3066... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14093... Training loss: 1.2669... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14094... Training loss: 1.3816... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14095... Training loss: 1.3259... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14096... Training loss: 1.3050... 0.0790 sec/batch
Epoch: 15/20... Training Step: 14097... Training loss: 1.3541... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14098... Training loss: 1.3405... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14099... Training loss: 1.2506... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14100... Training loss: 1.3672... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14101... Training loss: 1.2795... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14102... Training loss: 1.3575... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14103... Training loss: 1.3112... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14104... Training loss: 1.3284... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14105... Training loss: 1.2868... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14106... Training loss: 1.3464... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14107... Training loss: 1.3192... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14108... Training loss: 1.2671... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14109... Training loss: 1.2668... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14110... Training loss: 1.4148... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14111... Training loss: 1.3178... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14112... Training loss: 1.3300... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14113... Training loss: 1.3502... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14114... Training loss: 1.2756... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14115... Training loss: 1.3159... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14116... Training loss: 1.3413... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14117... Training loss: 1.2179... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14118... Training loss: 1.2630... 0.0796 sec/batch
Epoch: 15/20... Training Step: 14119... Training loss: 1.2959... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14120... Training loss: 1.3115... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14121... Training loss: 1.3139... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14122... Training loss: 1.4000... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14123... Training loss: 1.2972... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14124... Training loss: 1.3176... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14125... Training loss: 1.3032... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14126... Training loss: 1.3551... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14127... Training loss: 1.3674... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14128... Training loss: 1.3055... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14129... Training loss: 1.3228... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14130... Training loss: 1.3408... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14131... Training loss: 1.3714... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14132... Training loss: 1.3533... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14133... Training loss: 1.3118... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14134... Training loss: 1.3762... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14135... Training loss: 1.2534... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14136... Training loss: 1.2248... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14137... Training loss: 1.2996... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14138... Training loss: 1.3229... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14139... Training loss: 1.2985... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14140... Training loss: 1.3134... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14141... Training loss: 1.2936... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14142... Training loss: 1.2840... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14143... Training loss: 1.2507... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14144... Training loss: 1.3076... 0.0781 sec/batch
Epoch: 15/20... Training Step: 14145... Training loss: 1.3667... 0.0815 sec/batch
Epoch: 15/20... Training Step: 14146... Training loss: 1.2969... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14147... Training loss: 1.3128... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14148... Training loss: 1.3239... 0.0852 sec/batch
Epoch: 15/20... Training Step: 14149... Training loss: 1.3350... 0.0792 sec/batch
Epoch: 15/20... Training Step: 14150... Training loss: 1.3638... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14151... Training loss: 1.3422... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14152... Training loss: 1.3396... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14153... Training loss: 1.3477... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14154... Training loss: 1.2039... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14155... Training loss: 1.2756... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14156... Training loss: 1.3090... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14157... Training loss: 1.2367... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14158... Training loss: 1.2638... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14159... Training loss: 1.2994... 0.0815 sec/batch
Epoch: 15/20... Training Step: 14160... Training loss: 1.2771... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14161... Training loss: 1.2189... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14162... Training loss: 1.2279... 0.0740 sec/batch
Epoch: 15/20... Training Step: 14163... Training loss: 1.2878... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14164... Training loss: 1.3049... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14165... Training loss: 1.3100... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14166... Training loss: 1.1828... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14167... Training loss: 1.2626... 0.0803 sec/batch
Epoch: 15/20... Training Step: 14168... Training loss: 1.3645... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14169... Training loss: 1.2965... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14170... Training loss: 1.2782... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14171... Training loss: 1.3396... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14172... Training loss: 1.2769... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14173... Training loss: 1.1930... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14174... Training loss: 1.2420... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14175... Training loss: 1.3131... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14176... Training loss: 1.2676... 0.0739 sec/batch
Epoch: 15/20... Training Step: 14177... Training loss: 1.2798... 0.0791 sec/batch
Epoch: 15/20... Training Step: 14178... Training loss: 1.2645... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14179... Training loss: 1.2826... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14180... Training loss: 1.3151... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14181... Training loss: 1.2497... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14182... Training loss: 1.2532... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14183... Training loss: 1.3027... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14184... Training loss: 1.3110... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14185... Training loss: 1.2673... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14186... Training loss: 1.3897... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14187... Training loss: 1.3052... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14188... Training loss: 1.3319... 0.0739 sec/batch
Epoch: 15/20... Training Step: 14189... Training loss: 1.3671... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14190... Training loss: 1.2620... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14191... Training loss: 1.2805... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14192... Training loss: 1.3436... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14193... Training loss: 1.3086... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14194... Training loss: 1.3719... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14195... Training loss: 1.2712... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14196... Training loss: 1.2408... 0.0797 sec/batch
Epoch: 15/20... Training Step: 14197... Training loss: 1.4011... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14198... Training loss: 1.2935... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14199... Training loss: 1.3799... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14200... Training loss: 1.3152... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14201... Training loss: 1.2940... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14202... Training loss: 1.2813... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14203... Training loss: 1.3097... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14204... Training loss: 1.2667... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14205... Training loss: 1.2702... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14206... Training loss: 1.2411... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14207... Training loss: 1.2982... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14208... Training loss: 1.3234... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14209... Training loss: 1.3106... 0.0780 sec/batch
Epoch: 15/20... Training Step: 14210... Training loss: 1.2351... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14211... Training loss: 1.2835... 0.0787 sec/batch
Epoch: 15/20... Training Step: 14212... Training loss: 1.2837... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14213... Training loss: 1.3381... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14214... Training loss: 1.2939... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14215... Training loss: 1.3004... 0.0741 sec/batch
Epoch: 15/20... Training Step: 14216... Training loss: 1.2856... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14217... Training loss: 1.2715... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14218... Training loss: 1.3093... 0.0741 sec/batch
Epoch: 15/20... Training Step: 14219... Training loss: 1.3446... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14220... Training loss: 1.2436... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14221... Training loss: 1.3967... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14222... Training loss: 1.3978... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14223... Training loss: 1.3661... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14224... Training loss: 1.3281... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14225... Training loss: 1.2587... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14226... Training loss: 1.2994... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14227... Training loss: 1.2879... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14228... Training loss: 1.2956... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14229... Training loss: 1.2801... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14230... Training loss: 1.2709... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14231... Training loss: 1.2450... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14232... Training loss: 1.3211... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14233... Training loss: 1.2982... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14234... Training loss: 1.3238... 0.0780 sec/batch
Epoch: 15/20... Training Step: 14235... Training loss: 1.3396... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14236... Training loss: 1.3458... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14237... Training loss: 1.3668... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14238... Training loss: 1.3603... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14239... Training loss: 1.3281... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14240... Training loss: 1.2944... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14241... Training loss: 1.2919... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14242... Training loss: 1.3133... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14243... Training loss: 1.3443... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14244... Training loss: 1.3234... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14245... Training loss: 1.2914... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14246... Training loss: 1.2973... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14247... Training loss: 1.3043... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14248... Training loss: 1.3041... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14249... Training loss: 1.3099... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14250... Training loss: 1.3840... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14251... Training loss: 1.2828... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14252... Training loss: 1.3259... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14253... Training loss: 1.3015... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14254... Training loss: 1.2695... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14255... Training loss: 1.2953... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14256... Training loss: 1.2313... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14257... Training loss: 1.2780... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14258... Training loss: 1.2447... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14259... Training loss: 1.3625... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14260... Training loss: 1.2432... 0.0803 sec/batch
Epoch: 15/20... Training Step: 14261... Training loss: 1.2814... 0.0798 sec/batch
Epoch: 15/20... Training Step: 14262... Training loss: 1.2286... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14263... Training loss: 1.2822... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14264... Training loss: 1.2612... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14265... Training loss: 1.2736... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14266... Training loss: 1.3170... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14267... Training loss: 1.2903... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14268... Training loss: 1.3158... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14269... Training loss: 1.3016... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14270... Training loss: 1.2967... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14271... Training loss: 1.3330... 0.0740 sec/batch
Epoch: 15/20... Training Step: 14272... Training loss: 1.3090... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14273... Training loss: 1.3418... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14274... Training loss: 1.3172... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14275... Training loss: 1.3367... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14276... Training loss: 1.3159... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14277... Training loss: 1.3599... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14278... Training loss: 1.2684... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14279... Training loss: 1.3192... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14280... Training loss: 1.2985... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14281... Training loss: 1.3096... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14282... Training loss: 1.2634... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14283... Training loss: 1.3031... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14284... Training loss: 1.2812... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14285... Training loss: 1.2957... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14286... Training loss: 1.3489... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14287... Training loss: 1.3072... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14288... Training loss: 1.2985... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14289... Training loss: 1.2879... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14290... Training loss: 1.3212... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14291... Training loss: 1.2811... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14292... Training loss: 1.3506... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14293... Training loss: 1.2905... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14294... Training loss: 1.3104... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14295... Training loss: 1.2367... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14296... Training loss: 1.3650... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14297... Training loss: 1.2414... 0.0850 sec/batch
Epoch: 15/20... Training Step: 14298... Training loss: 1.2700... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14299... Training loss: 1.2002... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14300... Training loss: 1.2719... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14301... Training loss: 1.2804... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14302... Training loss: 1.2251... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14303... Training loss: 1.2555... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14304... Training loss: 1.3210... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14305... Training loss: 1.2352... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14306... Training loss: 1.2750... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14307... Training loss: 1.3276... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14308... Training loss: 1.3580... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14309... Training loss: 1.2387... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14310... Training loss: 1.3097... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14311... Training loss: 1.2972... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14312... Training loss: 1.3188... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14313... Training loss: 1.3167... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14314... Training loss: 1.3010... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14315... Training loss: 1.3520... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14316... Training loss: 1.3820... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14317... Training loss: 1.3677... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14318... Training loss: 1.3898... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14319... Training loss: 1.3012... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14320... Training loss: 1.3276... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14321... Training loss: 1.3771... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14322... Training loss: 1.3387... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14323... Training loss: 1.3545... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14324... Training loss: 1.2879... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14325... Training loss: 1.3405... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14326... Training loss: 1.3978... 0.0741 sec/batch
Epoch: 15/20... Training Step: 14327... Training loss: 1.3281... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14328... Training loss: 1.3372... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14329... Training loss: 1.2941... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14330... Training loss: 1.3406... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14331... Training loss: 1.3193... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14332... Training loss: 1.3216... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14333... Training loss: 1.2780... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14334... Training loss: 1.4090... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14335... Training loss: 1.3236... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14336... Training loss: 1.3292... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14337... Training loss: 1.4127... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14338... Training loss: 1.4057... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14339... Training loss: 1.3214... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14340... Training loss: 1.3001... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14341... Training loss: 1.3686... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14342... Training loss: 1.3274... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14343... Training loss: 1.3449... 0.0901 sec/batch
Epoch: 15/20... Training Step: 14344... Training loss: 1.3028... 0.0788 sec/batch
Epoch: 15/20... Training Step: 14345... Training loss: 1.2943... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14346... Training loss: 1.2871... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14347... Training loss: 1.2928... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14348... Training loss: 1.3123... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14349... Training loss: 1.3326... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14350... Training loss: 1.3111... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14351... Training loss: 1.2855... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14352... Training loss: 1.3328... 0.0797 sec/batch
Epoch: 15/20... Training Step: 14353... Training loss: 1.2899... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14354... Training loss: 1.2628... 0.0798 sec/batch
Epoch: 15/20... Training Step: 14355... Training loss: 1.2225... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14356... Training loss: 1.3025... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14357... Training loss: 1.2888... 0.0799 sec/batch
Epoch: 15/20... Training Step: 14358... Training loss: 1.3739... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14359... Training loss: 1.3417... 0.0781 sec/batch
Epoch: 15/20... Training Step: 14360... Training loss: 1.3009... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14361... Training loss: 1.3019... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14362... Training loss: 1.3038... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14363... Training loss: 1.1921... 0.0817 sec/batch
Epoch: 15/20... Training Step: 14364... Training loss: 1.2570... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14365... Training loss: 1.3082... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14366... Training loss: 1.3014... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14367... Training loss: 1.2418... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14368... Training loss: 1.2541... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14369... Training loss: 1.2691... 0.0787 sec/batch
Epoch: 15/20... Training Step: 14370... Training loss: 1.3256... 0.0787 sec/batch
Epoch: 15/20... Training Step: 14371... Training loss: 1.2897... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14372... Training loss: 1.3000... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14373... Training loss: 1.2327... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14374... Training loss: 1.2773... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14375... Training loss: 1.3066... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14376... Training loss: 1.2727... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14377... Training loss: 1.2679... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14378... Training loss: 1.2667... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14379... Training loss: 1.2706... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14380... Training loss: 1.2924... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14381... Training loss: 1.2702... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14382... Training loss: 1.2978... 0.0737 sec/batch
Epoch: 15/20... Training Step: 14383... Training loss: 1.2590... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14384... Training loss: 1.2447... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14385... Training loss: 1.3017... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14386... Training loss: 1.2877... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14387... Training loss: 1.2611... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14388... Training loss: 1.2761... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14389... Training loss: 1.2604... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14390... Training loss: 1.2933... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14391... Training loss: 1.2770... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14392... Training loss: 1.2835... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14393... Training loss: 1.2961... 0.0808 sec/batch
Epoch: 15/20... Training Step: 14394... Training loss: 1.2404... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14395... Training loss: 1.2814... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14396... Training loss: 1.2665... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14397... Training loss: 1.2618... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14398... Training loss: 1.2594... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14399... Training loss: 1.3157... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14400... Training loss: 1.3134... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14401... Training loss: 1.3533... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14402... Training loss: 1.3425... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14403... Training loss: 1.3074... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14404... Training loss: 1.3174... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14405... Training loss: 1.3280... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14406... Training loss: 1.3082... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14407... Training loss: 1.3236... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14408... Training loss: 1.2527... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14409... Training loss: 1.2991... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14410... Training loss: 1.2380... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14411... Training loss: 1.3910... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14412... Training loss: 1.2944... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14413... Training loss: 1.3038... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14414... Training loss: 1.2946... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14415... Training loss: 1.2913... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14416... Training loss: 1.2632... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14417... Training loss: 1.3074... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14418... Training loss: 1.3721... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14419... Training loss: 1.4089... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14420... Training loss: 1.3393... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14421... Training loss: 1.3484... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14422... Training loss: 1.3461... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14423... Training loss: 1.2998... 0.0808 sec/batch
Epoch: 15/20... Training Step: 14424... Training loss: 1.2737... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14425... Training loss: 1.2662... 0.0796 sec/batch
Epoch: 15/20... Training Step: 14426... Training loss: 1.2667... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14427... Training loss: 1.2772... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14428... Training loss: 1.2849... 0.0844 sec/batch
Epoch: 15/20... Training Step: 14429... Training loss: 1.3715... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14430... Training loss: 1.2848... 0.0742 sec/batch
Epoch: 15/20... Training Step: 14431... Training loss: 1.2735... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14432... Training loss: 1.2839... 0.0740 sec/batch
Epoch: 15/20... Training Step: 14433... Training loss: 1.2414... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14434... Training loss: 1.2743... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14435... Training loss: 1.3446... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14436... Training loss: 1.3166... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14437... Training loss: 1.2620... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14438... Training loss: 1.2754... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14439... Training loss: 1.2827... 0.0788 sec/batch
Epoch: 15/20... Training Step: 14440... Training loss: 1.2366... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14441... Training loss: 1.2573... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14442... Training loss: 1.2626... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14443... Training loss: 1.3075... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14444... Training loss: 1.2675... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14445... Training loss: 1.2834... 0.0784 sec/batch
Epoch: 15/20... Training Step: 14446... Training loss: 1.2774... 0.0781 sec/batch
Epoch: 15/20... Training Step: 14447... Training loss: 1.3040... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14448... Training loss: 1.2191... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14449... Training loss: 1.2805... 0.0838 sec/batch
Epoch: 15/20... Training Step: 14450... Training loss: 1.2897... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14451... Training loss: 1.2627... 0.0800 sec/batch
Epoch: 15/20... Training Step: 14452... Training loss: 1.2846... 0.0803 sec/batch
Epoch: 15/20... Training Step: 14453... Training loss: 1.2983... 0.0818 sec/batch
Epoch: 15/20... Training Step: 14454... Training loss: 1.3163... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14455... Training loss: 1.2663... 0.0781 sec/batch
Epoch: 15/20... Training Step: 14456... Training loss: 1.2357... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14457... Training loss: 1.2102... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14458... Training loss: 1.2412... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14459... Training loss: 1.2424... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14460... Training loss: 1.2584... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14461... Training loss: 1.3598... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14462... Training loss: 1.3356... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14463... Training loss: 1.3279... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14464... Training loss: 1.3051... 0.0815 sec/batch
Epoch: 15/20... Training Step: 14465... Training loss: 1.2985... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14466... Training loss: 1.4124... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14467... Training loss: 1.3463... 0.0800 sec/batch
Epoch: 15/20... Training Step: 14468... Training loss: 1.2626... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14469... Training loss: 1.2309... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14470... Training loss: 1.2309... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14471... Training loss: 1.3361... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14472... Training loss: 1.2347... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14473... Training loss: 1.2954... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14474... Training loss: 1.3213... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14475... Training loss: 1.3094... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14476... Training loss: 1.3243... 0.0825 sec/batch
Epoch: 15/20... Training Step: 14477... Training loss: 1.2837... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14478... Training loss: 1.2899... 0.0739 sec/batch
Epoch: 15/20... Training Step: 14479... Training loss: 1.2859... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14480... Training loss: 1.3159... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14481... Training loss: 1.2475... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14482... Training loss: 1.2543... 0.0797 sec/batch
Epoch: 15/20... Training Step: 14483... Training loss: 1.2055... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14484... Training loss: 1.3214... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14485... Training loss: 1.1826... 0.0784 sec/batch
Epoch: 15/20... Training Step: 14486... Training loss: 1.1995... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14487... Training loss: 1.2698... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14488... Training loss: 1.2709... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14489... Training loss: 1.3400... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14490... Training loss: 1.2617... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14491... Training loss: 1.2647... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14492... Training loss: 1.3049... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14493... Training loss: 1.2469... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14494... Training loss: 1.2538... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14495... Training loss: 1.3442... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14496... Training loss: 1.3219... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14497... Training loss: 1.3026... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14498... Training loss: 1.3311... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14499... Training loss: 1.1980... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14500... Training loss: 1.3194... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14501... Training loss: 1.3500... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14502... Training loss: 1.2626... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14503... Training loss: 1.3304... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14504... Training loss: 1.2897... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14505... Training loss: 1.3285... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14506... Training loss: 1.3141... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14507... Training loss: 1.2493... 0.0741 sec/batch
Epoch: 15/20... Training Step: 14508... Training loss: 1.2887... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14509... Training loss: 1.2479... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14510... Training loss: 1.2260... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14511... Training loss: 1.3319... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14512... Training loss: 1.3101... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14513... Training loss: 1.2569... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14514... Training loss: 1.2817... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14515... Training loss: 1.2763... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14516... Training loss: 1.3092... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14517... Training loss: 1.1367... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14518... Training loss: 1.3249... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14519... Training loss: 1.2777... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14520... Training loss: 1.2620... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14521... Training loss: 1.2475... 0.0867 sec/batch
Epoch: 15/20... Training Step: 14522... Training loss: 1.2733... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14523... Training loss: 1.3033... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14524... Training loss: 1.2867... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14525... Training loss: 1.3536... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14526... Training loss: 1.2687... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14527... Training loss: 1.3081... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14528... Training loss: 1.3632... 0.0798 sec/batch
Epoch: 15/20... Training Step: 14529... Training loss: 1.3708... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14530... Training loss: 1.3547... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14531... Training loss: 1.2619... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14532... Training loss: 1.3082... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14533... Training loss: 1.2538... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14534... Training loss: 1.3012... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14535... Training loss: 1.3221... 0.0787 sec/batch
Epoch: 15/20... Training Step: 14536... Training loss: 1.2133... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14537... Training loss: 1.2500... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14538... Training loss: 1.2915... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14539... Training loss: 1.3127... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14540... Training loss: 1.3265... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14541... Training loss: 1.3004... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14542... Training loss: 1.3709... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14543... Training loss: 1.3651... 0.0780 sec/batch
Epoch: 15/20... Training Step: 14544... Training loss: 1.3244... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14545... Training loss: 1.2331... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14546... Training loss: 1.2720... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14547... Training loss: 1.2812... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14548... Training loss: 1.3783... 0.0784 sec/batch
Epoch: 15/20... Training Step: 14549... Training loss: 1.3238... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14550... Training loss: 1.3656... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14551... Training loss: 1.3012... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14552... Training loss: 1.3885... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14553... Training loss: 1.3916... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14554... Training loss: 1.2720... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14555... Training loss: 1.3853... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14556... Training loss: 1.3282... 0.0781 sec/batch
Epoch: 15/20... Training Step: 14557... Training loss: 1.3154... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14558... Training loss: 1.2705... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14559... Training loss: 1.3361... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14560... Training loss: 1.3718... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14561... Training loss: 1.3155... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14562... Training loss: 1.3060... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14563... Training loss: 1.2656... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14564... Training loss: 1.2705... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14565... Training loss: 1.3751... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14566... Training loss: 1.2751... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14567... Training loss: 1.2282... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14568... Training loss: 1.3295... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14569... Training loss: 1.3643... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14570... Training loss: 1.3386... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14571... Training loss: 1.3247... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14572... Training loss: 1.2774... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14573... Training loss: 1.3613... 0.0785 sec/batch
Epoch: 15/20... Training Step: 14574... Training loss: 1.3348... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14575... Training loss: 1.3274... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14576... Training loss: 1.3071... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14577... Training loss: 1.2573... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14578... Training loss: 1.3471... 0.0799 sec/batch
Epoch: 15/20... Training Step: 14579... Training loss: 1.2838... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14580... Training loss: 1.2989... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14581... Training loss: 1.2894... 0.0796 sec/batch
Epoch: 15/20... Training Step: 14582... Training loss: 1.2269... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14583... Training loss: 1.3079... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14584... Training loss: 1.3041... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14585... Training loss: 1.3970... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14586... Training loss: 1.3231... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14587... Training loss: 1.2733... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14588... Training loss: 1.3019... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14589... Training loss: 1.3179... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14590... Training loss: 1.2923... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14591... Training loss: 1.2980... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14592... Training loss: 1.2259... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14593... Training loss: 1.2517... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14594... Training loss: 1.2699... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14595... Training loss: 1.2995... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14596... Training loss: 1.3239... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14597... Training loss: 1.3020... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14598... Training loss: 1.2863... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14599... Training loss: 1.2739... 0.0804 sec/batch
Epoch: 15/20... Training Step: 14600... Training loss: 1.3546... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14601... Training loss: 1.2712... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14602... Training loss: 1.3025... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14603... Training loss: 1.3088... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14604... Training loss: 1.3245... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14605... Training loss: 1.3580... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14606... Training loss: 1.3348... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14607... Training loss: 1.3490... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14608... Training loss: 1.3893... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14609... Training loss: 1.3401... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14610... Training loss: 1.3124... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14611... Training loss: 1.2816... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14612... Training loss: 1.3437... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14613... Training loss: 1.3615... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14614... Training loss: 1.3780... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14615... Training loss: 1.2974... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14616... Training loss: 1.3444... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14617... Training loss: 1.3438... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14618... Training loss: 1.2703... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14619... Training loss: 1.2940... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14620... Training loss: 1.3411... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14621... Training loss: 1.2779... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14622... Training loss: 1.3416... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14623... Training loss: 1.3001... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14624... Training loss: 1.3105... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14625... Training loss: 1.2781... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14626... Training loss: 1.3803... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14627... Training loss: 1.2986... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14628... Training loss: 1.2947... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14629... Training loss: 1.2719... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14630... Training loss: 1.2974... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14631... Training loss: 1.2875... 0.0786 sec/batch
Epoch: 15/20... Training Step: 14632... Training loss: 1.3551... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14633... Training loss: 1.3359... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14634... Training loss: 1.2886... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14635... Training loss: 1.2948... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14636... Training loss: 1.3046... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14637... Training loss: 1.2424... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14638... Training loss: 1.2624... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14639... Training loss: 1.3904... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14640... Training loss: 1.2686... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14641... Training loss: 1.3284... 0.0741 sec/batch
Epoch: 15/20... Training Step: 14642... Training loss: 1.3436... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14643... Training loss: 1.2237... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14644... Training loss: 1.3026... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14645... Training loss: 1.3353... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14646... Training loss: 1.3508... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14647... Training loss: 1.3637... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14648... Training loss: 1.3830... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14649... Training loss: 1.3174... 0.0804 sec/batch
Epoch: 15/20... Training Step: 14650... Training loss: 1.2517... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14651... Training loss: 1.3278... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14652... Training loss: 1.3025... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14653... Training loss: 1.3223... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14654... Training loss: 1.3684... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14655... Training loss: 1.3014... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14656... Training loss: 1.2696... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14657... Training loss: 1.2950... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14658... Training loss: 1.3018... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14659... Training loss: 1.3095... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14660... Training loss: 1.3680... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14661... Training loss: 1.3464... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14662... Training loss: 1.3626... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14663... Training loss: 1.2904... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14664... Training loss: 1.2225... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14665... Training loss: 1.2703... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14666... Training loss: 1.2875... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14667... Training loss: 1.2868... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14668... Training loss: 1.2881... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14669... Training loss: 1.3138... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14670... Training loss: 1.3128... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14671... Training loss: 1.3289... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14672... Training loss: 1.3249... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14673... Training loss: 1.3174... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14674... Training loss: 1.3171... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14675... Training loss: 1.2860... 0.0793 sec/batch
Epoch: 15/20... Training Step: 14676... Training loss: 1.3267... 0.0842 sec/batch
Epoch: 15/20... Training Step: 14677... Training loss: 1.3145... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14678... Training loss: 1.3279... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14679... Training loss: 1.2946... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14680... Training loss: 1.2921... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14681... Training loss: 1.3105... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14682... Training loss: 1.2335... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14683... Training loss: 1.2359... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14684... Training loss: 1.3105... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14685... Training loss: 1.2763... 0.0799 sec/batch
Epoch: 15/20... Training Step: 14686... Training loss: 1.2905... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14687... Training loss: 1.2731... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14688... Training loss: 1.2759... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14689... Training loss: 1.2955... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14690... Training loss: 1.3453... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14691... Training loss: 1.3857... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14692... Training loss: 1.2983... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14693... Training loss: 1.2568... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14694... Training loss: 1.2925... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14695... Training loss: 1.2959... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14696... Training loss: 1.3232... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14697... Training loss: 1.3338... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14698... Training loss: 1.3602... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14699... Training loss: 1.3099... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14700... Training loss: 1.3645... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14701... Training loss: 1.3158... 0.0811 sec/batch
Epoch: 15/20... Training Step: 14702... Training loss: 1.3557... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14703... Training loss: 1.3450... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14704... Training loss: 1.3349... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14705... Training loss: 1.3824... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14706... Training loss: 1.4445... 0.0803 sec/batch
Epoch: 15/20... Training Step: 14707... Training loss: 1.3407... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14708... Training loss: 1.3067... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14709... Training loss: 1.2928... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14710... Training loss: 1.2696... 0.0737 sec/batch
Epoch: 15/20... Training Step: 14711... Training loss: 1.3270... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14712... Training loss: 1.3236... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14713... Training loss: 1.3092... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14714... Training loss: 1.3283... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14715... Training loss: 1.3334... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14716... Training loss: 1.3798... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14717... Training loss: 1.3440... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14718... Training loss: 1.2888... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14719... Training loss: 1.3499... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14720... Training loss: 1.3187... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14721... Training loss: 1.3580... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14722... Training loss: 1.3195... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14723... Training loss: 1.2661... 0.0743 sec/batch
Epoch: 15/20... Training Step: 14724... Training loss: 1.2918... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14725... Training loss: 1.3319... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14726... Training loss: 1.3381... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14727... Training loss: 1.3347... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14728... Training loss: 1.3033... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14729... Training loss: 1.3208... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14730... Training loss: 1.4405... 0.0780 sec/batch
Epoch: 15/20... Training Step: 14731... Training loss: 1.3541... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14732... Training loss: 1.3472... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14733... Training loss: 1.3540... 0.0788 sec/batch
Epoch: 15/20... Training Step: 14734... Training loss: 1.2636... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14735... Training loss: 1.3191... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14736... Training loss: 1.3260... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14737... Training loss: 1.3244... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14738... Training loss: 1.3532... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14739... Training loss: 1.3009... 0.0760 sec/batch
Epoch: 15/20... Training Step: 14740... Training loss: 1.4139... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14741... Training loss: 1.3846... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14742... Training loss: 1.3046... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14743... Training loss: 1.3197... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14744... Training loss: 1.3099... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14745... Training loss: 1.3172... 0.0775 sec/batch
Epoch: 15/20... Training Step: 14746... Training loss: 1.3062... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14747... Training loss: 1.3205... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14748... Training loss: 1.3531... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14749... Training loss: 1.2998... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14750... Training loss: 1.2725... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14751... Training loss: 1.2828... 0.0790 sec/batch
Epoch: 15/20... Training Step: 14752... Training loss: 1.2985... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14753... Training loss: 1.2685... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14754... Training loss: 1.3072... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14755... Training loss: 1.3087... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14756... Training loss: 1.3359... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14757... Training loss: 1.3208... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14758... Training loss: 1.3736... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14759... Training loss: 1.3575... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14760... Training loss: 1.3106... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14761... Training loss: 1.3146... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14762... Training loss: 1.3001... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14763... Training loss: 1.3097... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14764... Training loss: 1.3570... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14765... Training loss: 1.3220... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14766... Training loss: 1.3471... 0.0769 sec/batch
Epoch: 15/20... Training Step: 14767... Training loss: 1.3091... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14768... Training loss: 1.3103... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14769... Training loss: 1.3159... 0.0798 sec/batch
Epoch: 15/20... Training Step: 14770... Training loss: 1.2234... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14771... Training loss: 1.3416... 0.0740 sec/batch
Epoch: 15/20... Training Step: 14772... Training loss: 1.3101... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14773... Training loss: 1.2980... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14774... Training loss: 1.3831... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14775... Training loss: 1.3217... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14776... Training loss: 1.3683... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14777... Training loss: 1.3466... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14778... Training loss: 1.3896... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14779... Training loss: 1.3585... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14780... Training loss: 1.2412... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14781... Training loss: 1.3044... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14782... Training loss: 1.2633... 0.0782 sec/batch
Epoch: 15/20... Training Step: 14783... Training loss: 1.3588... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14784... Training loss: 1.2743... 0.0788 sec/batch
Epoch: 15/20... Training Step: 14785... Training loss: 1.3273... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14786... Training loss: 1.3462... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14787... Training loss: 1.2811... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14788... Training loss: 1.2597... 0.0794 sec/batch
Epoch: 15/20... Training Step: 14789... Training loss: 1.2530... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14790... Training loss: 1.2898... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14791... Training loss: 1.2588... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14792... Training loss: 1.2629... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14793... Training loss: 1.2262... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14794... Training loss: 1.3174... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14795... Training loss: 1.2769... 0.0748 sec/batch
Epoch: 15/20... Training Step: 14796... Training loss: 1.2896... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14797... Training loss: 1.2470... 0.0744 sec/batch
Epoch: 15/20... Training Step: 14798... Training loss: 1.2311... 0.0797 sec/batch
Epoch: 15/20... Training Step: 14799... Training loss: 1.2530... 0.0789 sec/batch
Epoch: 15/20... Training Step: 14800... Training loss: 1.3060... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14801... Training loss: 1.3944... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14802... Training loss: 1.2984... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14803... Training loss: 1.2913... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14804... Training loss: 1.2936... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14805... Training loss: 1.3074... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14806... Training loss: 1.3592... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14807... Training loss: 1.2762... 0.0757 sec/batch
Epoch: 15/20... Training Step: 14808... Training loss: 1.3387... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14809... Training loss: 1.4141... 0.0773 sec/batch
Epoch: 15/20... Training Step: 14810... Training loss: 1.3509... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14811... Training loss: 1.3024... 0.0756 sec/batch
Epoch: 15/20... Training Step: 14812... Training loss: 1.2698... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14813... Training loss: 1.3243... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14814... Training loss: 1.3444... 0.0794 sec/batch
Epoch: 15/20... Training Step: 14815... Training loss: 1.3328... 0.0749 sec/batch
Epoch: 15/20... Training Step: 14816... Training loss: 1.3083... 0.0779 sec/batch
Epoch: 15/20... Training Step: 14817... Training loss: 1.4067... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14818... Training loss: 1.3834... 0.0738 sec/batch
Epoch: 15/20... Training Step: 14819... Training loss: 1.3246... 0.0790 sec/batch
Epoch: 15/20... Training Step: 14820... Training loss: 1.3985... 0.0792 sec/batch
Epoch: 15/20... Training Step: 14821... Training loss: 1.2988... 0.0795 sec/batch
Epoch: 15/20... Training Step: 14822... Training loss: 1.3133... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14823... Training loss: 1.3288... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14824... Training loss: 1.3370... 0.0765 sec/batch
Epoch: 15/20... Training Step: 14825... Training loss: 1.2792... 0.0793 sec/batch
Epoch: 15/20... Training Step: 14826... Training loss: 1.2936... 0.0752 sec/batch
Epoch: 15/20... Training Step: 14827... Training loss: 1.3034... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14828... Training loss: 1.3896... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14829... Training loss: 1.3280... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14830... Training loss: 1.3087... 0.0737 sec/batch
Epoch: 15/20... Training Step: 14831... Training loss: 1.3580... 0.0788 sec/batch
Epoch: 15/20... Training Step: 14832... Training loss: 1.3542... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14833... Training loss: 1.3609... 0.0768 sec/batch
Epoch: 15/20... Training Step: 14834... Training loss: 1.3084... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14835... Training loss: 1.2925... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14836... Training loss: 1.2838... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14837... Training loss: 1.3234... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14838... Training loss: 1.2543... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14839... Training loss: 1.3188... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14840... Training loss: 1.2603... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14841... Training loss: 1.2974... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14842... Training loss: 1.3442... 0.0774 sec/batch
Epoch: 15/20... Training Step: 14843... Training loss: 1.3606... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14844... Training loss: 1.3196... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14845... Training loss: 1.2950... 0.0746 sec/batch
Epoch: 15/20... Training Step: 14846... Training loss: 1.3335... 0.0794 sec/batch
Epoch: 15/20... Training Step: 14847... Training loss: 1.2754... 0.0783 sec/batch
Epoch: 15/20... Training Step: 14848... Training loss: 1.2877... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14849... Training loss: 1.3446... 0.0796 sec/batch
Epoch: 15/20... Training Step: 14850... Training loss: 1.3022... 0.0761 sec/batch
Epoch: 15/20... Training Step: 14851... Training loss: 1.3217... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14852... Training loss: 1.3016... 0.0797 sec/batch
Epoch: 15/20... Training Step: 14853... Training loss: 1.3591... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14854... Training loss: 1.3380... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14855... Training loss: 1.3625... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14856... Training loss: 1.3641... 0.0755 sec/batch
Epoch: 15/20... Training Step: 14857... Training loss: 1.3529... 0.0762 sec/batch
Epoch: 15/20... Training Step: 14858... Training loss: 1.5206... 0.0758 sec/batch
Epoch: 15/20... Training Step: 14859... Training loss: 1.3014... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14860... Training loss: 1.4046... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14861... Training loss: 1.4221... 0.0799 sec/batch
Epoch: 15/20... Training Step: 14862... Training loss: 1.2913... 0.0770 sec/batch
Epoch: 15/20... Training Step: 14863... Training loss: 1.3506... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14864... Training loss: 1.3743... 0.0767 sec/batch
Epoch: 15/20... Training Step: 14865... Training loss: 1.3847... 0.0753 sec/batch
Epoch: 15/20... Training Step: 14866... Training loss: 1.3587... 0.0759 sec/batch
Epoch: 15/20... Training Step: 14867... Training loss: 1.3763... 0.0772 sec/batch
Epoch: 15/20... Training Step: 14868... Training loss: 1.2783... 0.0771 sec/batch
Epoch: 15/20... Training Step: 14869... Training loss: 1.2552... 0.0778 sec/batch
Epoch: 15/20... Training Step: 14870... Training loss: 1.3922... 0.0754 sec/batch
Epoch: 15/20... Training Step: 14871... Training loss: 1.3198... 0.0766 sec/batch
Epoch: 15/20... Training Step: 14872... Training loss: 1.3217... 0.0776 sec/batch
Epoch: 15/20... Training Step: 14873... Training loss: 1.3201... 0.0777 sec/batch
Epoch: 15/20... Training Step: 14874... Training loss: 1.2837... 0.0750 sec/batch
Epoch: 15/20... Training Step: 14875... Training loss: 1.2804... 0.0763 sec/batch
Epoch: 15/20... Training Step: 14876... Training loss: 1.3021... 0.0745 sec/batch
Epoch: 15/20... Training Step: 14877... Training loss: 1.2961... 0.0764 sec/batch
Epoch: 15/20... Training Step: 14878... Training loss: 1.2163... 0.0751 sec/batch
Epoch: 15/20... Training Step: 14879... Training loss: 1.3035... 0.0747 sec/batch
Epoch: 15/20... Training Step: 14880... Training loss: 1.3562... 0.0771 sec/batch
Epoch: 16/20... Training Step: 14881... Training loss: 1.3814... 0.0778 sec/batch
Epoch: 16/20... Training Step: 14882... Training loss: 1.2673... 0.0808 sec/batch
Epoch: 16/20... Training Step: 14883... Training loss: 1.3599... 0.0776 sec/batch
Epoch: 16/20... Training Step: 14884... Training loss: 1.3519... 0.0766 sec/batch
Epoch: 16/20... Training Step: 14885... Training loss: 1.2245... 0.0762 sec/batch
Epoch: 16/20... Training Step: 14886... Training loss: 1.1833... 0.0795 sec/batch
Epoch: 16/20... Training Step: 14887... Training loss: 1.2977... 0.0812 sec/batch
Epoch: 16/20... Training Step: 14888... Training loss: 1.3034... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14889... Training loss: 1.3132... 0.0758 sec/batch
Epoch: 16/20... Training Step: 14890... Training loss: 1.2304... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14891... Training loss: 1.3745... 0.0759 sec/batch
Epoch: 16/20... Training Step: 14892... Training loss: 1.2518... 0.0783 sec/batch
Epoch: 16/20... Training Step: 14893... Training loss: 1.2998... 0.0750 sec/batch
Epoch: 16/20... Training Step: 14894... Training loss: 1.3028... 0.0769 sec/batch
Epoch: 16/20... Training Step: 14895... Training loss: 1.3508... 0.0746 sec/batch
Epoch: 16/20... Training Step: 14896... Training loss: 1.2696... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14897... Training loss: 1.2994... 0.0803 sec/batch
Epoch: 16/20... Training Step: 14898... Training loss: 1.3127... 0.0787 sec/batch
Epoch: 16/20... Training Step: 14899... Training loss: 1.2924... 0.0780 sec/batch
Epoch: 16/20... Training Step: 14900... Training loss: 1.2757... 0.0789 sec/batch
Epoch: 16/20... Training Step: 14901... Training loss: 1.3082... 0.0759 sec/batch
Epoch: 16/20... Training Step: 14902... Training loss: 1.2897... 0.0760 sec/batch
Epoch: 16/20... Training Step: 14903... Training loss: 1.2894... 0.0766 sec/batch
Epoch: 16/20... Training Step: 14904... Training loss: 1.2317... 0.0777 sec/batch
Epoch: 16/20... Training Step: 14905... Training loss: 1.3203... 0.0808 sec/batch
Epoch: 16/20... Training Step: 14906... Training loss: 1.2618... 0.0780 sec/batch
Epoch: 16/20... Training Step: 14907... Training loss: 1.2873... 0.0775 sec/batch
Epoch: 16/20... Training Step: 14908... Training loss: 1.2455... 0.0795 sec/batch
Epoch: 16/20... Training Step: 14909... Training loss: 1.2780... 0.0797 sec/batch
Epoch: 16/20... Training Step: 14910... Training loss: 1.3646... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14911... Training loss: 1.3427... 0.0781 sec/batch
Epoch: 16/20... Training Step: 14912... Training loss: 1.3313... 0.0798 sec/batch
Epoch: 16/20... Training Step: 14913... Training loss: 1.3045... 0.0766 sec/batch
Epoch: 16/20... Training Step: 14914... Training loss: 1.3545... 0.0792 sec/batch
Epoch: 16/20... Training Step: 14915... Training loss: 1.3796... 0.0785 sec/batch
Epoch: 16/20... Training Step: 14916... Training loss: 1.3463... 0.0748 sec/batch
Epoch: 16/20... Training Step: 14917... Training loss: 1.2883... 0.0771 sec/batch
Epoch: 16/20... Training Step: 14918... Training loss: 1.3464... 0.0788 sec/batch
Epoch: 16/20... Training Step: 14919... Training loss: 1.3280... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14920... Training loss: 1.2999... 0.0772 sec/batch
Epoch: 16/20... Training Step: 14921... Training loss: 1.3184... 0.0767 sec/batch
Epoch: 16/20... Training Step: 14922... Training loss: 1.2849... 0.0764 sec/batch
Epoch: 16/20... Training Step: 14923... Training loss: 1.2464... 0.0748 sec/batch
Epoch: 16/20... Training Step: 14924... Training loss: 1.2731... 0.0775 sec/batch
Epoch: 16/20... Training Step: 14925... Training loss: 1.3208... 0.0786 sec/batch
Epoch: 16/20... Training Step: 14926... Training loss: 1.2904... 0.0760 sec/batch
Epoch: 16/20... Training Step: 14927... Training loss: 1.3460... 0.0751 sec/batch
Epoch: 16/20... Training Step: 14928... Training loss: 1.2801... 0.0749 sec/batch
Epoch: 16/20... Training Step: 14929... Training loss: 1.3398... 0.0765 sec/batch
Epoch: 16/20... Training Step: 14930... Training loss: 1.3407... 0.0767 sec/batch
Epoch: 16/20... Training Step: 14931... Training loss: 1.2266... 0.0749 sec/batch
Epoch: 16/20... Training Step: 14932... Training loss: 1.3220... 0.0753 sec/batch
Epoch: 16/20... Training Step: 14933... Training loss: 1.2386... 0.0786 sec/batch
Epoch: 16/20... Training Step: 14934... Training loss: 1.2451... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14935... Training loss: 1.2885... 0.0781 sec/batch
Epoch: 16/20... Training Step: 14936... Training loss: 1.2559... 0.0758 sec/batch
Epoch: 16/20... Training Step: 14937... Training loss: 1.3253... 0.0776 sec/batch
Epoch: 16/20... Training Step: 14938... Training loss: 1.3514... 0.0748 sec/batch
Epoch: 16/20... Training Step: 14939... Training loss: 1.2978... 0.0759 sec/batch
Epoch: 16/20... Training Step: 14940... Training loss: 1.3154... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14941... Training loss: 1.2299... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14942... Training loss: 1.2297... 0.0754 sec/batch
Epoch: 16/20... Training Step: 14943... Training loss: 1.2987... 0.0766 sec/batch
Epoch: 16/20... Training Step: 14944... Training loss: 1.2496... 0.0747 sec/batch
Epoch: 16/20... Training Step: 14945... Training loss: 1.3237... 0.0753 sec/batch
Epoch: 16/20... Training Step: 14946... Training loss: 1.3077... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14947... Training loss: 1.2571... 0.0767 sec/batch
Epoch: 16/20... Training Step: 14948... Training loss: 1.2863... 0.0775 sec/batch
Epoch: 16/20... Training Step: 14949... Training loss: 1.3286... 0.0763 sec/batch
Epoch: 16/20... Training Step: 14950... Training loss: 1.3139... 0.0740 sec/batch
Epoch: 16/20... Training Step: 14951... Training loss: 1.3131... 0.0806 sec/batch
Epoch: 16/20... Training Step: 14952... Training loss: 1.3734... 0.0747 sec/batch
Epoch: 16/20... Training Step: 14953... Training loss: 1.3205... 0.0764 sec/batch
Epoch: 16/20... Training Step: 14954... Training loss: 1.2688... 0.0774 sec/batch
Epoch: 16/20... Training Step: 14955... Training loss: 1.2767... 0.0754 sec/batch
Epoch: 16/20... Training Step: 14956... Training loss: 1.2598... 0.0764 sec/batch
Epoch: 16/20... Training Step: 14957... Training loss: 1.2674... 0.0743 sec/batch
Epoch: 16/20... Training Step: 14958... Training loss: 1.2747... 0.0761 sec/batch
Epoch: 16/20... Training Step: 14959... Training loss: 1.2990... 0.0744 sec/batch
Epoch: 16/20... Training Step: 14960... Training loss: 1.3559... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14961... Training loss: 1.3246... 0.0767 sec/batch
Epoch: 16/20... Training Step: 14962... Training loss: 1.2586... 0.0745 sec/batch
Epoch: 16/20... Training Step: 14963... Training loss: 1.3125... 0.0769 sec/batch
Epoch: 16/20... Training Step: 14964... Training loss: 1.3247... 0.0761 sec/batch
Epoch: 16/20... Training Step: 14965... Training loss: 1.2734... 0.0751 sec/batch
Epoch: 16/20... Training Step: 14966... Training loss: 1.3417... 0.0839 sec/batch
Epoch: 16/20... Training Step: 14967... Training loss: 1.3007... 0.0763 sec/batch
Epoch: 16/20... Training Step: 14968... Training loss: 1.2819... 0.0770 sec/batch
Epoch: 16/20... Training Step: 14969... Training loss: 1.3361... 0.0751 sec/batch
Epoch: 16/20... Training Step: 14970... Training loss: 1.3296... 0.0773 sec/batch
Epoch: 16/20... Training Step: 14971... Training loss: 1.3506... 0.0774 sec/batch
Epoch: 16/20... Training Step: 14972... Training loss: 1.2704... 0.0771 sec/batch
Epoch: 16/20... Training Step: 14973... Training loss: 1.2543... 0.0756 sec/batch
Epoch: 16/20... Training Step: 14974... Training loss: 1.2162... 0.0768 sec/batch
Epoch: 16/20... Training Step: 14975... Training loss: 1.1791... 0.0775 sec/batch
Epoch: 16/20... Training Step: 14976... Training loss: 1.3185... 0.0753 sec/batch
Epoch: 16/20... Training Step: 14977... Training loss: 1.2723... 0.0787 sec/batch
Epoch: 16/20... Training Step: 14978... Training loss: 1.2473... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14979... Training loss: 1.2747... 0.0759 sec/batch
Epoch: 16/20... Training Step: 14980... Training loss: 1.2601... 0.0759 sec/batch
Epoch: 16/20... Training Step: 14981... Training loss: 1.2550... 0.0771 sec/batch
Epoch: 16/20... Training Step: 14982... Training loss: 1.2644... 0.0768 sec/batch
Epoch: 16/20... Training Step: 14983... Training loss: 1.2828... 0.0749 sec/batch
Epoch: 16/20... Training Step: 14984... Training loss: 1.3362... 0.0777 sec/batch
Epoch: 16/20... Training Step: 14985... Training loss: 1.3185... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14986... Training loss: 1.2835... 0.0750 sec/batch
Epoch: 16/20... Training Step: 14987... Training loss: 1.2450... 0.0766 sec/batch
Epoch: 16/20... Training Step: 14988... Training loss: 1.3309... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14989... Training loss: 1.3027... 0.0775 sec/batch
Epoch: 16/20... Training Step: 14990... Training loss: 1.3276... 0.0795 sec/batch
Epoch: 16/20... Training Step: 14991... Training loss: 1.2984... 0.0760 sec/batch
Epoch: 16/20... Training Step: 14992... Training loss: 1.2929... 0.0774 sec/batch
Epoch: 16/20... Training Step: 14993... Training loss: 1.3714... 0.0767 sec/batch
Epoch: 16/20... Training Step: 14994... Training loss: 1.3194... 0.0772 sec/batch
Epoch: 16/20... Training Step: 14995... Training loss: 1.2796... 0.0777 sec/batch
Epoch: 16/20... Training Step: 14996... Training loss: 1.4233... 0.0752 sec/batch
Epoch: 16/20... Training Step: 14997... Training loss: 1.3388... 0.0776 sec/batch
Epoch: 16/20... Training Step: 14998... Training loss: 1.4084... 0.0757 sec/batch
Epoch: 16/20... Training Step: 14999... Training loss: 1.3132... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15000... Training loss: 1.2979... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15001... Training loss: 1.4126... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15002... Training loss: 1.2565... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15003... Training loss: 1.3382... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15004... Training loss: 1.3063... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15005... Training loss: 1.3539... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15006... Training loss: 1.3757... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15007... Training loss: 1.3441... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15008... Training loss: 1.3064... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15009... Training loss: 1.2419... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15010... Training loss: 1.3292... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15011... Training loss: 1.2796... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15012... Training loss: 1.3589... 0.0788 sec/batch
Epoch: 16/20... Training Step: 15013... Training loss: 1.3349... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15014... Training loss: 1.2963... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15015... Training loss: 1.3465... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15016... Training loss: 1.3301... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15017... Training loss: 1.3576... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15018... Training loss: 1.3780... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15019... Training loss: 1.3511... 0.0813 sec/batch
Epoch: 16/20... Training Step: 15020... Training loss: 1.3023... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15021... Training loss: 1.3092... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15022... Training loss: 1.2999... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15023... Training loss: 1.3013... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15024... Training loss: 1.2937... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15025... Training loss: 1.1843... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15026... Training loss: 1.3533... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15027... Training loss: 1.3076... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15028... Training loss: 1.2526... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15029... Training loss: 1.2627... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15030... Training loss: 1.2658... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15031... Training loss: 1.1919... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15032... Training loss: 1.2986... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15033... Training loss: 1.2473... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15034... Training loss: 1.2847... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15035... Training loss: 1.2921... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15036... Training loss: 1.2513... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15037... Training loss: 1.2730... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15038... Training loss: 1.3027... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15039... Training loss: 1.2892... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15040... Training loss: 1.3187... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15041... Training loss: 1.2788... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15042... Training loss: 1.3121... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15043... Training loss: 1.2982... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15044... Training loss: 1.3571... 0.0797 sec/batch
Epoch: 16/20... Training Step: 15045... Training loss: 1.3137... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15046... Training loss: 1.3905... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15047... Training loss: 1.3404... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15048... Training loss: 1.3002... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15049... Training loss: 1.3312... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15050... Training loss: 1.3060... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15051... Training loss: 1.3047... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15052... Training loss: 1.3053... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15053... Training loss: 1.3129... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15054... Training loss: 1.3539... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15055... Training loss: 1.2606... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15056... Training loss: 1.3196... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15057... Training loss: 1.3368... 0.0926 sec/batch
Epoch: 16/20... Training Step: 15058... Training loss: 1.4046... 0.0911 sec/batch
Epoch: 16/20... Training Step: 15059... Training loss: 1.3398... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15060... Training loss: 1.2900... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15061... Training loss: 1.3131... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15062... Training loss: 1.3821... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15063... Training loss: 1.2971... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15064... Training loss: 1.3276... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15065... Training loss: 1.3068... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15066... Training loss: 1.3515... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15067... Training loss: 1.2908... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15068... Training loss: 1.2950... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15069... Training loss: 1.3175... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15070... Training loss: 1.3646... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15071... Training loss: 1.3794... 0.0815 sec/batch
Epoch: 16/20... Training Step: 15072... Training loss: 1.3389... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15073... Training loss: 1.3548... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15074... Training loss: 1.4008... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15075... Training loss: 1.3101... 0.0799 sec/batch
Epoch: 16/20... Training Step: 15076... Training loss: 1.2645... 0.0793 sec/batch
Epoch: 16/20... Training Step: 15077... Training loss: 1.3707... 0.0839 sec/batch
Epoch: 16/20... Training Step: 15078... Training loss: 1.3065... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15079... Training loss: 1.3464... 0.0797 sec/batch
Epoch: 16/20... Training Step: 15080... Training loss: 1.2861... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15081... Training loss: 1.3083... 0.0836 sec/batch
Epoch: 16/20... Training Step: 15082... Training loss: 1.3123... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15083... Training loss: 1.3118... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15084... Training loss: 1.2960... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15085... Training loss: 1.2684... 0.0788 sec/batch
Epoch: 16/20... Training Step: 15086... Training loss: 1.3485... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15087... Training loss: 1.2984... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15088... Training loss: 1.3043... 0.0794 sec/batch
Epoch: 16/20... Training Step: 15089... Training loss: 1.3354... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15090... Training loss: 1.3130... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15091... Training loss: 1.2781... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15092... Training loss: 1.3464... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15093... Training loss: 1.2839... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15094... Training loss: 1.3557... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15095... Training loss: 1.2771... 0.0868 sec/batch
Epoch: 16/20... Training Step: 15096... Training loss: 1.2966... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15097... Training loss: 1.2786... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15098... Training loss: 1.3247... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15099... Training loss: 1.3006... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15100... Training loss: 1.2438... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15101... Training loss: 1.2566... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15102... Training loss: 1.4089... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15103... Training loss: 1.3128... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15104... Training loss: 1.3077... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15105... Training loss: 1.3551... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15106... Training loss: 1.2764... 0.0823 sec/batch
Epoch: 16/20... Training Step: 15107... Training loss: 1.3173... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15108... Training loss: 1.3350... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15109... Training loss: 1.2295... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15110... Training loss: 1.2462... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15111... Training loss: 1.3162... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15112... Training loss: 1.3103... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15113... Training loss: 1.2977... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15114... Training loss: 1.3900... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15115... Training loss: 1.2782... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15116... Training loss: 1.3175... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15117... Training loss: 1.2933... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15118... Training loss: 1.3358... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15119... Training loss: 1.3396... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15120... Training loss: 1.2879... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15121... Training loss: 1.3080... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15122... Training loss: 1.3138... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15123... Training loss: 1.3550... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15124... Training loss: 1.3409... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15125... Training loss: 1.2895... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15126... Training loss: 1.3371... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15127... Training loss: 1.2680... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15128... Training loss: 1.2348... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15129... Training loss: 1.2816... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15130... Training loss: 1.3220... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15131... Training loss: 1.2917... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15132... Training loss: 1.2872... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15133... Training loss: 1.2629... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15134... Training loss: 1.2569... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15135... Training loss: 1.2463... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15136... Training loss: 1.3127... 0.0791 sec/batch
Epoch: 16/20... Training Step: 15137... Training loss: 1.3623... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15138... Training loss: 1.2917... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15139... Training loss: 1.2970... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15140... Training loss: 1.2895... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15141... Training loss: 1.3414... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15142... Training loss: 1.3474... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15143... Training loss: 1.3240... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15144... Training loss: 1.3210... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15145... Training loss: 1.3212... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15146... Training loss: 1.2096... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15147... Training loss: 1.2747... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15148... Training loss: 1.3007... 0.0820 sec/batch
Epoch: 16/20... Training Step: 15149... Training loss: 1.2523... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15150... Training loss: 1.2794... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15151... Training loss: 1.2970... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15152... Training loss: 1.2626... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15153... Training loss: 1.2270... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15154... Training loss: 1.2315... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15155... Training loss: 1.2942... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15156... Training loss: 1.2875... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15157... Training loss: 1.3214... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15158... Training loss: 1.2006... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15159... Training loss: 1.2589... 0.0739 sec/batch
Epoch: 16/20... Training Step: 15160... Training loss: 1.3693... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15161... Training loss: 1.3019... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15162... Training loss: 1.2672... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15163... Training loss: 1.3364... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15164... Training loss: 1.2782... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15165... Training loss: 1.1851... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15166... Training loss: 1.2654... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15167... Training loss: 1.3376... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15168... Training loss: 1.2933... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15169... Training loss: 1.2606... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15170... Training loss: 1.2613... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15171... Training loss: 1.2881... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15172... Training loss: 1.2957... 0.0837 sec/batch
Epoch: 16/20... Training Step: 15173... Training loss: 1.2297... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15174... Training loss: 1.2577... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15175... Training loss: 1.3091... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15176... Training loss: 1.3137... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15177... Training loss: 1.2741... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15178... Training loss: 1.3741... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15179... Training loss: 1.2961... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15180... Training loss: 1.3195... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15181... Training loss: 1.3598... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15182... Training loss: 1.2437... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15183... Training loss: 1.2671... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15184... Training loss: 1.3333... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15185... Training loss: 1.3017... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15186... Training loss: 1.3566... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15187... Training loss: 1.2703... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15188... Training loss: 1.2278... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15189... Training loss: 1.3812... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15190... Training loss: 1.2716... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15191... Training loss: 1.3826... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15192... Training loss: 1.3089... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15193... Training loss: 1.2824... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15194... Training loss: 1.2956... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15195... Training loss: 1.2946... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15196... Training loss: 1.2493... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15197... Training loss: 1.2544... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15198... Training loss: 1.2474... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15199... Training loss: 1.2959... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15200... Training loss: 1.3182... 0.0820 sec/batch
Epoch: 16/20... Training Step: 15201... Training loss: 1.2706... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15202... Training loss: 1.2327... 0.0729 sec/batch
Epoch: 16/20... Training Step: 15203... Training loss: 1.2670... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15204... Training loss: 1.2672... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15205... Training loss: 1.3503... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15206... Training loss: 1.2847... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15207... Training loss: 1.3013... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15208... Training loss: 1.2925... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15209... Training loss: 1.2914... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15210... Training loss: 1.3163... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15211... Training loss: 1.3318... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15212... Training loss: 1.2509... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15213... Training loss: 1.3808... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15214... Training loss: 1.3803... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15215... Training loss: 1.3613... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15216... Training loss: 1.3338... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15217... Training loss: 1.2602... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15218... Training loss: 1.2923... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15219... Training loss: 1.3079... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15220... Training loss: 1.2802... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15221... Training loss: 1.2960... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15222... Training loss: 1.2540... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15223... Training loss: 1.2565... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15224... Training loss: 1.3394... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15225... Training loss: 1.2871... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15226... Training loss: 1.3090... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15227... Training loss: 1.3416... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15228... Training loss: 1.3414... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15229... Training loss: 1.3540... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15230... Training loss: 1.3577... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15231... Training loss: 1.3376... 0.0796 sec/batch
Epoch: 16/20... Training Step: 15232... Training loss: 1.3021... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15233... Training loss: 1.2841... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15234... Training loss: 1.2842... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15235... Training loss: 1.3179... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15236... Training loss: 1.3071... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15237... Training loss: 1.2684... 0.0794 sec/batch
Epoch: 16/20... Training Step: 15238... Training loss: 1.2998... 0.0858 sec/batch
Epoch: 16/20... Training Step: 15239... Training loss: 1.3075... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15240... Training loss: 1.3003... 0.0837 sec/batch
Epoch: 16/20... Training Step: 15241... Training loss: 1.3262... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15242... Training loss: 1.3704... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15243... Training loss: 1.2879... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15244... Training loss: 1.3174... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15245... Training loss: 1.3164... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15246... Training loss: 1.2602... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15247... Training loss: 1.2908... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15248... Training loss: 1.2233... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15249... Training loss: 1.2509... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15250... Training loss: 1.2408... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15251... Training loss: 1.3586... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15252... Training loss: 1.2472... 0.0795 sec/batch
Epoch: 16/20... Training Step: 15253... Training loss: 1.2756... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15254... Training loss: 1.2182... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15255... Training loss: 1.2680... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15256... Training loss: 1.2418... 0.0791 sec/batch
Epoch: 16/20... Training Step: 15257... Training loss: 1.2864... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15258... Training loss: 1.3087... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15259... Training loss: 1.2785... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15260... Training loss: 1.3172... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15261... Training loss: 1.3121... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15262... Training loss: 1.2718... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15263... Training loss: 1.3343... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15264... Training loss: 1.2982... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15265... Training loss: 1.3088... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15266... Training loss: 1.2802... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15267... Training loss: 1.3452... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15268... Training loss: 1.3183... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15269... Training loss: 1.3510... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15270... Training loss: 1.2559... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15271... Training loss: 1.3303... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15272... Training loss: 1.2973... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15273... Training loss: 1.3056... 0.0834 sec/batch
Epoch: 16/20... Training Step: 15274... Training loss: 1.2508... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15275... Training loss: 1.2789... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15276... Training loss: 1.2748... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15277... Training loss: 1.2872... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15278... Training loss: 1.3430... 0.0795 sec/batch
Epoch: 16/20... Training Step: 15279... Training loss: 1.3054... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15280... Training loss: 1.2764... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15281... Training loss: 1.2658... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15282... Training loss: 1.3157... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15283... Training loss: 1.2710... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15284... Training loss: 1.3634... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15285... Training loss: 1.2920... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15286... Training loss: 1.3132... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15287... Training loss: 1.2204... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15288... Training loss: 1.3585... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15289... Training loss: 1.2309... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15290... Training loss: 1.2651... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15291... Training loss: 1.2011... 0.0825 sec/batch
Epoch: 16/20... Training Step: 15292... Training loss: 1.2627... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15293... Training loss: 1.3021... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15294... Training loss: 1.2124... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15295... Training loss: 1.2506... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15296... Training loss: 1.3455... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15297... Training loss: 1.2349... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15298... Training loss: 1.2733... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15299... Training loss: 1.3533... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15300... Training loss: 1.3521... 0.0788 sec/batch
Epoch: 16/20... Training Step: 15301... Training loss: 1.2288... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15302... Training loss: 1.3299... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15303... Training loss: 1.3070... 0.0825 sec/batch
Epoch: 16/20... Training Step: 15304... Training loss: 1.3070... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15305... Training loss: 1.2637... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15306... Training loss: 1.3024... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15307... Training loss: 1.3571... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15308... Training loss: 1.3750... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15309... Training loss: 1.3530... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15310... Training loss: 1.3903... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15311... Training loss: 1.2823... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15312... Training loss: 1.3206... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15313... Training loss: 1.3456... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15314... Training loss: 1.3104... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15315... Training loss: 1.3424... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15316... Training loss: 1.2821... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15317... Training loss: 1.3500... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15318... Training loss: 1.3882... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15319... Training loss: 1.3371... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15320... Training loss: 1.3220... 0.0797 sec/batch
Epoch: 16/20... Training Step: 15321... Training loss: 1.2890... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15322... Training loss: 1.3376... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15323... Training loss: 1.3166... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15324... Training loss: 1.3461... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15325... Training loss: 1.2865... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15326... Training loss: 1.3872... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15327... Training loss: 1.2986... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15328... Training loss: 1.3328... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15329... Training loss: 1.3822... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15330... Training loss: 1.4003... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15331... Training loss: 1.3066... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15332... Training loss: 1.3054... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15333... Training loss: 1.3434... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15334... Training loss: 1.3379... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15335... Training loss: 1.3271... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15336... Training loss: 1.2945... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15337... Training loss: 1.2872... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15338... Training loss: 1.2885... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15339... Training loss: 1.2924... 0.0815 sec/batch
Epoch: 16/20... Training Step: 15340... Training loss: 1.3295... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15341... Training loss: 1.3220... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15342... Training loss: 1.3063... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15343... Training loss: 1.2482... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15344... Training loss: 1.3275... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15345... Training loss: 1.2613... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15346... Training loss: 1.2716... 0.0832 sec/batch
Epoch: 16/20... Training Step: 15347... Training loss: 1.2235... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15348... Training loss: 1.3183... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15349... Training loss: 1.2788... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15350... Training loss: 1.3562... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15351... Training loss: 1.3186... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15352... Training loss: 1.2850... 0.0878 sec/batch
Epoch: 16/20... Training Step: 15353... Training loss: 1.3211... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15354... Training loss: 1.2761... 0.0793 sec/batch
Epoch: 16/20... Training Step: 15355... Training loss: 1.1863... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15356... Training loss: 1.2650... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15357... Training loss: 1.3234... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15358... Training loss: 1.2945... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15359... Training loss: 1.2367... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15360... Training loss: 1.2605... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15361... Training loss: 1.2572... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15362... Training loss: 1.3033... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15363... Training loss: 1.2759... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15364... Training loss: 1.2914... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15365... Training loss: 1.2220... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15366... Training loss: 1.2909... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15367... Training loss: 1.3322... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15368... Training loss: 1.2752... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15369... Training loss: 1.2507... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15370... Training loss: 1.2502... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15371... Training loss: 1.2745... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15372... Training loss: 1.2588... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15373... Training loss: 1.2915... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15374... Training loss: 1.3001... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15375... Training loss: 1.2392... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15376... Training loss: 1.2461... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15377... Training loss: 1.2861... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15378... Training loss: 1.2877... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15379... Training loss: 1.2510... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15380... Training loss: 1.3083... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15381... Training loss: 1.2331... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15382... Training loss: 1.2771... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15383... Training loss: 1.2935... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15384... Training loss: 1.2666... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15385... Training loss: 1.2875... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15386... Training loss: 1.2323... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15387... Training loss: 1.2701... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15388... Training loss: 1.2719... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15389... Training loss: 1.2711... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15390... Training loss: 1.2715... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15391... Training loss: 1.3138... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15392... Training loss: 1.3135... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15393... Training loss: 1.3308... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15394... Training loss: 1.3334... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15395... Training loss: 1.2769... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15396... Training loss: 1.3136... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15397... Training loss: 1.3042... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15398... Training loss: 1.2977... 0.0786 sec/batch
Epoch: 16/20... Training Step: 15399... Training loss: 1.3012... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15400... Training loss: 1.2482... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15401... Training loss: 1.2956... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15402... Training loss: 1.2542... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15403... Training loss: 1.3657... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15404... Training loss: 1.3000... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15405... Training loss: 1.2856... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15406... Training loss: 1.2525... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15407... Training loss: 1.2546... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15408... Training loss: 1.2768... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15409... Training loss: 1.3032... 0.0735 sec/batch
Epoch: 16/20... Training Step: 15410... Training loss: 1.3362... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15411... Training loss: 1.3739... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15412... Training loss: 1.3226... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15413... Training loss: 1.3423... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15414... Training loss: 1.3449... 0.0795 sec/batch
Epoch: 16/20... Training Step: 15415... Training loss: 1.3054... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15416... Training loss: 1.2850... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15417... Training loss: 1.2517... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15418... Training loss: 1.2701... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15419... Training loss: 1.2670... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15420... Training loss: 1.2799... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15421... Training loss: 1.3686... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15422... Training loss: 1.2791... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15423... Training loss: 1.2741... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15424... Training loss: 1.3003... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15425... Training loss: 1.2231... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15426... Training loss: 1.2639... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15427... Training loss: 1.3523... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15428... Training loss: 1.3233... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15429... Training loss: 1.2527... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15430... Training loss: 1.2722... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15431... Training loss: 1.2851... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15432... Training loss: 1.2283... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15433... Training loss: 1.2399... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15434... Training loss: 1.2693... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15435... Training loss: 1.3012... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15436... Training loss: 1.2494... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15437... Training loss: 1.2319... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15438... Training loss: 1.2521... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15439... Training loss: 1.3171... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15440... Training loss: 1.2068... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15441... Training loss: 1.2602... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15442... Training loss: 1.3116... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15443... Training loss: 1.2518... 0.0834 sec/batch
Epoch: 16/20... Training Step: 15444... Training loss: 1.2912... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15445... Training loss: 1.2968... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15446... Training loss: 1.2910... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15447... Training loss: 1.2642... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15448... Training loss: 1.2285... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15449... Training loss: 1.1991... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15450... Training loss: 1.2416... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15451... Training loss: 1.2303... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15452... Training loss: 1.2646... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15453... Training loss: 1.3354... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15454... Training loss: 1.3088... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15455... Training loss: 1.3058... 0.0786 sec/batch
Epoch: 16/20... Training Step: 15456... Training loss: 1.2935... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15457... Training loss: 1.2685... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15458... Training loss: 1.4104... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15459... Training loss: 1.3409... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15460... Training loss: 1.2483... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15461... Training loss: 1.2299... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15462... Training loss: 1.2735... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15463... Training loss: 1.3400... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15464... Training loss: 1.2336... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15465... Training loss: 1.2644... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15466... Training loss: 1.3227... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15467... Training loss: 1.3008... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15468... Training loss: 1.3364... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15469... Training loss: 1.2914... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15470... Training loss: 1.2664... 0.0853 sec/batch
Epoch: 16/20... Training Step: 15471... Training loss: 1.2898... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15472... Training loss: 1.3197... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15473... Training loss: 1.2348... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15474... Training loss: 1.2612... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15475... Training loss: 1.2133... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15476... Training loss: 1.2996... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15477... Training loss: 1.1771... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15478... Training loss: 1.2265... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15479... Training loss: 1.2417... 0.0786 sec/batch
Epoch: 16/20... Training Step: 15480... Training loss: 1.2686... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15481... Training loss: 1.3251... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15482... Training loss: 1.2694... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15483... Training loss: 1.2674... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15484... Training loss: 1.2828... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15485... Training loss: 1.2288... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15486... Training loss: 1.2730... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15487... Training loss: 1.3237... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15488... Training loss: 1.3201... 0.0851 sec/batch
Epoch: 16/20... Training Step: 15489... Training loss: 1.2757... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15490... Training loss: 1.3470... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15491... Training loss: 1.1985... 0.0832 sec/batch
Epoch: 16/20... Training Step: 15492... Training loss: 1.3091... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15493... Training loss: 1.3231... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15494... Training loss: 1.2541... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15495... Training loss: 1.3342... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15496... Training loss: 1.3005... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15497... Training loss: 1.3007... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15498... Training loss: 1.3327... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15499... Training loss: 1.2545... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15500... Training loss: 1.2750... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15501... Training loss: 1.2564... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15502... Training loss: 1.2265... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15503... Training loss: 1.3209... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15504... Training loss: 1.3195... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15505... Training loss: 1.2423... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15506... Training loss: 1.2631... 0.0889 sec/batch
Epoch: 16/20... Training Step: 15507... Training loss: 1.2490... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15508... Training loss: 1.2963... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15509... Training loss: 1.1295... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15510... Training loss: 1.2853... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15511... Training loss: 1.2695... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15512... Training loss: 1.2737... 0.0822 sec/batch
Epoch: 16/20... Training Step: 15513... Training loss: 1.2387... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15514... Training loss: 1.2803... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15515... Training loss: 1.3009... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15516... Training loss: 1.2753... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15517... Training loss: 1.3390... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15518... Training loss: 1.2470... 0.0802 sec/batch
Epoch: 16/20... Training Step: 15519... Training loss: 1.3026... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15520... Training loss: 1.3567... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15521... Training loss: 1.3821... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15522... Training loss: 1.3407... 0.0738 sec/batch
Epoch: 16/20... Training Step: 15523... Training loss: 1.2716... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15524... Training loss: 1.2925... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15525... Training loss: 1.2654... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15526... Training loss: 1.3150... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15527... Training loss: 1.2880... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15528... Training loss: 1.2156... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15529... Training loss: 1.2463... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15530... Training loss: 1.2711... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15531... Training loss: 1.2711... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15532... Training loss: 1.3332... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15533... Training loss: 1.2769... 0.0838 sec/batch
Epoch: 16/20... Training Step: 15534... Training loss: 1.3386... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15535... Training loss: 1.3461... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15536... Training loss: 1.3276... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15537... Training loss: 1.2339... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15538... Training loss: 1.2601... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15539... Training loss: 1.2540... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15540... Training loss: 1.3536... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15541... Training loss: 1.3146... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15542... Training loss: 1.3553... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15543... Training loss: 1.3106... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15544... Training loss: 1.3727... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15545... Training loss: 1.3745... 0.0799 sec/batch
Epoch: 16/20... Training Step: 15546... Training loss: 1.2766... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15547... Training loss: 1.4160... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15548... Training loss: 1.3384... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15549... Training loss: 1.3298... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15550... Training loss: 1.2707... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15551... Training loss: 1.3319... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15552... Training loss: 1.3669... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15553... Training loss: 1.2975... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15554... Training loss: 1.2942... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15555... Training loss: 1.2443... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15556... Training loss: 1.2552... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15557... Training loss: 1.3768... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15558... Training loss: 1.2889... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15559... Training loss: 1.2310... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15560... Training loss: 1.3282... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15561... Training loss: 1.3321... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15562... Training loss: 1.3196... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15563... Training loss: 1.3161... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15564... Training loss: 1.2780... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15565... Training loss: 1.3615... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15566... Training loss: 1.3275... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15567... Training loss: 1.2961... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15568... Training loss: 1.3127... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15569... Training loss: 1.2643... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15570... Training loss: 1.3208... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15571... Training loss: 1.2875... 0.0811 sec/batch
Epoch: 16/20... Training Step: 15572... Training loss: 1.3066... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15573... Training loss: 1.2938... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15574... Training loss: 1.2344... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15575... Training loss: 1.3106... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15576... Training loss: 1.2875... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15577... Training loss: 1.3612... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15578... Training loss: 1.2992... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15579... Training loss: 1.2652... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15580... Training loss: 1.2999... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15581... Training loss: 1.3295... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15582... Training loss: 1.2836... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15583... Training loss: 1.2883... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15584... Training loss: 1.2377... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15585... Training loss: 1.2347... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15586... Training loss: 1.2524... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15587... Training loss: 1.2940... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15588... Training loss: 1.3037... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15589... Training loss: 1.3197... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15590... Training loss: 1.2729... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15591... Training loss: 1.2768... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15592... Training loss: 1.3393... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15593... Training loss: 1.2620... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15594... Training loss: 1.2834... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15595... Training loss: 1.3144... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15596... Training loss: 1.3229... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15597... Training loss: 1.3563... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15598... Training loss: 1.3286... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15599... Training loss: 1.3394... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15600... Training loss: 1.3895... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15601... Training loss: 1.3391... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15602... Training loss: 1.2939... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15603... Training loss: 1.3032... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15604... Training loss: 1.3493... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15605... Training loss: 1.3502... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15606... Training loss: 1.3906... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15607... Training loss: 1.2820... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15608... Training loss: 1.3463... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15609... Training loss: 1.3287... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15610... Training loss: 1.2610... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15611... Training loss: 1.2806... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15612... Training loss: 1.3456... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15613... Training loss: 1.2751... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15614... Training loss: 1.3370... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15615... Training loss: 1.2973... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15616... Training loss: 1.2910... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15617... Training loss: 1.2791... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15618... Training loss: 1.3644... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15619... Training loss: 1.2622... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15620... Training loss: 1.3006... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15621... Training loss: 1.2573... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15622... Training loss: 1.2970... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15623... Training loss: 1.3104... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15624... Training loss: 1.3662... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15625... Training loss: 1.3113... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15626... Training loss: 1.2889... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15627... Training loss: 1.2815... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15628... Training loss: 1.3005... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15629... Training loss: 1.2351... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15630... Training loss: 1.2437... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15631... Training loss: 1.3734... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15632... Training loss: 1.2735... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15633... Training loss: 1.3349... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15634... Training loss: 1.3097... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15635... Training loss: 1.2248... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15636... Training loss: 1.3012... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15637... Training loss: 1.2974... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15638... Training loss: 1.3494... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15639... Training loss: 1.3570... 0.0809 sec/batch
Epoch: 16/20... Training Step: 15640... Training loss: 1.3759... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15641... Training loss: 1.3006... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15642... Training loss: 1.2299... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15643... Training loss: 1.3013... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15644... Training loss: 1.2958... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15645... Training loss: 1.3456... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15646... Training loss: 1.3440... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15647... Training loss: 1.3015... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15648... Training loss: 1.2586... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15649... Training loss: 1.2682... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15650... Training loss: 1.3180... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15651... Training loss: 1.3189... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15652... Training loss: 1.3581... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15653... Training loss: 1.3338... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15654... Training loss: 1.3292... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15655... Training loss: 1.2977... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15656... Training loss: 1.2343... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15657... Training loss: 1.2535... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15658... Training loss: 1.3034... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15659... Training loss: 1.2515... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15660... Training loss: 1.2889... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15661... Training loss: 1.3089... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15662... Training loss: 1.3040... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15663... Training loss: 1.3171... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15664... Training loss: 1.3171... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15665... Training loss: 1.3045... 0.0805 sec/batch
Epoch: 16/20... Training Step: 15666... Training loss: 1.3275... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15667... Training loss: 1.2924... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15668... Training loss: 1.3288... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15669... Training loss: 1.3173... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15670... Training loss: 1.3196... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15671... Training loss: 1.2863... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15672... Training loss: 1.2790... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15673... Training loss: 1.3087... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15674... Training loss: 1.2602... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15675... Training loss: 1.2166... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15676... Training loss: 1.2904... 0.0819 sec/batch
Epoch: 16/20... Training Step: 15677... Training loss: 1.2714... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15678... Training loss: 1.2816... 0.0774 sec/batch
Epoch: 16/20... Training Step: 15679... Training loss: 1.2687... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15680... Training loss: 1.2770... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15681... Training loss: 1.3034... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15682... Training loss: 1.3396... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15683... Training loss: 1.3693... 0.0739 sec/batch
Epoch: 16/20... Training Step: 15684... Training loss: 1.3130... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15685... Training loss: 1.2525... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15686... Training loss: 1.3093... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15687... Training loss: 1.2862... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15688... Training loss: 1.3255... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15689... Training loss: 1.3515... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15690... Training loss: 1.3837... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15691... Training loss: 1.3260... 0.0815 sec/batch
Epoch: 16/20... Training Step: 15692... Training loss: 1.3701... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15693... Training loss: 1.3110... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15694... Training loss: 1.3333... 0.0789 sec/batch
Epoch: 16/20... Training Step: 15695... Training loss: 1.3461... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15696... Training loss: 1.3188... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15697... Training loss: 1.3612... 0.0790 sec/batch
Epoch: 16/20... Training Step: 15698... Training loss: 1.3952... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15699... Training loss: 1.3153... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15700... Training loss: 1.2930... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15701... Training loss: 1.2617... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15702... Training loss: 1.2542... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15703... Training loss: 1.3042... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15704... Training loss: 1.2995... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15705... Training loss: 1.3065... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15706... Training loss: 1.3321... 0.0852 sec/batch
Epoch: 16/20... Training Step: 15707... Training loss: 1.3021... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15708... Training loss: 1.3645... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15709... Training loss: 1.3471... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15710... Training loss: 1.2579... 0.0780 sec/batch
Epoch: 16/20... Training Step: 15711... Training loss: 1.3433... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15712... Training loss: 1.3107... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15713... Training loss: 1.3454... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15714... Training loss: 1.3350... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15715... Training loss: 1.2531... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15716... Training loss: 1.2831... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15717... Training loss: 1.2997... 0.0788 sec/batch
Epoch: 16/20... Training Step: 15718... Training loss: 1.3271... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15719... Training loss: 1.3375... 0.0742 sec/batch
Epoch: 16/20... Training Step: 15720... Training loss: 1.2978... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15721... Training loss: 1.3115... 0.0838 sec/batch
Epoch: 16/20... Training Step: 15722... Training loss: 1.4186... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15723... Training loss: 1.3349... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15724... Training loss: 1.3490... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15725... Training loss: 1.3282... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15726... Training loss: 1.2592... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15727... Training loss: 1.3178... 0.0768 sec/batch
Epoch: 16/20... Training Step: 15728... Training loss: 1.2937... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15729... Training loss: 1.3160... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15730... Training loss: 1.3405... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15731... Training loss: 1.2953... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15732... Training loss: 1.3969... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15733... Training loss: 1.3903... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15734... Training loss: 1.3130... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15735... Training loss: 1.2959... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15736... Training loss: 1.3150... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15737... Training loss: 1.3179... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15738... Training loss: 1.2910... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15739... Training loss: 1.3170... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15740... Training loss: 1.3459... 0.0800 sec/batch
Epoch: 16/20... Training Step: 15741... Training loss: 1.3179... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15742... Training loss: 1.2648... 0.0771 sec/batch
Epoch: 16/20... Training Step: 15743... Training loss: 1.2658... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15744... Training loss: 1.2908... 0.0806 sec/batch
Epoch: 16/20... Training Step: 15745... Training loss: 1.2464... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15746... Training loss: 1.3088... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15747... Training loss: 1.3100... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15748... Training loss: 1.3491... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15749... Training loss: 1.3005... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15750... Training loss: 1.3758... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15751... Training loss: 1.3466... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15752... Training loss: 1.3049... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15753... Training loss: 1.3260... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15754... Training loss: 1.3210... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15755... Training loss: 1.3206... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15756... Training loss: 1.3516... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15757... Training loss: 1.3164... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15758... Training loss: 1.3353... 0.0735 sec/batch
Epoch: 16/20... Training Step: 15759... Training loss: 1.2870... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15760... Training loss: 1.2956... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15761... Training loss: 1.3226... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15762... Training loss: 1.2472... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15763... Training loss: 1.3244... 0.0766 sec/batch
Epoch: 16/20... Training Step: 15764... Training loss: 1.2956... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15765... Training loss: 1.3074... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15766... Training loss: 1.3502... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15767... Training loss: 1.3139... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15768... Training loss: 1.3568... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15769... Training loss: 1.3432... 0.0812 sec/batch
Epoch: 16/20... Training Step: 15770... Training loss: 1.3718... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15771... Training loss: 1.3694... 0.0783 sec/batch
Epoch: 16/20... Training Step: 15772... Training loss: 1.2474... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15773... Training loss: 1.2886... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15774... Training loss: 1.2232... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15775... Training loss: 1.3467... 0.0793 sec/batch
Epoch: 16/20... Training Step: 15776... Training loss: 1.2466... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15777... Training loss: 1.3178... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15778... Training loss: 1.3501... 0.0784 sec/batch
Epoch: 16/20... Training Step: 15779... Training loss: 1.2708... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15780... Training loss: 1.2448... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15781... Training loss: 1.2873... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15782... Training loss: 1.2695... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15783... Training loss: 1.2759... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15784... Training loss: 1.2692... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15785... Training loss: 1.2295... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15786... Training loss: 1.3294... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15787... Training loss: 1.2424... 0.0793 sec/batch
Epoch: 16/20... Training Step: 15788... Training loss: 1.2880... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15789... Training loss: 1.2296... 0.0743 sec/batch
Epoch: 16/20... Training Step: 15790... Training loss: 1.1946... 0.0772 sec/batch
Epoch: 16/20... Training Step: 15791... Training loss: 1.2455... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15792... Training loss: 1.3056... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15793... Training loss: 1.4025... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15794... Training loss: 1.3023... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15795... Training loss: 1.3083... 0.0810 sec/batch
Epoch: 16/20... Training Step: 15796... Training loss: 1.2654... 0.0779 sec/batch
Epoch: 16/20... Training Step: 15797... Training loss: 1.2755... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15798... Training loss: 1.3585... 0.0785 sec/batch
Epoch: 16/20... Training Step: 15799... Training loss: 1.2877... 0.0782 sec/batch
Epoch: 16/20... Training Step: 15800... Training loss: 1.3489... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15801... Training loss: 1.3919... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15802... Training loss: 1.3278... 0.0744 sec/batch
Epoch: 16/20... Training Step: 15803... Training loss: 1.3175... 0.0777 sec/batch
Epoch: 16/20... Training Step: 15804... Training loss: 1.2629... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15805... Training loss: 1.2998... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15806... Training loss: 1.3417... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15807... Training loss: 1.3044... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15808... Training loss: 1.2988... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15809... Training loss: 1.3941... 0.0738 sec/batch
Epoch: 16/20... Training Step: 15810... Training loss: 1.3668... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15811... Training loss: 1.3132... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15812... Training loss: 1.3992... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15813... Training loss: 1.2907... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15814... Training loss: 1.3317... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15815... Training loss: 1.3198... 0.0741 sec/batch
Epoch: 16/20... Training Step: 15816... Training loss: 1.3389... 0.0811 sec/batch
Epoch: 16/20... Training Step: 15817... Training loss: 1.2759... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15818... Training loss: 1.3143... 0.0753 sec/batch
Epoch: 16/20... Training Step: 15819... Training loss: 1.3106... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15820... Training loss: 1.3912... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15821... Training loss: 1.3009... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15822... Training loss: 1.3375... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15823... Training loss: 1.3715... 0.0761 sec/batch
Epoch: 16/20... Training Step: 15824... Training loss: 1.3554... 0.0778 sec/batch
Epoch: 16/20... Training Step: 15825... Training loss: 1.3433... 0.0806 sec/batch
Epoch: 16/20... Training Step: 15826... Training loss: 1.3007... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15827... Training loss: 1.2864... 0.0767 sec/batch
Epoch: 16/20... Training Step: 15828... Training loss: 1.2760... 0.0756 sec/batch
Epoch: 16/20... Training Step: 15829... Training loss: 1.3002... 0.0769 sec/batch
Epoch: 16/20... Training Step: 15830... Training loss: 1.2381... 0.0740 sec/batch
Epoch: 16/20... Training Step: 15831... Training loss: 1.3031... 0.0757 sec/batch
Epoch: 16/20... Training Step: 15832... Training loss: 1.2740... 0.0749 sec/batch
Epoch: 16/20... Training Step: 15833... Training loss: 1.3136... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15834... Training loss: 1.3268... 0.0758 sec/batch
Epoch: 16/20... Training Step: 15835... Training loss: 1.3094... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15836... Training loss: 1.2994... 0.0750 sec/batch
Epoch: 16/20... Training Step: 15837... Training loss: 1.3099... 0.0781 sec/batch
Epoch: 16/20... Training Step: 15838... Training loss: 1.3039... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15839... Training loss: 1.2491... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15840... Training loss: 1.2815... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15841... Training loss: 1.3280... 0.0755 sec/batch
Epoch: 16/20... Training Step: 15842... Training loss: 1.2890... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15843... Training loss: 1.3031... 0.0787 sec/batch
Epoch: 16/20... Training Step: 15844... Training loss: 1.2988... 0.0754 sec/batch
Epoch: 16/20... Training Step: 15845... Training loss: 1.3668... 0.0765 sec/batch
Epoch: 16/20... Training Step: 15846... Training loss: 1.3373... 0.0762 sec/batch
Epoch: 16/20... Training Step: 15847... Training loss: 1.3490... 0.0770 sec/batch
Epoch: 16/20... Training Step: 15848... Training loss: 1.3445... 0.0737 sec/batch
Epoch: 16/20... Training Step: 15849... Training loss: 1.3611... 0.0813 sec/batch
Epoch: 16/20... Training Step: 15850... Training loss: 1.4883... 0.0843 sec/batch
Epoch: 16/20... Training Step: 15851... Training loss: 1.2746... 0.0798 sec/batch
Epoch: 16/20... Training Step: 15852... Training loss: 1.4064... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15853... Training loss: 1.4074... 0.0764 sec/batch
Epoch: 16/20... Training Step: 15854... Training loss: 1.2687... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15855... Training loss: 1.3378... 0.0773 sec/batch
Epoch: 16/20... Training Step: 15856... Training loss: 1.3614... 0.0752 sec/batch
Epoch: 16/20... Training Step: 15857... Training loss: 1.3485... 0.0747 sec/batch
Epoch: 16/20... Training Step: 15858... Training loss: 1.3403... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15859... Training loss: 1.3520... 0.0763 sec/batch
Epoch: 16/20... Training Step: 15860... Training loss: 1.2731... 0.0745 sec/batch
Epoch: 16/20... Training Step: 15861... Training loss: 1.2376... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15862... Training loss: 1.3964... 0.0751 sec/batch
Epoch: 16/20... Training Step: 15863... Training loss: 1.3198... 0.0759 sec/batch
Epoch: 16/20... Training Step: 15864... Training loss: 1.3291... 0.0776 sec/batch
Epoch: 16/20... Training Step: 15865... Training loss: 1.3243... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15866... Training loss: 1.2671... 0.0760 sec/batch
Epoch: 16/20... Training Step: 15867... Training loss: 1.2658... 0.0786 sec/batch
Epoch: 16/20... Training Step: 15868... Training loss: 1.3125... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15869... Training loss: 1.3063... 0.0746 sec/batch
Epoch: 16/20... Training Step: 15870... Training loss: 1.1912... 0.0775 sec/batch
Epoch: 16/20... Training Step: 15871... Training loss: 1.2811... 0.0748 sec/batch
Epoch: 16/20... Training Step: 15872... Training loss: 1.3559... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15873... Training loss: 1.3781... 0.0752 sec/batch
Epoch: 17/20... Training Step: 15874... Training loss: 1.2720... 0.0754 sec/batch
Epoch: 17/20... Training Step: 15875... Training loss: 1.3650... 0.0747 sec/batch
Epoch: 17/20... Training Step: 15876... Training loss: 1.3279... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15877... Training loss: 1.2284... 0.0750 sec/batch
Epoch: 17/20... Training Step: 15878... Training loss: 1.1781... 0.0742 sec/batch
Epoch: 17/20... Training Step: 15879... Training loss: 1.2684... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15880... Training loss: 1.3330... 0.0753 sec/batch
Epoch: 17/20... Training Step: 15881... Training loss: 1.3117... 0.0785 sec/batch
Epoch: 17/20... Training Step: 15882... Training loss: 1.1874... 0.0800 sec/batch
Epoch: 17/20... Training Step: 15883... Training loss: 1.3747... 0.0763 sec/batch
Epoch: 17/20... Training Step: 15884... Training loss: 1.2631... 0.0747 sec/batch
Epoch: 17/20... Training Step: 15885... Training loss: 1.2868... 0.0823 sec/batch
Epoch: 17/20... Training Step: 15886... Training loss: 1.3018... 0.0745 sec/batch
Epoch: 17/20... Training Step: 15887... Training loss: 1.3647... 0.0753 sec/batch
Epoch: 17/20... Training Step: 15888... Training loss: 1.2649... 0.0772 sec/batch
Epoch: 17/20... Training Step: 15889... Training loss: 1.2943... 0.0792 sec/batch
Epoch: 17/20... Training Step: 15890... Training loss: 1.3331... 0.0763 sec/batch
Epoch: 17/20... Training Step: 15891... Training loss: 1.2903... 0.0768 sec/batch
Epoch: 17/20... Training Step: 15892... Training loss: 1.2591... 0.0754 sec/batch
Epoch: 17/20... Training Step: 15893... Training loss: 1.2955... 0.0765 sec/batch
Epoch: 17/20... Training Step: 15894... Training loss: 1.2693... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15895... Training loss: 1.2874... 0.0781 sec/batch
Epoch: 17/20... Training Step: 15896... Training loss: 1.2445... 0.0755 sec/batch
Epoch: 17/20... Training Step: 15897... Training loss: 1.3058... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15898... Training loss: 1.2478... 0.0754 sec/batch
Epoch: 17/20... Training Step: 15899... Training loss: 1.2760... 0.0767 sec/batch
Epoch: 17/20... Training Step: 15900... Training loss: 1.2581... 0.0768 sec/batch
Epoch: 17/20... Training Step: 15901... Training loss: 1.3166... 0.0749 sec/batch
Epoch: 17/20... Training Step: 15902... Training loss: 1.3396... 0.0752 sec/batch
Epoch: 17/20... Training Step: 15903... Training loss: 1.3411... 0.0760 sec/batch
Epoch: 17/20... Training Step: 15904... Training loss: 1.3402... 0.0762 sec/batch
Epoch: 17/20... Training Step: 15905... Training loss: 1.3140... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15906... Training loss: 1.3426... 0.0783 sec/batch
Epoch: 17/20... Training Step: 15907... Training loss: 1.4003... 0.0782 sec/batch
Epoch: 17/20... Training Step: 15908... Training loss: 1.3302... 0.0760 sec/batch
Epoch: 17/20... Training Step: 15909... Training loss: 1.2873... 0.0774 sec/batch
Epoch: 17/20... Training Step: 15910... Training loss: 1.3597... 0.0781 sec/batch
Epoch: 17/20... Training Step: 15911... Training loss: 1.3288... 0.0777 sec/batch
Epoch: 17/20... Training Step: 15912... Training loss: 1.2532... 0.0795 sec/batch
Epoch: 17/20... Training Step: 15913... Training loss: 1.3034... 0.0765 sec/batch
Epoch: 17/20... Training Step: 15914... Training loss: 1.2706... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15915... Training loss: 1.2502... 0.0754 sec/batch
Epoch: 17/20... Training Step: 15916... Training loss: 1.2740... 0.0762 sec/batch
Epoch: 17/20... Training Step: 15917... Training loss: 1.3289... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15918... Training loss: 1.2691... 0.0822 sec/batch
Epoch: 17/20... Training Step: 15919... Training loss: 1.3421... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15920... Training loss: 1.2693... 0.0747 sec/batch
Epoch: 17/20... Training Step: 15921... Training loss: 1.3240... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15922... Training loss: 1.3182... 0.0741 sec/batch
Epoch: 17/20... Training Step: 15923... Training loss: 1.2425... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15924... Training loss: 1.3199... 0.0786 sec/batch
Epoch: 17/20... Training Step: 15925... Training loss: 1.2316... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15926... Training loss: 1.2647... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15927... Training loss: 1.2690... 0.0751 sec/batch
Epoch: 17/20... Training Step: 15928... Training loss: 1.2530... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15929... Training loss: 1.2988... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15930... Training loss: 1.3335... 0.0778 sec/batch
Epoch: 17/20... Training Step: 15931... Training loss: 1.3095... 0.0766 sec/batch
Epoch: 17/20... Training Step: 15932... Training loss: 1.2977... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15933... Training loss: 1.2059... 0.0791 sec/batch
Epoch: 17/20... Training Step: 15934... Training loss: 1.2413... 0.0742 sec/batch
Epoch: 17/20... Training Step: 15935... Training loss: 1.3004... 0.0771 sec/batch
Epoch: 17/20... Training Step: 15936... Training loss: 1.2555... 0.0774 sec/batch
Epoch: 17/20... Training Step: 15937... Training loss: 1.3098... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15938... Training loss: 1.2945... 0.0751 sec/batch
Epoch: 17/20... Training Step: 15939... Training loss: 1.2358... 0.0770 sec/batch
Epoch: 17/20... Training Step: 15940... Training loss: 1.2743... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15941... Training loss: 1.3301... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15942... Training loss: 1.3003... 0.0799 sec/batch
Epoch: 17/20... Training Step: 15943... Training loss: 1.3009... 0.0760 sec/batch
Epoch: 17/20... Training Step: 15944... Training loss: 1.3584... 0.0757 sec/batch
Epoch: 17/20... Training Step: 15945... Training loss: 1.3250... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15946... Training loss: 1.2417... 0.0753 sec/batch
Epoch: 17/20... Training Step: 15947... Training loss: 1.2796... 0.0774 sec/batch
Epoch: 17/20... Training Step: 15948... Training loss: 1.2667... 0.0762 sec/batch
Epoch: 17/20... Training Step: 15949... Training loss: 1.2703... 0.0751 sec/batch
Epoch: 17/20... Training Step: 15950... Training loss: 1.2774... 0.0750 sec/batch
Epoch: 17/20... Training Step: 15951... Training loss: 1.2905... 0.0772 sec/batch
Epoch: 17/20... Training Step: 15952... Training loss: 1.3298... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15953... Training loss: 1.2981... 0.0772 sec/batch
Epoch: 17/20... Training Step: 15954... Training loss: 1.2968... 0.0756 sec/batch
Epoch: 17/20... Training Step: 15955... Training loss: 1.3057... 0.0760 sec/batch
Epoch: 17/20... Training Step: 15956... Training loss: 1.3261... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15957... Training loss: 1.2562... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15958... Training loss: 1.3333... 0.0766 sec/batch
Epoch: 17/20... Training Step: 15959... Training loss: 1.3044... 0.0786 sec/batch
Epoch: 17/20... Training Step: 15960... Training loss: 1.2882... 0.0780 sec/batch
Epoch: 17/20... Training Step: 15961... Training loss: 1.3133... 0.0749 sec/batch
Epoch: 17/20... Training Step: 15962... Training loss: 1.3495... 0.0741 sec/batch
Epoch: 17/20... Training Step: 15963... Training loss: 1.3696... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15964... Training loss: 1.2729... 0.0753 sec/batch
Epoch: 17/20... Training Step: 15965... Training loss: 1.2645... 0.0752 sec/batch
Epoch: 17/20... Training Step: 15966... Training loss: 1.2217... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15967... Training loss: 1.2217... 0.0743 sec/batch
Epoch: 17/20... Training Step: 15968... Training loss: 1.3040... 0.0797 sec/batch
Epoch: 17/20... Training Step: 15969... Training loss: 1.2593... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15970... Training loss: 1.2385... 0.0754 sec/batch
Epoch: 17/20... Training Step: 15971... Training loss: 1.2576... 0.0783 sec/batch
Epoch: 17/20... Training Step: 15972... Training loss: 1.2630... 0.0770 sec/batch
Epoch: 17/20... Training Step: 15973... Training loss: 1.2628... 0.0780 sec/batch
Epoch: 17/20... Training Step: 15974... Training loss: 1.2689... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15975... Training loss: 1.2968... 0.0761 sec/batch
Epoch: 17/20... Training Step: 15976... Training loss: 1.3277... 0.0756 sec/batch
Epoch: 17/20... Training Step: 15977... Training loss: 1.3166... 0.0758 sec/batch
Epoch: 17/20... Training Step: 15978... Training loss: 1.2977... 0.0780 sec/batch
Epoch: 17/20... Training Step: 15979... Training loss: 1.2067... 0.0742 sec/batch
Epoch: 17/20... Training Step: 15980... Training loss: 1.3216... 0.0759 sec/batch
Epoch: 17/20... Training Step: 15981... Training loss: 1.2724... 0.0759 sec/batch
Epoch: 17/20... Training Step: 15982... Training loss: 1.3265... 0.0749 sec/batch
Epoch: 17/20... Training Step: 15983... Training loss: 1.2963... 0.0749 sec/batch
Epoch: 17/20... Training Step: 15984... Training loss: 1.2937... 0.0770 sec/batch
Epoch: 17/20... Training Step: 15985... Training loss: 1.3722... 0.0765 sec/batch
Epoch: 17/20... Training Step: 15986... Training loss: 1.3087... 0.0757 sec/batch
Epoch: 17/20... Training Step: 15987... Training loss: 1.2229... 0.0773 sec/batch
Epoch: 17/20... Training Step: 15988... Training loss: 1.4034... 0.0762 sec/batch
Epoch: 17/20... Training Step: 15989... Training loss: 1.3492... 0.0797 sec/batch
Epoch: 17/20... Training Step: 15990... Training loss: 1.3989... 0.0748 sec/batch
Epoch: 17/20... Training Step: 15991... Training loss: 1.3070... 0.0753 sec/batch
Epoch: 17/20... Training Step: 15992... Training loss: 1.2837... 0.0760 sec/batch
Epoch: 17/20... Training Step: 15993... Training loss: 1.4018... 0.0775 sec/batch
Epoch: 17/20... Training Step: 15994... Training loss: 1.2661... 0.0792 sec/batch
Epoch: 17/20... Training Step: 15995... Training loss: 1.3106... 0.0746 sec/batch
Epoch: 17/20... Training Step: 15996... Training loss: 1.2910... 0.0759 sec/batch
Epoch: 17/20... Training Step: 15997... Training loss: 1.3583... 0.0778 sec/batch
Epoch: 17/20... Training Step: 15998... Training loss: 1.3848... 0.0762 sec/batch
Epoch: 17/20... Training Step: 15999... Training loss: 1.3556... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16000... Training loss: 1.3174... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16001... Training loss: 1.2442... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16002... Training loss: 1.3218... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16003... Training loss: 1.2735... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16004... Training loss: 1.3462... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16005... Training loss: 1.3165... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16006... Training loss: 1.2949... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16007... Training loss: 1.3273... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16008... Training loss: 1.3598... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16009... Training loss: 1.3584... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16010... Training loss: 1.3861... 0.0841 sec/batch
Epoch: 17/20... Training Step: 16011... Training loss: 1.3365... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16012... Training loss: 1.3043... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16013... Training loss: 1.3080... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16014... Training loss: 1.2864... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16015... Training loss: 1.2954... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16016... Training loss: 1.3102... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16017... Training loss: 1.1866... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16018... Training loss: 1.3497... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16019... Training loss: 1.2845... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16020... Training loss: 1.2556... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16021... Training loss: 1.2599... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16022... Training loss: 1.2497... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16023... Training loss: 1.1848... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16024... Training loss: 1.2933... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16025... Training loss: 1.2323... 0.0833 sec/batch
Epoch: 17/20... Training Step: 16026... Training loss: 1.2704... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16027... Training loss: 1.2835... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16028... Training loss: 1.2581... 0.0793 sec/batch
Epoch: 17/20... Training Step: 16029... Training loss: 1.2822... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16030... Training loss: 1.3040... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16031... Training loss: 1.2920... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16032... Training loss: 1.3220... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16033... Training loss: 1.2645... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16034... Training loss: 1.2828... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16035... Training loss: 1.2965... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16036... Training loss: 1.3433... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16037... Training loss: 1.3081... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16038... Training loss: 1.3869... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16039... Training loss: 1.2996... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16040... Training loss: 1.3001... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16041... Training loss: 1.3223... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16042... Training loss: 1.2986... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16043... Training loss: 1.3109... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16044... Training loss: 1.2807... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16045... Training loss: 1.3048... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16046... Training loss: 1.3353... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16047... Training loss: 1.2788... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16048... Training loss: 1.3358... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16049... Training loss: 1.3000... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16050... Training loss: 1.3992... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16051... Training loss: 1.3328... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16052... Training loss: 1.2739... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16053... Training loss: 1.2871... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16054... Training loss: 1.3763... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16055... Training loss: 1.2886... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16056... Training loss: 1.3293... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16057... Training loss: 1.3182... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16058... Training loss: 1.3475... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16059... Training loss: 1.3063... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16060... Training loss: 1.2878... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16061... Training loss: 1.3180... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16062... Training loss: 1.3416... 0.0790 sec/batch
Epoch: 17/20... Training Step: 16063... Training loss: 1.3570... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16064... Training loss: 1.3232... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16065... Training loss: 1.3366... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16066... Training loss: 1.4018... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16067... Training loss: 1.3000... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16068... Training loss: 1.2725... 0.0787 sec/batch
Epoch: 17/20... Training Step: 16069... Training loss: 1.3433... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16070... Training loss: 1.3114... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16071... Training loss: 1.3190... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16072... Training loss: 1.2722... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16073... Training loss: 1.2928... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16074... Training loss: 1.3045... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16075... Training loss: 1.2904... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16076... Training loss: 1.2740... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16077... Training loss: 1.2667... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16078... Training loss: 1.3727... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16079... Training loss: 1.2873... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16080... Training loss: 1.2845... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16081... Training loss: 1.3405... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16082... Training loss: 1.3050... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16083... Training loss: 1.2609... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16084... Training loss: 1.3621... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16085... Training loss: 1.2633... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16086... Training loss: 1.3472... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16087... Training loss: 1.2731... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16088... Training loss: 1.3210... 0.0826 sec/batch
Epoch: 17/20... Training Step: 16089... Training loss: 1.2687... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16090... Training loss: 1.3081... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16091... Training loss: 1.2829... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16092... Training loss: 1.2349... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16093... Training loss: 1.2355... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16094... Training loss: 1.3922... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16095... Training loss: 1.3032... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16096... Training loss: 1.3100... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16097... Training loss: 1.3472... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16098... Training loss: 1.2701... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16099... Training loss: 1.3069... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16100... Training loss: 1.3486... 0.0866 sec/batch
Epoch: 17/20... Training Step: 16101... Training loss: 1.2220... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16102... Training loss: 1.2396... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16103... Training loss: 1.2884... 0.0801 sec/batch
Epoch: 17/20... Training Step: 16104... Training loss: 1.3015... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16105... Training loss: 1.2911... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16106... Training loss: 1.3591... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16107... Training loss: 1.2842... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16108... Training loss: 1.3299... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16109... Training loss: 1.2744... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16110... Training loss: 1.3172... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16111... Training loss: 1.3408... 0.0738 sec/batch
Epoch: 17/20... Training Step: 16112... Training loss: 1.2778... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16113... Training loss: 1.2917... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16114... Training loss: 1.2990... 0.0792 sec/batch
Epoch: 17/20... Training Step: 16115... Training loss: 1.3465... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16116... Training loss: 1.3233... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16117... Training loss: 1.2918... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16118... Training loss: 1.3309... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16119... Training loss: 1.2557... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16120... Training loss: 1.2119... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16121... Training loss: 1.3021... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16122... Training loss: 1.3058... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16123... Training loss: 1.2830... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16124... Training loss: 1.2786... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16125... Training loss: 1.2595... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16126... Training loss: 1.2574... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16127... Training loss: 1.2407... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16128... Training loss: 1.2960... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16129... Training loss: 1.3873... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16130... Training loss: 1.2710... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16131... Training loss: 1.2771... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16132... Training loss: 1.3119... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16133... Training loss: 1.3221... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16134... Training loss: 1.3336... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16135... Training loss: 1.3519... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16136... Training loss: 1.3187... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16137... Training loss: 1.3043... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16138... Training loss: 1.2079... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16139... Training loss: 1.2680... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16140... Training loss: 1.2903... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16141... Training loss: 1.2299... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16142... Training loss: 1.2571... 0.0815 sec/batch
Epoch: 17/20... Training Step: 16143... Training loss: 1.2960... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16144... Training loss: 1.2724... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16145... Training loss: 1.2142... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16146... Training loss: 1.2239... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16147... Training loss: 1.2817... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16148... Training loss: 1.3076... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16149... Training loss: 1.3117... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16150... Training loss: 1.1952... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16151... Training loss: 1.2459... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16152... Training loss: 1.3511... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16153... Training loss: 1.2810... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16154... Training loss: 1.2644... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16155... Training loss: 1.3241... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16156... Training loss: 1.2533... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16157... Training loss: 1.1734... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16158... Training loss: 1.2520... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16159... Training loss: 1.2986... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16160... Training loss: 1.2782... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16161... Training loss: 1.2542... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16162... Training loss: 1.2524... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16163... Training loss: 1.2705... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16164... Training loss: 1.2826... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16165... Training loss: 1.2452... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16166... Training loss: 1.2385... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16167... Training loss: 1.2990... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16168... Training loss: 1.2860... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16169... Training loss: 1.2551... 0.0791 sec/batch
Epoch: 17/20... Training Step: 16170... Training loss: 1.3612... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16171... Training loss: 1.3029... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16172... Training loss: 1.3301... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16173... Training loss: 1.3714... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16174... Training loss: 1.2212... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16175... Training loss: 1.2687... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16176... Training loss: 1.3199... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16177... Training loss: 1.2929... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16178... Training loss: 1.3499... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16179... Training loss: 1.2501... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16180... Training loss: 1.2403... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16181... Training loss: 1.3622... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16182... Training loss: 1.2763... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16183... Training loss: 1.3684... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16184... Training loss: 1.3113... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16185... Training loss: 1.2827... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16186... Training loss: 1.2975... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16187... Training loss: 1.2803... 0.0799 sec/batch
Epoch: 17/20... Training Step: 16188... Training loss: 1.2550... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16189... Training loss: 1.2606... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16190... Training loss: 1.2287... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16191... Training loss: 1.2802... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16192... Training loss: 1.2998... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16193... Training loss: 1.3077... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16194... Training loss: 1.2263... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16195... Training loss: 1.2435... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16196... Training loss: 1.2636... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16197... Training loss: 1.3101... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16198... Training loss: 1.2910... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16199... Training loss: 1.2744... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16200... Training loss: 1.2810... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16201... Training loss: 1.2590... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16202... Training loss: 1.3140... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16203... Training loss: 1.3402... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16204... Training loss: 1.2431... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16205... Training loss: 1.3774... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16206... Training loss: 1.3777... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16207... Training loss: 1.3662... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16208... Training loss: 1.3210... 0.0793 sec/batch
Epoch: 17/20... Training Step: 16209... Training loss: 1.2810... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16210... Training loss: 1.2843... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16211... Training loss: 1.3092... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16212... Training loss: 1.2580... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16213... Training loss: 1.2737... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16214... Training loss: 1.2243... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16215... Training loss: 1.2468... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16216... Training loss: 1.3175... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16217... Training loss: 1.2854... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16218... Training loss: 1.2863... 0.0789 sec/batch
Epoch: 17/20... Training Step: 16219... Training loss: 1.3258... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16220... Training loss: 1.3096... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16221... Training loss: 1.3410... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16222... Training loss: 1.3656... 0.0829 sec/batch
Epoch: 17/20... Training Step: 16223... Training loss: 1.3377... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16224... Training loss: 1.2669... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16225... Training loss: 1.2739... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16226... Training loss: 1.3027... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16227... Training loss: 1.3119... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16228... Training loss: 1.3099... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16229... Training loss: 1.2802... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16230... Training loss: 1.2886... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16231... Training loss: 1.3013... 0.0790 sec/batch
Epoch: 17/20... Training Step: 16232... Training loss: 1.2978... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16233... Training loss: 1.3308... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16234... Training loss: 1.3525... 0.0804 sec/batch
Epoch: 17/20... Training Step: 16235... Training loss: 1.2899... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16236... Training loss: 1.3073... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16237... Training loss: 1.2837... 0.0822 sec/batch
Epoch: 17/20... Training Step: 16238... Training loss: 1.2609... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16239... Training loss: 1.2988... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16240... Training loss: 1.2262... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16241... Training loss: 1.2492... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16242... Training loss: 1.2482... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16243... Training loss: 1.3595... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16244... Training loss: 1.2335... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16245... Training loss: 1.2794... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16246... Training loss: 1.2172... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16247... Training loss: 1.2792... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16248... Training loss: 1.2561... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16249... Training loss: 1.2831... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16250... Training loss: 1.3271... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16251... Training loss: 1.2582... 0.0792 sec/batch
Epoch: 17/20... Training Step: 16252... Training loss: 1.3118... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16253... Training loss: 1.3147... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16254... Training loss: 1.2762... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16255... Training loss: 1.3108... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16256... Training loss: 1.3083... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16257... Training loss: 1.3333... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16258... Training loss: 1.3089... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16259... Training loss: 1.3176... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16260... Training loss: 1.3086... 0.0793 sec/batch
Epoch: 17/20... Training Step: 16261... Training loss: 1.3135... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16262... Training loss: 1.2580... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16263... Training loss: 1.3060... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16264... Training loss: 1.2908... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16265... Training loss: 1.3181... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16266... Training loss: 1.2692... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16267... Training loss: 1.2996... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16268... Training loss: 1.2838... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16269... Training loss: 1.2866... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16270... Training loss: 1.3441... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16271... Training loss: 1.2919... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16272... Training loss: 1.2826... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16273... Training loss: 1.2381... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16274... Training loss: 1.3194... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16275... Training loss: 1.2635... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16276... Training loss: 1.3255... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16277... Training loss: 1.2925... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16278... Training loss: 1.3066... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16279... Training loss: 1.2306... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16280... Training loss: 1.3594... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16281... Training loss: 1.2330... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16282... Training loss: 1.2552... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16283... Training loss: 1.1858... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16284... Training loss: 1.2376... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16285... Training loss: 1.2764... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16286... Training loss: 1.2219... 0.0787 sec/batch
Epoch: 17/20... Training Step: 16287... Training loss: 1.2269... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16288... Training loss: 1.3288... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16289... Training loss: 1.2405... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16290... Training loss: 1.2434... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16291... Training loss: 1.3206... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16292... Training loss: 1.3450... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16293... Training loss: 1.2261... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16294... Training loss: 1.3068... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16295... Training loss: 1.2760... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16296... Training loss: 1.2952... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16297... Training loss: 1.2799... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16298... Training loss: 1.2968... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16299... Training loss: 1.3673... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16300... Training loss: 1.3789... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16301... Training loss: 1.3484... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16302... Training loss: 1.3706... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16303... Training loss: 1.2783... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16304... Training loss: 1.2905... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16305... Training loss: 1.3316... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16306... Training loss: 1.3113... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16307... Training loss: 1.3251... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16308... Training loss: 1.2526... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16309... Training loss: 1.3410... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16310... Training loss: 1.3987... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16311... Training loss: 1.3202... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16312... Training loss: 1.3286... 0.0799 sec/batch
Epoch: 17/20... Training Step: 16313... Training loss: 1.2915... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16314... Training loss: 1.3417... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16315... Training loss: 1.2868... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16316... Training loss: 1.3291... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16317... Training loss: 1.3067... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16318... Training loss: 1.3860... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16319... Training loss: 1.2902... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16320... Training loss: 1.3290... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16321... Training loss: 1.3769... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16322... Training loss: 1.3952... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16323... Training loss: 1.2917... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16324... Training loss: 1.3027... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16325... Training loss: 1.3553... 0.0792 sec/batch
Epoch: 17/20... Training Step: 16326... Training loss: 1.3255... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16327... Training loss: 1.3336... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16328... Training loss: 1.2804... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16329... Training loss: 1.2622... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16330... Training loss: 1.2950... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16331... Training loss: 1.3046... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16332... Training loss: 1.3281... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16333... Training loss: 1.3109... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16334... Training loss: 1.2939... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16335... Training loss: 1.2819... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16336... Training loss: 1.3216... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16337... Training loss: 1.2701... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16338... Training loss: 1.2772... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16339... Training loss: 1.2043... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16340... Training loss: 1.3099... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16341... Training loss: 1.2861... 0.0734 sec/batch
Epoch: 17/20... Training Step: 16342... Training loss: 1.3469... 0.0790 sec/batch
Epoch: 17/20... Training Step: 16343... Training loss: 1.3030... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16344... Training loss: 1.2663... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16345... Training loss: 1.2827... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16346... Training loss: 1.2673... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16347... Training loss: 1.1870... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16348... Training loss: 1.2672... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16349... Training loss: 1.3018... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16350... Training loss: 1.3028... 0.0794 sec/batch
Epoch: 17/20... Training Step: 16351... Training loss: 1.2330... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16352... Training loss: 1.2358... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16353... Training loss: 1.2547... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16354... Training loss: 1.2975... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16355... Training loss: 1.2848... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16356... Training loss: 1.3017... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16357... Training loss: 1.2236... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16358... Training loss: 1.2592... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16359... Training loss: 1.3121... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16360... Training loss: 1.2638... 0.0797 sec/batch
Epoch: 17/20... Training Step: 16361... Training loss: 1.2652... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16362... Training loss: 1.2395... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16363... Training loss: 1.2576... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16364... Training loss: 1.2490... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16365... Training loss: 1.2822... 0.0814 sec/batch
Epoch: 17/20... Training Step: 16366... Training loss: 1.2644... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16367... Training loss: 1.2446... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16368... Training loss: 1.2489... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16369... Training loss: 1.2986... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16370... Training loss: 1.2739... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16371... Training loss: 1.2526... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16372... Training loss: 1.2848... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16373... Training loss: 1.2424... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16374... Training loss: 1.2843... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16375... Training loss: 1.2825... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16376... Training loss: 1.2699... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16377... Training loss: 1.2835... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16378... Training loss: 1.2155... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16379... Training loss: 1.2733... 0.0741 sec/batch
Epoch: 17/20... Training Step: 16380... Training loss: 1.2421... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16381... Training loss: 1.2793... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16382... Training loss: 1.2601... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16383... Training loss: 1.3217... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16384... Training loss: 1.2936... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16385... Training loss: 1.3220... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16386... Training loss: 1.3332... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16387... Training loss: 1.2851... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16388... Training loss: 1.2758... 0.0741 sec/batch
Epoch: 17/20... Training Step: 16389... Training loss: 1.3079... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16390... Training loss: 1.2884... 0.0855 sec/batch
Epoch: 17/20... Training Step: 16391... Training loss: 1.2915... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16392... Training loss: 1.2473... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16393... Training loss: 1.2837... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16394... Training loss: 1.2439... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16395... Training loss: 1.3892... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16396... Training loss: 1.2759... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16397... Training loss: 1.2734... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16398... Training loss: 1.2634... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16399... Training loss: 1.2537... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16400... Training loss: 1.2663... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16401... Training loss: 1.2939... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16402... Training loss: 1.3412... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16403... Training loss: 1.4189... 0.0741 sec/batch
Epoch: 17/20... Training Step: 16404... Training loss: 1.3366... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16405... Training loss: 1.3525... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16406... Training loss: 1.3392... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16407... Training loss: 1.3052... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16408... Training loss: 1.2311... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16409... Training loss: 1.2423... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16410... Training loss: 1.2606... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16411... Training loss: 1.2639... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16412... Training loss: 1.3107... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16413... Training loss: 1.3655... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16414... Training loss: 1.2873... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16415... Training loss: 1.2433... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16416... Training loss: 1.2851... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16417... Training loss: 1.2203... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16418... Training loss: 1.2661... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16419... Training loss: 1.3447... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16420... Training loss: 1.3042... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16421... Training loss: 1.2471... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16422... Training loss: 1.2573... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16423... Training loss: 1.2765... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16424... Training loss: 1.2187... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16425... Training loss: 1.2285... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16426... Training loss: 1.2847... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16427... Training loss: 1.3076... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16428... Training loss: 1.2532... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16429... Training loss: 1.2446... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16430... Training loss: 1.2616... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16431... Training loss: 1.2990... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16432... Training loss: 1.2047... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16433... Training loss: 1.2424... 0.0795 sec/batch
Epoch: 17/20... Training Step: 16434... Training loss: 1.2988... 0.0831 sec/batch
Epoch: 17/20... Training Step: 16435... Training loss: 1.2596... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16436... Training loss: 1.2839... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16437... Training loss: 1.3001... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16438... Training loss: 1.2907... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16439... Training loss: 1.2570... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16440... Training loss: 1.2151... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16441... Training loss: 1.1900... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16442... Training loss: 1.2427... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16443... Training loss: 1.2159... 0.0789 sec/batch
Epoch: 17/20... Training Step: 16444... Training loss: 1.2446... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16445... Training loss: 1.3412... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16446... Training loss: 1.3227... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16447... Training loss: 1.3333... 0.0741 sec/batch
Epoch: 17/20... Training Step: 16448... Training loss: 1.2911... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16449... Training loss: 1.2951... 0.0791 sec/batch
Epoch: 17/20... Training Step: 16450... Training loss: 1.4205... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16451... Training loss: 1.3294... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16452... Training loss: 1.2503... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16453... Training loss: 1.2275... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16454... Training loss: 1.2462... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16455... Training loss: 1.3164... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16456... Training loss: 1.2031... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16457... Training loss: 1.2745... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16458... Training loss: 1.3169... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16459... Training loss: 1.3148... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16460... Training loss: 1.3171... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16461... Training loss: 1.2720... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16462... Training loss: 1.2712... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16463... Training loss: 1.2757... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16464... Training loss: 1.3072... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16465... Training loss: 1.2326... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16466... Training loss: 1.2628... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16467... Training loss: 1.2004... 0.0823 sec/batch
Epoch: 17/20... Training Step: 16468... Training loss: 1.2909... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16469... Training loss: 1.1692... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16470... Training loss: 1.2014... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16471... Training loss: 1.2467... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16472... Training loss: 1.2352... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16473... Training loss: 1.3161... 0.0785 sec/batch
Epoch: 17/20... Training Step: 16474... Training loss: 1.2549... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16475... Training loss: 1.2508... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16476... Training loss: 1.2870... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16477... Training loss: 1.2330... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16478... Training loss: 1.2473... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16479... Training loss: 1.3174... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16480... Training loss: 1.3031... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16481... Training loss: 1.2849... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16482... Training loss: 1.3326... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16483... Training loss: 1.2152... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16484... Training loss: 1.2862... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16485... Training loss: 1.3111... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16486... Training loss: 1.2491... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16487... Training loss: 1.3310... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16488... Training loss: 1.2937... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16489... Training loss: 1.3014... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16490... Training loss: 1.3056... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16491... Training loss: 1.2189... 0.0798 sec/batch
Epoch: 17/20... Training Step: 16492... Training loss: 1.2693... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16493... Training loss: 1.2290... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16494... Training loss: 1.2107... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16495... Training loss: 1.3109... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16496... Training loss: 1.3084... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16497... Training loss: 1.2562... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16498... Training loss: 1.2503... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16499... Training loss: 1.2513... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16500... Training loss: 1.2801... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16501... Training loss: 1.1103... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16502... Training loss: 1.2877... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16503... Training loss: 1.2744... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16504... Training loss: 1.2611... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16505... Training loss: 1.2242... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16506... Training loss: 1.2751... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16507... Training loss: 1.3096... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16508... Training loss: 1.2638... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16509... Training loss: 1.3195... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16510... Training loss: 1.2453... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16511... Training loss: 1.2960... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16512... Training loss: 1.3673... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16513... Training loss: 1.3781... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16514... Training loss: 1.3373... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16515... Training loss: 1.2472... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16516... Training loss: 1.2831... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16517... Training loss: 1.2523... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16518... Training loss: 1.3006... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16519... Training loss: 1.3283... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16520... Training loss: 1.2057... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16521... Training loss: 1.2549... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16522... Training loss: 1.2824... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16523... Training loss: 1.2676... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16524... Training loss: 1.3207... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16525... Training loss: 1.2902... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16526... Training loss: 1.3301... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16527... Training loss: 1.3697... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16528... Training loss: 1.3183... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16529... Training loss: 1.2353... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16530... Training loss: 1.2551... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16531... Training loss: 1.2418... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16532... Training loss: 1.3496... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16533... Training loss: 1.3320... 0.0834 sec/batch
Epoch: 17/20... Training Step: 16534... Training loss: 1.3589... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16535... Training loss: 1.3158... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16536... Training loss: 1.3719... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16537... Training loss: 1.3865... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16538... Training loss: 1.2571... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16539... Training loss: 1.3923... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16540... Training loss: 1.3167... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16541... Training loss: 1.3170... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16542... Training loss: 1.2462... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16543... Training loss: 1.3273... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16544... Training loss: 1.3604... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16545... Training loss: 1.3083... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16546... Training loss: 1.2991... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16547... Training loss: 1.2243... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16548... Training loss: 1.2650... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16549... Training loss: 1.3668... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16550... Training loss: 1.2650... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16551... Training loss: 1.2469... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16552... Training loss: 1.3417... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16553... Training loss: 1.3356... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16554... Training loss: 1.3167... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16555... Training loss: 1.3172... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16556... Training loss: 1.2646... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16557... Training loss: 1.3442... 0.0782 sec/batch
Epoch: 17/20... Training Step: 16558... Training loss: 1.3513... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16559... Training loss: 1.2729... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16560... Training loss: 1.2827... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16561... Training loss: 1.2448... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16562... Training loss: 1.3285... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16563... Training loss: 1.2825... 0.0825 sec/batch
Epoch: 17/20... Training Step: 16564... Training loss: 1.2953... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16565... Training loss: 1.2852... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16566... Training loss: 1.2268... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16567... Training loss: 1.2908... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16568... Training loss: 1.2901... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16569... Training loss: 1.3774... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16570... Training loss: 1.3059... 0.0738 sec/batch
Epoch: 17/20... Training Step: 16571... Training loss: 1.2963... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16572... Training loss: 1.2994... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16573... Training loss: 1.3034... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16574... Training loss: 1.2825... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16575... Training loss: 1.2897... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16576... Training loss: 1.2210... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16577... Training loss: 1.2301... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16578... Training loss: 1.2436... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16579... Training loss: 1.2817... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16580... Training loss: 1.3023... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16581... Training loss: 1.3046... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16582... Training loss: 1.2869... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16583... Training loss: 1.2808... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16584... Training loss: 1.3389... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16585... Training loss: 1.2482... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16586... Training loss: 1.2856... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16587... Training loss: 1.2787... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16588... Training loss: 1.2851... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16589... Training loss: 1.3369... 0.0785 sec/batch
Epoch: 17/20... Training Step: 16590... Training loss: 1.3346... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16591... Training loss: 1.3338... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16592... Training loss: 1.3879... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16593... Training loss: 1.3340... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16594... Training loss: 1.3069... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16595... Training loss: 1.2828... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16596... Training loss: 1.3246... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16597... Training loss: 1.3438... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16598... Training loss: 1.3723... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16599... Training loss: 1.2784... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16600... Training loss: 1.3490... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16601... Training loss: 1.3293... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16602... Training loss: 1.2629... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16603... Training loss: 1.2804... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16604... Training loss: 1.3323... 0.0813 sec/batch
Epoch: 17/20... Training Step: 16605... Training loss: 1.2878... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16606... Training loss: 1.3326... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16607... Training loss: 1.2889... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16608... Training loss: 1.2832... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16609... Training loss: 1.2724... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16610... Training loss: 1.3565... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16611... Training loss: 1.2726... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16612... Training loss: 1.2881... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16613... Training loss: 1.2558... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16614... Training loss: 1.2708... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16615... Training loss: 1.2984... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16616... Training loss: 1.3423... 0.0792 sec/batch
Epoch: 17/20... Training Step: 16617... Training loss: 1.3380... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16618... Training loss: 1.2903... 0.0786 sec/batch
Epoch: 17/20... Training Step: 16619... Training loss: 1.2849... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16620... Training loss: 1.3133... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16621... Training loss: 1.2429... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16622... Training loss: 1.2560... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16623... Training loss: 1.3787... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16624... Training loss: 1.2604... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16625... Training loss: 1.3126... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16626... Training loss: 1.3177... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16627... Training loss: 1.2057... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16628... Training loss: 1.3030... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16629... Training loss: 1.3008... 0.0738 sec/batch
Epoch: 17/20... Training Step: 16630... Training loss: 1.3093... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16631... Training loss: 1.3355... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16632... Training loss: 1.3508... 0.0739 sec/batch
Epoch: 17/20... Training Step: 16633... Training loss: 1.3017... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16634... Training loss: 1.2426... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16635... Training loss: 1.2926... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16636... Training loss: 1.2949... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16637... Training loss: 1.3216... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16638... Training loss: 1.3433... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16639... Training loss: 1.3106... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16640... Training loss: 1.2393... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16641... Training loss: 1.2664... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16642... Training loss: 1.3090... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16643... Training loss: 1.3165... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16644... Training loss: 1.3535... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16645... Training loss: 1.3396... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16646... Training loss: 1.2965... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16647... Training loss: 1.2865... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16648... Training loss: 1.2092... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16649... Training loss: 1.2585... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16650... Training loss: 1.2909... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16651... Training loss: 1.2651... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16652... Training loss: 1.2778... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16653... Training loss: 1.2791... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16654... Training loss: 1.2825... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16655... Training loss: 1.3300... 0.0796 sec/batch
Epoch: 17/20... Training Step: 16656... Training loss: 1.3333... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16657... Training loss: 1.2900... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16658... Training loss: 1.2894... 0.0791 sec/batch
Epoch: 17/20... Training Step: 16659... Training loss: 1.2949... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16660... Training loss: 1.3198... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16661... Training loss: 1.2835... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16662... Training loss: 1.3097... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16663... Training loss: 1.2833... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16664... Training loss: 1.2698... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16665... Training loss: 1.2919... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16666... Training loss: 1.2366... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16667... Training loss: 1.1966... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16668... Training loss: 1.2842... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16669... Training loss: 1.2595... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16670... Training loss: 1.2855... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16671... Training loss: 1.2391... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16672... Training loss: 1.2587... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16673... Training loss: 1.3139... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16674... Training loss: 1.3494... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16675... Training loss: 1.3648... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16676... Training loss: 1.3057... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16677... Training loss: 1.2504... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16678... Training loss: 1.2677... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16679... Training loss: 1.2693... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16680... Training loss: 1.3303... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16681... Training loss: 1.3449... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16682... Training loss: 1.3415... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16683... Training loss: 1.2811... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16684... Training loss: 1.3488... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16685... Training loss: 1.3090... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16686... Training loss: 1.3496... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16687... Training loss: 1.3425... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16688... Training loss: 1.3310... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16689... Training loss: 1.3670... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16690... Training loss: 1.4195... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16691... Training loss: 1.3139... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16692... Training loss: 1.2978... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16693... Training loss: 1.2698... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16694... Training loss: 1.2444... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16695... Training loss: 1.2904... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16696... Training loss: 1.2982... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16697... Training loss: 1.2626... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16698... Training loss: 1.3245... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16699... Training loss: 1.3259... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16700... Training loss: 1.3617... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16701... Training loss: 1.3271... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16702... Training loss: 1.2537... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16703... Training loss: 1.3436... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16704... Training loss: 1.3011... 0.0739 sec/batch
Epoch: 17/20... Training Step: 16705... Training loss: 1.3637... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16706... Training loss: 1.3319... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16707... Training loss: 1.2499... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16708... Training loss: 1.2873... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16709... Training loss: 1.3275... 0.0837 sec/batch
Epoch: 17/20... Training Step: 16710... Training loss: 1.3246... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16711... Training loss: 1.3131... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16712... Training loss: 1.2898... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16713... Training loss: 1.3022... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16714... Training loss: 1.4307... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16715... Training loss: 1.3403... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16716... Training loss: 1.3321... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16717... Training loss: 1.3253... 0.0753 sec/batch
Epoch: 17/20... Training Step: 16718... Training loss: 1.2590... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16719... Training loss: 1.3098... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16720... Training loss: 1.2971... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16721... Training loss: 1.2915... 0.0776 sec/batch
Epoch: 17/20... Training Step: 16722... Training loss: 1.3330... 0.0755 sec/batch
Epoch: 17/20... Training Step: 16723... Training loss: 1.2619... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16724... Training loss: 1.3856... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16725... Training loss: 1.3821... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16726... Training loss: 1.3017... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16727... Training loss: 1.3118... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16728... Training loss: 1.2949... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16729... Training loss: 1.3189... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16730... Training loss: 1.2783... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16731... Training loss: 1.3286... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16732... Training loss: 1.3543... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16733... Training loss: 1.3065... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16734... Training loss: 1.2591... 0.0783 sec/batch
Epoch: 17/20... Training Step: 16735... Training loss: 1.2735... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16736... Training loss: 1.2863... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16737... Training loss: 1.2589... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16738... Training loss: 1.2924... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16739... Training loss: 1.2814... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16740... Training loss: 1.3448... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16741... Training loss: 1.3093... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16742... Training loss: 1.3735... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16743... Training loss: 1.3294... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16744... Training loss: 1.3138... 0.0768 sec/batch
Epoch: 17/20... Training Step: 16745... Training loss: 1.3058... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16746... Training loss: 1.3249... 0.0778 sec/batch
Epoch: 17/20... Training Step: 16747... Training loss: 1.3106... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16748... Training loss: 1.3192... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16749... Training loss: 1.3047... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16750... Training loss: 1.3333... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16751... Training loss: 1.3270... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16752... Training loss: 1.2837... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16753... Training loss: 1.3090... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16754... Training loss: 1.2274... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16755... Training loss: 1.3290... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16756... Training loss: 1.2621... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16757... Training loss: 1.3352... 0.0774 sec/batch
Epoch: 17/20... Training Step: 16758... Training loss: 1.3680... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16759... Training loss: 1.2973... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16760... Training loss: 1.3353... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16761... Training loss: 1.3568... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16762... Training loss: 1.3536... 0.0791 sec/batch
Epoch: 17/20... Training Step: 16763... Training loss: 1.3605... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16764... Training loss: 1.2399... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16765... Training loss: 1.2710... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16766... Training loss: 1.2487... 0.0787 sec/batch
Epoch: 17/20... Training Step: 16767... Training loss: 1.3471... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16768... Training loss: 1.2570... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16769... Training loss: 1.3379... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16770... Training loss: 1.3406... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16771... Training loss: 1.2618... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16772... Training loss: 1.2226... 0.0784 sec/batch
Epoch: 17/20... Training Step: 16773... Training loss: 1.2632... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16774... Training loss: 1.2830... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16775... Training loss: 1.2841... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16776... Training loss: 1.2295... 0.0752 sec/batch
Epoch: 17/20... Training Step: 16777... Training loss: 1.2385... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16778... Training loss: 1.3160... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16779... Training loss: 1.2449... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16780... Training loss: 1.3021... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16781... Training loss: 1.2315... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16782... Training loss: 1.2064... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16783... Training loss: 1.2309... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16784... Training loss: 1.2860... 0.0767 sec/batch
Epoch: 17/20... Training Step: 16785... Training loss: 1.3679... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16786... Training loss: 1.2821... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16787... Training loss: 1.2845... 0.0763 sec/batch
Epoch: 17/20... Training Step: 16788... Training loss: 1.2717... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16789... Training loss: 1.3032... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16790... Training loss: 1.3385... 0.0785 sec/batch
Epoch: 17/20... Training Step: 16791... Training loss: 1.2725... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16792... Training loss: 1.3223... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16793... Training loss: 1.4241... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16794... Training loss: 1.3386... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16795... Training loss: 1.3068... 0.0779 sec/batch
Epoch: 17/20... Training Step: 16796... Training loss: 1.2567... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16797... Training loss: 1.2999... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16798... Training loss: 1.3505... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16799... Training loss: 1.3191... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16800... Training loss: 1.2921... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16801... Training loss: 1.3777... 0.0773 sec/batch
Epoch: 17/20... Training Step: 16802... Training loss: 1.3500... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16803... Training loss: 1.3043... 0.0741 sec/batch
Epoch: 17/20... Training Step: 16804... Training loss: 1.3941... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16805... Training loss: 1.2709... 0.0740 sec/batch
Epoch: 17/20... Training Step: 16806... Training loss: 1.3084... 0.0742 sec/batch
Epoch: 17/20... Training Step: 16807... Training loss: 1.2945... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16808... Training loss: 1.3242... 0.0769 sec/batch
Epoch: 17/20... Training Step: 16809... Training loss: 1.2755... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16810... Training loss: 1.2943... 0.0785 sec/batch
Epoch: 17/20... Training Step: 16811... Training loss: 1.2929... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16812... Training loss: 1.3710... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16813... Training loss: 1.3036... 0.0794 sec/batch
Epoch: 17/20... Training Step: 16814... Training loss: 1.3081... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16815... Training loss: 1.3474... 0.0748 sec/batch
Epoch: 17/20... Training Step: 16816... Training loss: 1.3208... 0.0799 sec/batch
Epoch: 17/20... Training Step: 16817... Training loss: 1.3533... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16818... Training loss: 1.2928... 0.0739 sec/batch
Epoch: 17/20... Training Step: 16819... Training loss: 1.3017... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16820... Training loss: 1.2688... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16821... Training loss: 1.3191... 0.0751 sec/batch
Epoch: 17/20... Training Step: 16822... Training loss: 1.2084... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16823... Training loss: 1.2977... 0.0770 sec/batch
Epoch: 17/20... Training Step: 16824... Training loss: 1.2426... 0.0743 sec/batch
Epoch: 17/20... Training Step: 16825... Training loss: 1.2736... 0.0772 sec/batch
Epoch: 17/20... Training Step: 16826... Training loss: 1.3382... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16827... Training loss: 1.3145... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16828... Training loss: 1.3034... 0.0766 sec/batch
Epoch: 17/20... Training Step: 16829... Training loss: 1.2896... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16830... Training loss: 1.3161... 0.0787 sec/batch
Epoch: 17/20... Training Step: 16831... Training loss: 1.2595... 0.0756 sec/batch
Epoch: 17/20... Training Step: 16832... Training loss: 1.2676... 0.0758 sec/batch
Epoch: 17/20... Training Step: 16833... Training loss: 1.2980... 0.0760 sec/batch
Epoch: 17/20... Training Step: 16834... Training loss: 1.2878... 0.0787 sec/batch
Epoch: 17/20... Training Step: 16835... Training loss: 1.3007... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16836... Training loss: 1.2926... 0.0771 sec/batch
Epoch: 17/20... Training Step: 16837... Training loss: 1.3514... 0.0785 sec/batch
Epoch: 17/20... Training Step: 16838... Training loss: 1.3285... 0.0750 sec/batch
Epoch: 17/20... Training Step: 16839... Training loss: 1.3581... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16840... Training loss: 1.3304... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16841... Training loss: 1.3581... 0.0749 sec/batch
Epoch: 17/20... Training Step: 16842... Training loss: 1.4646... 0.0745 sec/batch
Epoch: 17/20... Training Step: 16843... Training loss: 1.2865... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16844... Training loss: 1.3770... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16845... Training loss: 1.4069... 0.0777 sec/batch
Epoch: 17/20... Training Step: 16846... Training loss: 1.2721... 0.0781 sec/batch
Epoch: 17/20... Training Step: 16847... Training loss: 1.3283... 0.0747 sec/batch
Epoch: 17/20... Training Step: 16848... Training loss: 1.3348... 0.0757 sec/batch
Epoch: 17/20... Training Step: 16849... Training loss: 1.3798... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16850... Training loss: 1.3447... 0.0759 sec/batch
Epoch: 17/20... Training Step: 16851... Training loss: 1.3469... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16852... Training loss: 1.2757... 0.0780 sec/batch
Epoch: 17/20... Training Step: 16853... Training loss: 1.2463... 0.0744 sec/batch
Epoch: 17/20... Training Step: 16854... Training loss: 1.3952... 0.0775 sec/batch
Epoch: 17/20... Training Step: 16855... Training loss: 1.3296... 0.0765 sec/batch
Epoch: 17/20... Training Step: 16856... Training loss: 1.3196... 0.0761 sec/batch
Epoch: 17/20... Training Step: 16857... Training loss: 1.2973... 0.0795 sec/batch
Epoch: 17/20... Training Step: 16858... Training loss: 1.2801... 0.0788 sec/batch
Epoch: 17/20... Training Step: 16859... Training loss: 1.2733... 0.0746 sec/batch
Epoch: 17/20... Training Step: 16860... Training loss: 1.2762... 0.0764 sec/batch
Epoch: 17/20... Training Step: 16861... Training loss: 1.3100... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16862... Training loss: 1.2068... 0.0754 sec/batch
Epoch: 17/20... Training Step: 16863... Training loss: 1.2792... 0.0762 sec/batch
Epoch: 17/20... Training Step: 16864... Training loss: 1.3221... 0.0767 sec/batch
Epoch: 18/20... Training Step: 16865... Training loss: 1.3720... 0.0746 sec/batch
Epoch: 18/20... Training Step: 16866... Training loss: 1.2557... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16867... Training loss: 1.3646... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16868... Training loss: 1.3168... 0.0777 sec/batch
Epoch: 18/20... Training Step: 16869... Training loss: 1.2239... 0.0773 sec/batch
Epoch: 18/20... Training Step: 16870... Training loss: 1.1733... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16871... Training loss: 1.2895... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16872... Training loss: 1.3067... 0.0779 sec/batch
Epoch: 18/20... Training Step: 16873... Training loss: 1.3104... 0.0820 sec/batch
Epoch: 18/20... Training Step: 16874... Training loss: 1.2143... 0.0748 sec/batch
Epoch: 18/20... Training Step: 16875... Training loss: 1.3726... 0.0744 sec/batch
Epoch: 18/20... Training Step: 16876... Training loss: 1.2451... 0.0768 sec/batch
Epoch: 18/20... Training Step: 16877... Training loss: 1.2828... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16878... Training loss: 1.3021... 0.0744 sec/batch
Epoch: 18/20... Training Step: 16879... Training loss: 1.3266... 0.0756 sec/batch
Epoch: 18/20... Training Step: 16880... Training loss: 1.2486... 0.0753 sec/batch
Epoch: 18/20... Training Step: 16881... Training loss: 1.2751... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16882... Training loss: 1.3102... 0.0793 sec/batch
Epoch: 18/20... Training Step: 16883... Training loss: 1.2915... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16884... Training loss: 1.2739... 0.0771 sec/batch
Epoch: 18/20... Training Step: 16885... Training loss: 1.3086... 0.0766 sec/batch
Epoch: 18/20... Training Step: 16886... Training loss: 1.2495... 0.0745 sec/batch
Epoch: 18/20... Training Step: 16887... Training loss: 1.2770... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16888... Training loss: 1.2147... 0.0782 sec/batch
Epoch: 18/20... Training Step: 16889... Training loss: 1.3157... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16890... Training loss: 1.2348... 0.0781 sec/batch
Epoch: 18/20... Training Step: 16891... Training loss: 1.2850... 0.0804 sec/batch
Epoch: 18/20... Training Step: 16892... Training loss: 1.2139... 0.0761 sec/batch
Epoch: 18/20... Training Step: 16893... Training loss: 1.2921... 0.0756 sec/batch
Epoch: 18/20... Training Step: 16894... Training loss: 1.3347... 0.0771 sec/batch
Epoch: 18/20... Training Step: 16895... Training loss: 1.3326... 0.0763 sec/batch
Epoch: 18/20... Training Step: 16896... Training loss: 1.3197... 0.0763 sec/batch
Epoch: 18/20... Training Step: 16897... Training loss: 1.2838... 0.0795 sec/batch
Epoch: 18/20... Training Step: 16898... Training loss: 1.3392... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16899... Training loss: 1.3755... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16900... Training loss: 1.3589... 0.0745 sec/batch
Epoch: 18/20... Training Step: 16901... Training loss: 1.2849... 0.0740 sec/batch
Epoch: 18/20... Training Step: 16902... Training loss: 1.3540... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16903... Training loss: 1.3023... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16904... Training loss: 1.2707... 0.0761 sec/batch
Epoch: 18/20... Training Step: 16905... Training loss: 1.3016... 0.0768 sec/batch
Epoch: 18/20... Training Step: 16906... Training loss: 1.2386... 0.0847 sec/batch
Epoch: 18/20... Training Step: 16907... Training loss: 1.2404... 0.0767 sec/batch
Epoch: 18/20... Training Step: 16908... Training loss: 1.2781... 0.0743 sec/batch
Epoch: 18/20... Training Step: 16909... Training loss: 1.3122... 0.0771 sec/batch
Epoch: 18/20... Training Step: 16910... Training loss: 1.2822... 0.0780 sec/batch
Epoch: 18/20... Training Step: 16911... Training loss: 1.3262... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16912... Training loss: 1.2336... 0.0753 sec/batch
Epoch: 18/20... Training Step: 16913... Training loss: 1.3465... 0.0759 sec/batch
Epoch: 18/20... Training Step: 16914... Training loss: 1.3112... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16915... Training loss: 1.2245... 0.0818 sec/batch
Epoch: 18/20... Training Step: 16916... Training loss: 1.3362... 0.0756 sec/batch
Epoch: 18/20... Training Step: 16917... Training loss: 1.2218... 0.0766 sec/batch
Epoch: 18/20... Training Step: 16918... Training loss: 1.2384... 0.0798 sec/batch
Epoch: 18/20... Training Step: 16919... Training loss: 1.2697... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16920... Training loss: 1.2413... 0.0765 sec/batch
Epoch: 18/20... Training Step: 16921... Training loss: 1.2925... 0.0763 sec/batch
Epoch: 18/20... Training Step: 16922... Training loss: 1.3409... 0.0765 sec/batch
Epoch: 18/20... Training Step: 16923... Training loss: 1.2868... 0.0766 sec/batch
Epoch: 18/20... Training Step: 16924... Training loss: 1.3024... 0.0743 sec/batch
Epoch: 18/20... Training Step: 16925... Training loss: 1.1992... 0.0771 sec/batch
Epoch: 18/20... Training Step: 16926... Training loss: 1.2194... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16927... Training loss: 1.2820... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16928... Training loss: 1.2317... 0.0756 sec/batch
Epoch: 18/20... Training Step: 16929... Training loss: 1.3088... 0.0769 sec/batch
Epoch: 18/20... Training Step: 16930... Training loss: 1.2766... 0.0775 sec/batch
Epoch: 18/20... Training Step: 16931... Training loss: 1.2427... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16932... Training loss: 1.2922... 0.0758 sec/batch
Epoch: 18/20... Training Step: 16933... Training loss: 1.3320... 0.0789 sec/batch
Epoch: 18/20... Training Step: 16934... Training loss: 1.2982... 0.0744 sec/batch
Epoch: 18/20... Training Step: 16935... Training loss: 1.3193... 0.0771 sec/batch
Epoch: 18/20... Training Step: 16936... Training loss: 1.3458... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16937... Training loss: 1.2932... 0.0753 sec/batch
Epoch: 18/20... Training Step: 16938... Training loss: 1.2409... 0.0761 sec/batch
Epoch: 18/20... Training Step: 16939... Training loss: 1.2644... 0.0757 sec/batch
Epoch: 18/20... Training Step: 16940... Training loss: 1.2371... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16941... Training loss: 1.2634... 0.0753 sec/batch
Epoch: 18/20... Training Step: 16942... Training loss: 1.2653... 0.0765 sec/batch
Epoch: 18/20... Training Step: 16943... Training loss: 1.2888... 0.0741 sec/batch
Epoch: 18/20... Training Step: 16944... Training loss: 1.3370... 0.0757 sec/batch
Epoch: 18/20... Training Step: 16945... Training loss: 1.3087... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16946... Training loss: 1.2503... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16947... Training loss: 1.3122... 0.0776 sec/batch
Epoch: 18/20... Training Step: 16948... Training loss: 1.2903... 0.0760 sec/batch
Epoch: 18/20... Training Step: 16949... Training loss: 1.2598... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16950... Training loss: 1.3394... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16951... Training loss: 1.2890... 0.0782 sec/batch
Epoch: 18/20... Training Step: 16952... Training loss: 1.2916... 0.0744 sec/batch
Epoch: 18/20... Training Step: 16953... Training loss: 1.3285... 0.0766 sec/batch
Epoch: 18/20... Training Step: 16954... Training loss: 1.3411... 0.0770 sec/batch
Epoch: 18/20... Training Step: 16955... Training loss: 1.3429... 0.0750 sec/batch
Epoch: 18/20... Training Step: 16956... Training loss: 1.2625... 0.0742 sec/batch
Epoch: 18/20... Training Step: 16957... Training loss: 1.2500... 0.0759 sec/batch
Epoch: 18/20... Training Step: 16958... Training loss: 1.2054... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16959... Training loss: 1.1978... 0.0744 sec/batch
Epoch: 18/20... Training Step: 16960... Training loss: 1.2923... 0.0780 sec/batch
Epoch: 18/20... Training Step: 16961... Training loss: 1.2743... 0.0766 sec/batch
Epoch: 18/20... Training Step: 16962... Training loss: 1.2609... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16963... Training loss: 1.2544... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16964... Training loss: 1.2483... 0.0758 sec/batch
Epoch: 18/20... Training Step: 16965... Training loss: 1.2388... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16966... Training loss: 1.2595... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16967... Training loss: 1.2931... 0.0757 sec/batch
Epoch: 18/20... Training Step: 16968... Training loss: 1.3202... 0.0769 sec/batch
Epoch: 18/20... Training Step: 16969... Training loss: 1.3079... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16970... Training loss: 1.2904... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16971... Training loss: 1.2211... 0.0753 sec/batch
Epoch: 18/20... Training Step: 16972... Training loss: 1.3080... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16973... Training loss: 1.2898... 0.0758 sec/batch
Epoch: 18/20... Training Step: 16974... Training loss: 1.3265... 0.0764 sec/batch
Epoch: 18/20... Training Step: 16975... Training loss: 1.3093... 0.0791 sec/batch
Epoch: 18/20... Training Step: 16976... Training loss: 1.2885... 0.0752 sec/batch
Epoch: 18/20... Training Step: 16977... Training loss: 1.3531... 0.0776 sec/batch
Epoch: 18/20... Training Step: 16978... Training loss: 1.3111... 0.0795 sec/batch
Epoch: 18/20... Training Step: 16979... Training loss: 1.2418... 0.0765 sec/batch
Epoch: 18/20... Training Step: 16980... Training loss: 1.3810... 0.0772 sec/batch
Epoch: 18/20... Training Step: 16981... Training loss: 1.3351... 0.0759 sec/batch
Epoch: 18/20... Training Step: 16982... Training loss: 1.3646... 0.0774 sec/batch
Epoch: 18/20... Training Step: 16983... Training loss: 1.3046... 0.0780 sec/batch
Epoch: 18/20... Training Step: 16984... Training loss: 1.2780... 0.0787 sec/batch
Epoch: 18/20... Training Step: 16985... Training loss: 1.4049... 0.0749 sec/batch
Epoch: 18/20... Training Step: 16986... Training loss: 1.2710... 0.0755 sec/batch
Epoch: 18/20... Training Step: 16987... Training loss: 1.3101... 0.0780 sec/batch
Epoch: 18/20... Training Step: 16988... Training loss: 1.2845... 0.0761 sec/batch
Epoch: 18/20... Training Step: 16989... Training loss: 1.3359... 0.0778 sec/batch
Epoch: 18/20... Training Step: 16990... Training loss: 1.3451... 0.0779 sec/batch
Epoch: 18/20... Training Step: 16991... Training loss: 1.3423... 0.0765 sec/batch
Epoch: 18/20... Training Step: 16992... Training loss: 1.3227... 0.0761 sec/batch
Epoch: 18/20... Training Step: 16993... Training loss: 1.2413... 0.0783 sec/batch
Epoch: 18/20... Training Step: 16994... Training loss: 1.3115... 0.0763 sec/batch
Epoch: 18/20... Training Step: 16995... Training loss: 1.2745... 0.0762 sec/batch
Epoch: 18/20... Training Step: 16996... Training loss: 1.3279... 0.0773 sec/batch
Epoch: 18/20... Training Step: 16997... Training loss: 1.3309... 0.0772 sec/batch
Epoch: 18/20... Training Step: 16998... Training loss: 1.2829... 0.0741 sec/batch
Epoch: 18/20... Training Step: 16999... Training loss: 1.3297... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17000... Training loss: 1.3121... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17001... Training loss: 1.3483... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17002... Training loss: 1.3884... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17003... Training loss: 1.3338... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17004... Training loss: 1.2862... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17005... Training loss: 1.3012... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17006... Training loss: 1.2700... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17007... Training loss: 1.2844... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17008... Training loss: 1.2919... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17009... Training loss: 1.1740... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17010... Training loss: 1.3437... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17011... Training loss: 1.2704... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17012... Training loss: 1.2221... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17013... Training loss: 1.2622... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17014... Training loss: 1.2335... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17015... Training loss: 1.1790... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17016... Training loss: 1.2663... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17017... Training loss: 1.2259... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17018... Training loss: 1.2670... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17019... Training loss: 1.2682... 0.0790 sec/batch
Epoch: 18/20... Training Step: 17020... Training loss: 1.2543... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17021... Training loss: 1.2804... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17022... Training loss: 1.2994... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17023... Training loss: 1.2917... 0.0786 sec/batch
Epoch: 18/20... Training Step: 17024... Training loss: 1.3268... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17025... Training loss: 1.2497... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17026... Training loss: 1.2685... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17027... Training loss: 1.2682... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17028... Training loss: 1.3299... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17029... Training loss: 1.3034... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17030... Training loss: 1.3682... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17031... Training loss: 1.3041... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17032... Training loss: 1.2859... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17033... Training loss: 1.3360... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17034... Training loss: 1.2783... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17035... Training loss: 1.3072... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17036... Training loss: 1.2898... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17037... Training loss: 1.2848... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17038... Training loss: 1.3207... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17039... Training loss: 1.2673... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17040... Training loss: 1.3045... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17041... Training loss: 1.2970... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17042... Training loss: 1.3826... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17043... Training loss: 1.3461... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17044... Training loss: 1.2646... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17045... Training loss: 1.3161... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17046... Training loss: 1.3808... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17047... Training loss: 1.2845... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17048... Training loss: 1.3190... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17049... Training loss: 1.3011... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17050... Training loss: 1.3431... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17051... Training loss: 1.3124... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17052... Training loss: 1.2713... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17053... Training loss: 1.2916... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17054... Training loss: 1.3443... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17055... Training loss: 1.3855... 0.0795 sec/batch
Epoch: 18/20... Training Step: 17056... Training loss: 1.3217... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17057... Training loss: 1.3286... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17058... Training loss: 1.3893... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17059... Training loss: 1.2995... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17060... Training loss: 1.2499... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17061... Training loss: 1.3529... 0.0794 sec/batch
Epoch: 18/20... Training Step: 17062... Training loss: 1.3103... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17063... Training loss: 1.3480... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17064... Training loss: 1.2829... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17065... Training loss: 1.3069... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17066... Training loss: 1.2722... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17067... Training loss: 1.2877... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17068... Training loss: 1.2826... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17069... Training loss: 1.2669... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17070... Training loss: 1.3533... 0.0842 sec/batch
Epoch: 18/20... Training Step: 17071... Training loss: 1.2916... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17072... Training loss: 1.2745... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17073... Training loss: 1.3256... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17074... Training loss: 1.3068... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17075... Training loss: 1.2407... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17076... Training loss: 1.3226... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17077... Training loss: 1.2595... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17078... Training loss: 1.3579... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17079... Training loss: 1.2730... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17080... Training loss: 1.2803... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17081... Training loss: 1.2749... 0.0790 sec/batch
Epoch: 18/20... Training Step: 17082... Training loss: 1.3047... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17083... Training loss: 1.2893... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17084... Training loss: 1.2408... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17085... Training loss: 1.2479... 0.0790 sec/batch
Epoch: 18/20... Training Step: 17086... Training loss: 1.4032... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17087... Training loss: 1.2927... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17088... Training loss: 1.3109... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17089... Training loss: 1.3400... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17090... Training loss: 1.2608... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17091... Training loss: 1.3057... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17092... Training loss: 1.3227... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17093... Training loss: 1.2299... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17094... Training loss: 1.2390... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17095... Training loss: 1.3068... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17096... Training loss: 1.2978... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17097... Training loss: 1.2824... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17098... Training loss: 1.3512... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17099... Training loss: 1.2878... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17100... Training loss: 1.3167... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17101... Training loss: 1.2437... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17102... Training loss: 1.3301... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17103... Training loss: 1.3565... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17104... Training loss: 1.2911... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17105... Training loss: 1.2862... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17106... Training loss: 1.2995... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17107... Training loss: 1.3277... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17108... Training loss: 1.3281... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17109... Training loss: 1.2759... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17110... Training loss: 1.3192... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17111... Training loss: 1.2537... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17112... Training loss: 1.2250... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17113... Training loss: 1.2744... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17114... Training loss: 1.2987... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17115... Training loss: 1.2634... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17116... Training loss: 1.3013... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17117... Training loss: 1.2579... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17118... Training loss: 1.2564... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17119... Training loss: 1.2114... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17120... Training loss: 1.2578... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17121... Training loss: 1.3570... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17122... Training loss: 1.2788... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17123... Training loss: 1.2827... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17124... Training loss: 1.2854... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17125... Training loss: 1.3347... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17126... Training loss: 1.3397... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17127... Training loss: 1.3327... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17128... Training loss: 1.3003... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17129... Training loss: 1.3236... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17130... Training loss: 1.2046... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17131... Training loss: 1.2504... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17132... Training loss: 1.2828... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17133... Training loss: 1.2351... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17134... Training loss: 1.2422... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17135... Training loss: 1.2927... 0.0739 sec/batch
Epoch: 18/20... Training Step: 17136... Training loss: 1.2661... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17137... Training loss: 1.2058... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17138... Training loss: 1.2191... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17139... Training loss: 1.2722... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17140... Training loss: 1.2882... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17141... Training loss: 1.2864... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17142... Training loss: 1.1914... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17143... Training loss: 1.2430... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17144... Training loss: 1.3507... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17145... Training loss: 1.2797... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17146... Training loss: 1.2573... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17147... Training loss: 1.3058... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17148... Training loss: 1.2369... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17149... Training loss: 1.1617... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17150... Training loss: 1.2263... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17151... Training loss: 1.2959... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17152... Training loss: 1.2869... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17153... Training loss: 1.2363... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17154... Training loss: 1.2223... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17155... Training loss: 1.2578... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17156... Training loss: 1.2835... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17157... Training loss: 1.2192... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17158... Training loss: 1.2499... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17159... Training loss: 1.2773... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17160... Training loss: 1.3165... 0.0805 sec/batch
Epoch: 18/20... Training Step: 17161... Training loss: 1.2453... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17162... Training loss: 1.3717... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17163... Training loss: 1.2579... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17164... Training loss: 1.3216... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17165... Training loss: 1.3663... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17166... Training loss: 1.2379... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17167... Training loss: 1.2581... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17168... Training loss: 1.2939... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17169... Training loss: 1.2709... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17170... Training loss: 1.3538... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17171... Training loss: 1.2556... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17172... Training loss: 1.2294... 0.0782 sec/batch
Epoch: 18/20... Training Step: 17173... Training loss: 1.3713... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17174... Training loss: 1.2688... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17175... Training loss: 1.3634... 0.0797 sec/batch
Epoch: 18/20... Training Step: 17176... Training loss: 1.2858... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17177... Training loss: 1.2687... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17178... Training loss: 1.2853... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17179... Training loss: 1.2950... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17180... Training loss: 1.2466... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17181... Training loss: 1.2523... 0.0809 sec/batch
Epoch: 18/20... Training Step: 17182... Training loss: 1.2343... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17183... Training loss: 1.2797... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17184... Training loss: 1.3263... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17185... Training loss: 1.3007... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17186... Training loss: 1.2156... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17187... Training loss: 1.2702... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17188... Training loss: 1.2771... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17189... Training loss: 1.3000... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17190... Training loss: 1.2882... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17191... Training loss: 1.2657... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17192... Training loss: 1.2730... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17193... Training loss: 1.3029... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17194... Training loss: 1.2935... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17195... Training loss: 1.3229... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17196... Training loss: 1.2458... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17197... Training loss: 1.3810... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17198... Training loss: 1.3946... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17199... Training loss: 1.3694... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17200... Training loss: 1.2938... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17201... Training loss: 1.2413... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17202... Training loss: 1.2692... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17203... Training loss: 1.2990... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17204... Training loss: 1.2650... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17205... Training loss: 1.2653... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17206... Training loss: 1.2541... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17207... Training loss: 1.2231... 0.0829 sec/batch
Epoch: 18/20... Training Step: 17208... Training loss: 1.3221... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17209... Training loss: 1.2702... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17210... Training loss: 1.2984... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17211... Training loss: 1.3065... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17212... Training loss: 1.3279... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17213... Training loss: 1.3412... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17214... Training loss: 1.3444... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17215... Training loss: 1.3124... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17216... Training loss: 1.2573... 0.0837 sec/batch
Epoch: 18/20... Training Step: 17217... Training loss: 1.2823... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17218... Training loss: 1.2685... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17219... Training loss: 1.2877... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17220... Training loss: 1.2898... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17221... Training loss: 1.2803... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17222... Training loss: 1.2889... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17223... Training loss: 1.3065... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17224... Training loss: 1.2904... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17225... Training loss: 1.3080... 0.0797 sec/batch
Epoch: 18/20... Training Step: 17226... Training loss: 1.3506... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17227... Training loss: 1.2598... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17228... Training loss: 1.3057... 0.0793 sec/batch
Epoch: 18/20... Training Step: 17229... Training loss: 1.3041... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17230... Training loss: 1.2500... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17231... Training loss: 1.2815... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17232... Training loss: 1.2161... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17233... Training loss: 1.2486... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17234... Training loss: 1.2054... 0.0824 sec/batch
Epoch: 18/20... Training Step: 17235... Training loss: 1.3499... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17236... Training loss: 1.2361... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17237... Training loss: 1.2650... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17238... Training loss: 1.2417... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17239... Training loss: 1.2551... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17240... Training loss: 1.2416... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17241... Training loss: 1.2640... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17242... Training loss: 1.3068... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17243... Training loss: 1.2516... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17244... Training loss: 1.3229... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17245... Training loss: 1.3030... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17246... Training loss: 1.2502... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17247... Training loss: 1.3030... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17248... Training loss: 1.2965... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17249... Training loss: 1.2915... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17250... Training loss: 1.2956... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17251... Training loss: 1.3190... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17252... Training loss: 1.3233... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17253... Training loss: 1.3239... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17254... Training loss: 1.2614... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17255... Training loss: 1.3071... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17256... Training loss: 1.2881... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17257... Training loss: 1.3218... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17258... Training loss: 1.2588... 0.0793 sec/batch
Epoch: 18/20... Training Step: 17259... Training loss: 1.2851... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17260... Training loss: 1.2481... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17261... Training loss: 1.2456... 0.0799 sec/batch
Epoch: 18/20... Training Step: 17262... Training loss: 1.3471... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17263... Training loss: 1.3026... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17264... Training loss: 1.2586... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17265... Training loss: 1.2488... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17266... Training loss: 1.2893... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17267... Training loss: 1.2567... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17268... Training loss: 1.3258... 0.0797 sec/batch
Epoch: 18/20... Training Step: 17269... Training loss: 1.2677... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17270... Training loss: 1.2720... 0.0798 sec/batch
Epoch: 18/20... Training Step: 17271... Training loss: 1.2087... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17272... Training loss: 1.3381... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17273... Training loss: 1.2459... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17274... Training loss: 1.2200... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17275... Training loss: 1.2014... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17276... Training loss: 1.2442... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17277... Training loss: 1.2604... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17278... Training loss: 1.2069... 0.0737 sec/batch
Epoch: 18/20... Training Step: 17279... Training loss: 1.2387... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17280... Training loss: 1.2970... 0.0800 sec/batch
Epoch: 18/20... Training Step: 17281... Training loss: 1.2474... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17282... Training loss: 1.2391... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17283... Training loss: 1.3195... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17284... Training loss: 1.3404... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17285... Training loss: 1.2061... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17286... Training loss: 1.2773... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17287... Training loss: 1.3033... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17288... Training loss: 1.2836... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17289... Training loss: 1.2858... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17290... Training loss: 1.2885... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17291... Training loss: 1.3383... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17292... Training loss: 1.3683... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17293... Training loss: 1.3488... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17294... Training loss: 1.3913... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17295... Training loss: 1.2500... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17296... Training loss: 1.3057... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17297... Training loss: 1.3324... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17298... Training loss: 1.3153... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17299... Training loss: 1.3332... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17300... Training loss: 1.2547... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17301... Training loss: 1.3168... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17302... Training loss: 1.3894... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17303... Training loss: 1.3042... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17304... Training loss: 1.3280... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17305... Training loss: 1.2726... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17306... Training loss: 1.3373... 0.0800 sec/batch
Epoch: 18/20... Training Step: 17307... Training loss: 1.3263... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17308... Training loss: 1.3177... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17309... Training loss: 1.2839... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17310... Training loss: 1.3959... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17311... Training loss: 1.3093... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17312... Training loss: 1.3047... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17313... Training loss: 1.3681... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17314... Training loss: 1.3719... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17315... Training loss: 1.2934... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17316... Training loss: 1.2908... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17317... Training loss: 1.3314... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17318... Training loss: 1.3078... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17319... Training loss: 1.3052... 0.0793 sec/batch
Epoch: 18/20... Training Step: 17320... Training loss: 1.2884... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17321... Training loss: 1.2867... 0.0795 sec/batch
Epoch: 18/20... Training Step: 17322... Training loss: 1.2888... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17323... Training loss: 1.2773... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17324... Training loss: 1.3084... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17325... Training loss: 1.3054... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17326... Training loss: 1.2912... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17327... Training loss: 1.2515... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17328... Training loss: 1.3328... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17329... Training loss: 1.2912... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17330... Training loss: 1.2539... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17331... Training loss: 1.1942... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17332... Training loss: 1.2875... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17333... Training loss: 1.2808... 0.0803 sec/batch
Epoch: 18/20... Training Step: 17334... Training loss: 1.3547... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17335... Training loss: 1.3337... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17336... Training loss: 1.2750... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17337... Training loss: 1.2734... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17338... Training loss: 1.3063... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17339... Training loss: 1.1790... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17340... Training loss: 1.2566... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17341... Training loss: 1.3014... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17342... Training loss: 1.2762... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17343... Training loss: 1.2302... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17344... Training loss: 1.2472... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17345... Training loss: 1.2376... 0.0805 sec/batch
Epoch: 18/20... Training Step: 17346... Training loss: 1.3122... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17347... Training loss: 1.2867... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17348... Training loss: 1.2765... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17349... Training loss: 1.2227... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17350... Training loss: 1.2650... 0.0740 sec/batch
Epoch: 18/20... Training Step: 17351... Training loss: 1.3198... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17352... Training loss: 1.2632... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17353... Training loss: 1.2161... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17354... Training loss: 1.2255... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17355... Training loss: 1.2502... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17356... Training loss: 1.2656... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17357... Training loss: 1.2719... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17358... Training loss: 1.2582... 0.0782 sec/batch
Epoch: 18/20... Training Step: 17359... Training loss: 1.2420... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17360... Training loss: 1.2411... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17361... Training loss: 1.2593... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17362... Training loss: 1.2672... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17363... Training loss: 1.2425... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17364... Training loss: 1.2738... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17365... Training loss: 1.2393... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17366... Training loss: 1.2623... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17367... Training loss: 1.2805... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17368... Training loss: 1.2665... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17369... Training loss: 1.2814... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17370... Training loss: 1.2087... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17371... Training loss: 1.2681... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17372... Training loss: 1.2610... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17373... Training loss: 1.2504... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17374... Training loss: 1.2408... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17375... Training loss: 1.3225... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17376... Training loss: 1.2760... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17377... Training loss: 1.3296... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17378... Training loss: 1.3115... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17379... Training loss: 1.2895... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17380... Training loss: 1.2841... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17381... Training loss: 1.2876... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17382... Training loss: 1.2878... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17383... Training loss: 1.3020... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17384... Training loss: 1.2437... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17385... Training loss: 1.2949... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17386... Training loss: 1.2472... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17387... Training loss: 1.3715... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17388... Training loss: 1.2777... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17389... Training loss: 1.2834... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17390... Training loss: 1.2732... 0.0814 sec/batch
Epoch: 18/20... Training Step: 17391... Training loss: 1.2453... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17392... Training loss: 1.2774... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17393... Training loss: 1.3030... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17394... Training loss: 1.3337... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17395... Training loss: 1.3807... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17396... Training loss: 1.3244... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17397... Training loss: 1.3235... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17398... Training loss: 1.3395... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17399... Training loss: 1.2853... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17400... Training loss: 1.2582... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17401... Training loss: 1.2397... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17402... Training loss: 1.2571... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17403... Training loss: 1.2632... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17404... Training loss: 1.2743... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17405... Training loss: 1.3734... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17406... Training loss: 1.2750... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17407... Training loss: 1.2487... 0.0787 sec/batch
Epoch: 18/20... Training Step: 17408... Training loss: 1.2756... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17409... Training loss: 1.2344... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17410... Training loss: 1.2583... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17411... Training loss: 1.3223... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17412... Training loss: 1.2934... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17413... Training loss: 1.2393... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17414... Training loss: 1.2494... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17415... Training loss: 1.2741... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17416... Training loss: 1.2045... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17417... Training loss: 1.2291... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17418... Training loss: 1.2626... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17419... Training loss: 1.2945... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17420... Training loss: 1.2361... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17421... Training loss: 1.2333... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17422... Training loss: 1.2561... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17423... Training loss: 1.2918... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17424... Training loss: 1.1988... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17425... Training loss: 1.2723... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17426... Training loss: 1.2441... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17427... Training loss: 1.2506... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17428... Training loss: 1.2642... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17429... Training loss: 1.2722... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17430... Training loss: 1.2991... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17431... Training loss: 1.2412... 0.0784 sec/batch
Epoch: 18/20... Training Step: 17432... Training loss: 1.2104... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17433... Training loss: 1.1877... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17434... Training loss: 1.2181... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17435... Training loss: 1.2129... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17436... Training loss: 1.2490... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17437... Training loss: 1.3249... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17438... Training loss: 1.3197... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17439... Training loss: 1.3037... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17440... Training loss: 1.2689... 0.0805 sec/batch
Epoch: 18/20... Training Step: 17441... Training loss: 1.2691... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17442... Training loss: 1.4117... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17443... Training loss: 1.2985... 0.0813 sec/batch
Epoch: 18/20... Training Step: 17444... Training loss: 1.2547... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17445... Training loss: 1.2268... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17446... Training loss: 1.2363... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17447... Training loss: 1.3165... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17448... Training loss: 1.2014... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17449... Training loss: 1.2731... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17450... Training loss: 1.3180... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17451... Training loss: 1.3174... 0.0786 sec/batch
Epoch: 18/20... Training Step: 17452... Training loss: 1.3219... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17453... Training loss: 1.2596... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17454... Training loss: 1.2573... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17455... Training loss: 1.2826... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17456... Training loss: 1.2918... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17457... Training loss: 1.2160... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17458... Training loss: 1.2559... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17459... Training loss: 1.1913... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17460... Training loss: 1.2905... 0.0789 sec/batch
Epoch: 18/20... Training Step: 17461... Training loss: 1.1774... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17462... Training loss: 1.1936... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17463... Training loss: 1.2447... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17464... Training loss: 1.2456... 0.0784 sec/batch
Epoch: 18/20... Training Step: 17465... Training loss: 1.3198... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17466... Training loss: 1.2503... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17467... Training loss: 1.2461... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17468... Training loss: 1.2746... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17469... Training loss: 1.2293... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17470... Training loss: 1.2604... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17471... Training loss: 1.3261... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17472... Training loss: 1.3326... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17473... Training loss: 1.2476... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17474... Training loss: 1.3179... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17475... Training loss: 1.1973... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17476... Training loss: 1.2970... 0.0786 sec/batch
Epoch: 18/20... Training Step: 17477... Training loss: 1.2943... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17478... Training loss: 1.2661... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17479... Training loss: 1.3383... 0.0795 sec/batch
Epoch: 18/20... Training Step: 17480... Training loss: 1.2824... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17481... Training loss: 1.2971... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17482... Training loss: 1.2931... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17483... Training loss: 1.2306... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17484... Training loss: 1.2512... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17485... Training loss: 1.2402... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17486... Training loss: 1.2141... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17487... Training loss: 1.3194... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17488... Training loss: 1.3012... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17489... Training loss: 1.2358... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17490... Training loss: 1.2545... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17491... Training loss: 1.2655... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17492... Training loss: 1.2781... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17493... Training loss: 1.0967... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17494... Training loss: 1.2950... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17495... Training loss: 1.2590... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17496... Training loss: 1.2512... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17497... Training loss: 1.2154... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17498... Training loss: 1.2602... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17499... Training loss: 1.2923... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17500... Training loss: 1.2814... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17501... Training loss: 1.3503... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17502... Training loss: 1.2338... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17503... Training loss: 1.3143... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17504... Training loss: 1.3495... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17505... Training loss: 1.3785... 0.0787 sec/batch
Epoch: 18/20... Training Step: 17506... Training loss: 1.3268... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17507... Training loss: 1.2505... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17508... Training loss: 1.2939... 0.0782 sec/batch
Epoch: 18/20... Training Step: 17509... Training loss: 1.2447... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17510... Training loss: 1.2861... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17511... Training loss: 1.2930... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17512... Training loss: 1.2023... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17513... Training loss: 1.2386... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17514... Training loss: 1.2678... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17515... Training loss: 1.2538... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17516... Training loss: 1.3199... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17517... Training loss: 1.2590... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17518... Training loss: 1.3251... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17519... Training loss: 1.3435... 0.0793 sec/batch
Epoch: 18/20... Training Step: 17520... Training loss: 1.2944... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17521... Training loss: 1.1951... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17522... Training loss: 1.2293... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17523... Training loss: 1.2481... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17524... Training loss: 1.3359... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17525... Training loss: 1.3110... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17526... Training loss: 1.3589... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17527... Training loss: 1.2980... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17528... Training loss: 1.3702... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17529... Training loss: 1.3792... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17530... Training loss: 1.2499... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17531... Training loss: 1.3778... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17532... Training loss: 1.2908... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17533... Training loss: 1.2952... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17534... Training loss: 1.2636... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17535... Training loss: 1.3143... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17536... Training loss: 1.3598... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17537... Training loss: 1.2918... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17538... Training loss: 1.3001... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17539... Training loss: 1.2344... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17540... Training loss: 1.2656... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17541... Training loss: 1.3849... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17542... Training loss: 1.2751... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17543... Training loss: 1.2137... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17544... Training loss: 1.3214... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17545... Training loss: 1.3421... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17546... Training loss: 1.3109... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17547... Training loss: 1.3242... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17548... Training loss: 1.2576... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17549... Training loss: 1.3348... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17550... Training loss: 1.3337... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17551... Training loss: 1.2851... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17552... Training loss: 1.2898... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17553... Training loss: 1.2526... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17554... Training loss: 1.3130... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17555... Training loss: 1.2897... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17556... Training loss: 1.2942... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17557... Training loss: 1.2741... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17558... Training loss: 1.2157... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17559... Training loss: 1.2888... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17560... Training loss: 1.2617... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17561... Training loss: 1.3547... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17562... Training loss: 1.3254... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17563... Training loss: 1.2728... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17564... Training loss: 1.2659... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17565... Training loss: 1.3167... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17566... Training loss: 1.2751... 0.0830 sec/batch
Epoch: 18/20... Training Step: 17567... Training loss: 1.2640... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17568... Training loss: 1.2040... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17569... Training loss: 1.2185... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17570... Training loss: 1.2313... 0.0739 sec/batch
Epoch: 18/20... Training Step: 17571... Training loss: 1.2867... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17572... Training loss: 1.2895... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17573... Training loss: 1.2971... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17574... Training loss: 1.2639... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17575... Training loss: 1.2572... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17576... Training loss: 1.3404... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17577... Training loss: 1.2321... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17578... Training loss: 1.2766... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17579... Training loss: 1.2737... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17580... Training loss: 1.3483... 0.0739 sec/batch
Epoch: 18/20... Training Step: 17581... Training loss: 1.3363... 0.0734 sec/batch
Epoch: 18/20... Training Step: 17582... Training loss: 1.3177... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17583... Training loss: 1.3224... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17584... Training loss: 1.3799... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17585... Training loss: 1.3295... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17586... Training loss: 1.3155... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17587... Training loss: 1.2864... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17588... Training loss: 1.3300... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17589... Training loss: 1.3272... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17590... Training loss: 1.3614... 0.0850 sec/batch
Epoch: 18/20... Training Step: 17591... Training loss: 1.2636... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17592... Training loss: 1.3445... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17593... Training loss: 1.3213... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17594... Training loss: 1.2450... 0.0782 sec/batch
Epoch: 18/20... Training Step: 17595... Training loss: 1.2500... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17596... Training loss: 1.3180... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17597... Training loss: 1.2780... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17598... Training loss: 1.3333... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17599... Training loss: 1.2947... 0.0792 sec/batch
Epoch: 18/20... Training Step: 17600... Training loss: 1.2658... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17601... Training loss: 1.2805... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17602... Training loss: 1.3638... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17603... Training loss: 1.2634... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17604... Training loss: 1.2907... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17605... Training loss: 1.2348... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17606... Training loss: 1.2674... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17607... Training loss: 1.2876... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17608... Training loss: 1.3393... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17609... Training loss: 1.2994... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17610... Training loss: 1.2838... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17611... Training loss: 1.2863... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17612... Training loss: 1.2945... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17613... Training loss: 1.2203... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17614... Training loss: 1.2506... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17615... Training loss: 1.3803... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17616... Training loss: 1.2604... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17617... Training loss: 1.3178... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17618... Training loss: 1.3284... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17619... Training loss: 1.2196... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17620... Training loss: 1.2870... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17621... Training loss: 1.2913... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17622... Training loss: 1.3393... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17623... Training loss: 1.3081... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17624... Training loss: 1.3461... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17625... Training loss: 1.2839... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17626... Training loss: 1.2157... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17627... Training loss: 1.2980... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17628... Training loss: 1.2836... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17629... Training loss: 1.3277... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17630... Training loss: 1.3301... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17631... Training loss: 1.2719... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17632... Training loss: 1.2585... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17633... Training loss: 1.2661... 0.0785 sec/batch
Epoch: 18/20... Training Step: 17634... Training loss: 1.2909... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17635... Training loss: 1.3024... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17636... Training loss: 1.3447... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17637... Training loss: 1.3187... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17638... Training loss: 1.3281... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17639... Training loss: 1.2853... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17640... Training loss: 1.1806... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17641... Training loss: 1.2191... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17642... Training loss: 1.2907... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17643... Training loss: 1.2519... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17644... Training loss: 1.2741... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17645... Training loss: 1.2907... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17646... Training loss: 1.2812... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17647... Training loss: 1.3015... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17648... Training loss: 1.3050... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17649... Training loss: 1.2976... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17650... Training loss: 1.2881... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17651... Training loss: 1.2861... 0.0792 sec/batch
Epoch: 18/20... Training Step: 17652... Training loss: 1.2951... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17653... Training loss: 1.2904... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17654... Training loss: 1.3112... 0.0740 sec/batch
Epoch: 18/20... Training Step: 17655... Training loss: 1.2892... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17656... Training loss: 1.2835... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17657... Training loss: 1.2999... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17658... Training loss: 1.2471... 0.0849 sec/batch
Epoch: 18/20... Training Step: 17659... Training loss: 1.1890... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17660... Training loss: 1.2693... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17661... Training loss: 1.2518... 0.0840 sec/batch
Epoch: 18/20... Training Step: 17662... Training loss: 1.2777... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17663... Training loss: 1.2843... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17664... Training loss: 1.2687... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17665... Training loss: 1.3419... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17666... Training loss: 1.3404... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17667... Training loss: 1.3518... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17668... Training loss: 1.2868... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17669... Training loss: 1.2329... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17670... Training loss: 1.2926... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17671... Training loss: 1.2573... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17672... Training loss: 1.3159... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17673... Training loss: 1.3300... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17674... Training loss: 1.3501... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17675... Training loss: 1.2826... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17676... Training loss: 1.3415... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17677... Training loss: 1.2858... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17678... Training loss: 1.3309... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17679... Training loss: 1.3402... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17680... Training loss: 1.3202... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17681... Training loss: 1.3419... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17682... Training loss: 1.3832... 0.0835 sec/batch
Epoch: 18/20... Training Step: 17683... Training loss: 1.2981... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17684... Training loss: 1.2944... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17685... Training loss: 1.2660... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17686... Training loss: 1.2205... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17687... Training loss: 1.2891... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17688... Training loss: 1.2821... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17689... Training loss: 1.2878... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17690... Training loss: 1.3199... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17691... Training loss: 1.3288... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17692... Training loss: 1.3590... 0.0778 sec/batch
Epoch: 18/20... Training Step: 17693... Training loss: 1.3355... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17694... Training loss: 1.2905... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17695... Training loss: 1.3285... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17696... Training loss: 1.3208... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17697... Training loss: 1.3372... 0.0792 sec/batch
Epoch: 18/20... Training Step: 17698... Training loss: 1.3274... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17699... Training loss: 1.2295... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17700... Training loss: 1.2510... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17701... Training loss: 1.3059... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17702... Training loss: 1.3095... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17703... Training loss: 1.3002... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17704... Training loss: 1.2649... 0.0784 sec/batch
Epoch: 18/20... Training Step: 17705... Training loss: 1.3064... 0.0742 sec/batch
Epoch: 18/20... Training Step: 17706... Training loss: 1.4116... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17707... Training loss: 1.3277... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17708... Training loss: 1.3297... 0.0735 sec/batch
Epoch: 18/20... Training Step: 17709... Training loss: 1.3297... 0.0786 sec/batch
Epoch: 18/20... Training Step: 17710... Training loss: 1.2388... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17711... Training loss: 1.3130... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17712... Training loss: 1.3038... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17713... Training loss: 1.2909... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17714... Training loss: 1.3264... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17715... Training loss: 1.2761... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17716... Training loss: 1.3730... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17717... Training loss: 1.3798... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17718... Training loss: 1.2965... 0.0784 sec/batch
Epoch: 18/20... Training Step: 17719... Training loss: 1.3055... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17720... Training loss: 1.2902... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17721... Training loss: 1.2949... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17722... Training loss: 1.3009... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17723... Training loss: 1.3110... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17724... Training loss: 1.3198... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17725... Training loss: 1.3009... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17726... Training loss: 1.2510... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17727... Training loss: 1.2342... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17728... Training loss: 1.2837... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17729... Training loss: 1.2769... 0.0803 sec/batch
Epoch: 18/20... Training Step: 17730... Training loss: 1.2982... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17731... Training loss: 1.2995... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17732... Training loss: 1.3225... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17733... Training loss: 1.3031... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17734... Training loss: 1.3618... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17735... Training loss: 1.3329... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17736... Training loss: 1.3294... 0.0774 sec/batch
Epoch: 18/20... Training Step: 17737... Training loss: 1.2999... 0.0752 sec/batch
Epoch: 18/20... Training Step: 17738... Training loss: 1.2923... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17739... Training loss: 1.2755... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17740... Training loss: 1.3085... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17741... Training loss: 1.3000... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17742... Training loss: 1.3369... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17743... Training loss: 1.3037... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17744... Training loss: 1.2914... 0.0748 sec/batch
Epoch: 18/20... Training Step: 17745... Training loss: 1.3083... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17746... Training loss: 1.2374... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17747... Training loss: 1.3261... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17748... Training loss: 1.2891... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17749... Training loss: 1.2892... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17750... Training loss: 1.3645... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17751... Training loss: 1.2863... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17752... Training loss: 1.3279... 0.0740 sec/batch
Epoch: 18/20... Training Step: 17753... Training loss: 1.3273... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17754... Training loss: 1.3492... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17755... Training loss: 1.3211... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17756... Training loss: 1.2369... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17757... Training loss: 1.2835... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17758... Training loss: 1.2315... 0.0763 sec/batch
Epoch: 18/20... Training Step: 17759... Training loss: 1.3201... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17760... Training loss: 1.2416... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17761... Training loss: 1.3302... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17762... Training loss: 1.3134... 0.0747 sec/batch
Epoch: 18/20... Training Step: 17763... Training loss: 1.2545... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17764... Training loss: 1.2288... 0.0788 sec/batch
Epoch: 18/20... Training Step: 17765... Training loss: 1.2356... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17766... Training loss: 1.2761... 0.0844 sec/batch
Epoch: 18/20... Training Step: 17767... Training loss: 1.2593... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17768... Training loss: 1.2524... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17769... Training loss: 1.2004... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17770... Training loss: 1.3122... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17771... Training loss: 1.2457... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17772... Training loss: 1.2741... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17773... Training loss: 1.2154... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17774... Training loss: 1.1804... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17775... Training loss: 1.2363... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17776... Training loss: 1.2897... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17777... Training loss: 1.3635... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17778... Training loss: 1.2838... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17779... Training loss: 1.2691... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17780... Training loss: 1.2843... 0.0780 sec/batch
Epoch: 18/20... Training Step: 17781... Training loss: 1.2661... 0.0782 sec/batch
Epoch: 18/20... Training Step: 17782... Training loss: 1.3225... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17783... Training loss: 1.2889... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17784... Training loss: 1.3329... 0.0777 sec/batch
Epoch: 18/20... Training Step: 17785... Training loss: 1.4006... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17786... Training loss: 1.3329... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17787... Training loss: 1.2885... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17788... Training loss: 1.2612... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17789... Training loss: 1.3146... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17790... Training loss: 1.3257... 0.0750 sec/batch
Epoch: 18/20... Training Step: 17791... Training loss: 1.3015... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17792... Training loss: 1.2807... 0.0757 sec/batch
Epoch: 18/20... Training Step: 17793... Training loss: 1.3847... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17794... Training loss: 1.3559... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17795... Training loss: 1.3023... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17796... Training loss: 1.3935... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17797... Training loss: 1.2774... 0.0756 sec/batch
Epoch: 18/20... Training Step: 17798... Training loss: 1.2964... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17799... Training loss: 1.3150... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17800... Training loss: 1.3478... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17801... Training loss: 1.2770... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17802... Training loss: 1.2941... 0.0743 sec/batch
Epoch: 18/20... Training Step: 17803... Training loss: 1.2906... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17804... Training loss: 1.3714... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17805... Training loss: 1.3026... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17806... Training loss: 1.3172... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17807... Training loss: 1.3544... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17808... Training loss: 1.3087... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17809... Training loss: 1.3429... 0.0740 sec/batch
Epoch: 18/20... Training Step: 17810... Training loss: 1.2890... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17811... Training loss: 1.2873... 0.0744 sec/batch
Epoch: 18/20... Training Step: 17812... Training loss: 1.2630... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17813... Training loss: 1.2976... 0.0771 sec/batch
Epoch: 18/20... Training Step: 17814... Training loss: 1.2318... 0.0789 sec/batch
Epoch: 18/20... Training Step: 17815... Training loss: 1.3008... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17816... Training loss: 1.2468... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17817... Training loss: 1.2843... 0.0740 sec/batch
Epoch: 18/20... Training Step: 17818... Training loss: 1.3172... 0.0737 sec/batch
Epoch: 18/20... Training Step: 17819... Training loss: 1.3144... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17820... Training loss: 1.3269... 0.0749 sec/batch
Epoch: 18/20... Training Step: 17821... Training loss: 1.2886... 0.0762 sec/batch
Epoch: 18/20... Training Step: 17822... Training loss: 1.2968... 0.0761 sec/batch
Epoch: 18/20... Training Step: 17823... Training loss: 1.2557... 0.0765 sec/batch
Epoch: 18/20... Training Step: 17824... Training loss: 1.2446... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17825... Training loss: 1.2873... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17826... Training loss: 1.2800... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17827... Training loss: 1.2906... 0.0767 sec/batch
Epoch: 18/20... Training Step: 17828... Training loss: 1.2579... 0.0746 sec/batch
Epoch: 18/20... Training Step: 17829... Training loss: 1.3424... 0.0760 sec/batch
Epoch: 18/20... Training Step: 17830... Training loss: 1.3430... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17831... Training loss: 1.3504... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17832... Training loss: 1.3428... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17833... Training loss: 1.3331... 0.0745 sec/batch
Epoch: 18/20... Training Step: 17834... Training loss: 1.4671... 0.0764 sec/batch
Epoch: 18/20... Training Step: 17835... Training loss: 1.2706... 0.0768 sec/batch
Epoch: 18/20... Training Step: 17836... Training loss: 1.3675... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17837... Training loss: 1.4097... 0.0751 sec/batch
Epoch: 18/20... Training Step: 17838... Training loss: 1.2617... 0.0759 sec/batch
Epoch: 18/20... Training Step: 17839... Training loss: 1.3318... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17840... Training loss: 1.3395... 0.0769 sec/batch
Epoch: 18/20... Training Step: 17841... Training loss: 1.3443... 0.0741 sec/batch
Epoch: 18/20... Training Step: 17842... Training loss: 1.3539... 0.0776 sec/batch
Epoch: 18/20... Training Step: 17843... Training loss: 1.3254... 0.0766 sec/batch
Epoch: 18/20... Training Step: 17844... Training loss: 1.2495... 0.0754 sec/batch
Epoch: 18/20... Training Step: 17845... Training loss: 1.2351... 0.0753 sec/batch
Epoch: 18/20... Training Step: 17846... Training loss: 1.4003... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17847... Training loss: 1.2946... 0.0773 sec/batch
Epoch: 18/20... Training Step: 17848... Training loss: 1.2990... 0.0783 sec/batch
Epoch: 18/20... Training Step: 17849... Training loss: 1.2999... 0.0770 sec/batch
Epoch: 18/20... Training Step: 17850... Training loss: 1.2711... 0.0779 sec/batch
Epoch: 18/20... Training Step: 17851... Training loss: 1.2597... 0.0758 sec/batch
Epoch: 18/20... Training Step: 17852... Training loss: 1.2626... 0.0772 sec/batch
Epoch: 18/20... Training Step: 17853... Training loss: 1.3107... 0.0775 sec/batch
Epoch: 18/20... Training Step: 17854... Training loss: 1.2054... 0.0755 sec/batch
Epoch: 18/20... Training Step: 17855... Training loss: 1.2862... 0.0781 sec/batch
Epoch: 18/20... Training Step: 17856... Training loss: 1.3406... 0.0747 sec/batch
Epoch: 19/20... Training Step: 17857... Training loss: 1.3731... 0.0773 sec/batch
Epoch: 19/20... Training Step: 17858... Training loss: 1.2492... 0.0783 sec/batch
Epoch: 19/20... Training Step: 17859... Training loss: 1.3526... 0.0746 sec/batch
Epoch: 19/20... Training Step: 17860... Training loss: 1.3243... 0.0756 sec/batch
Epoch: 19/20... Training Step: 17861... Training loss: 1.2322... 0.0759 sec/batch
Epoch: 19/20... Training Step: 17862... Training loss: 1.1656... 0.0766 sec/batch
Epoch: 19/20... Training Step: 17863... Training loss: 1.2644... 0.0745 sec/batch
Epoch: 19/20... Training Step: 17864... Training loss: 1.2833... 0.0828 sec/batch
Epoch: 19/20... Training Step: 17865... Training loss: 1.3093... 0.0742 sec/batch
Epoch: 19/20... Training Step: 17866... Training loss: 1.2265... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17867... Training loss: 1.3587... 0.0785 sec/batch
Epoch: 19/20... Training Step: 17868... Training loss: 1.2580... 0.0775 sec/batch
Epoch: 19/20... Training Step: 17869... Training loss: 1.2815... 0.0759 sec/batch
Epoch: 19/20... Training Step: 17870... Training loss: 1.2890... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17871... Training loss: 1.3430... 0.0773 sec/batch
Epoch: 19/20... Training Step: 17872... Training loss: 1.2508... 0.0759 sec/batch
Epoch: 19/20... Training Step: 17873... Training loss: 1.2695... 0.0767 sec/batch
Epoch: 19/20... Training Step: 17874... Training loss: 1.3084... 0.0757 sec/batch
Epoch: 19/20... Training Step: 17875... Training loss: 1.2550... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17876... Training loss: 1.2502... 0.0792 sec/batch
Epoch: 19/20... Training Step: 17877... Training loss: 1.2971... 0.0776 sec/batch
Epoch: 19/20... Training Step: 17878... Training loss: 1.2603... 0.0790 sec/batch
Epoch: 19/20... Training Step: 17879... Training loss: 1.2672... 0.0762 sec/batch
Epoch: 19/20... Training Step: 17880... Training loss: 1.2081... 0.0757 sec/batch
Epoch: 19/20... Training Step: 17881... Training loss: 1.3125... 0.0772 sec/batch
Epoch: 19/20... Training Step: 17882... Training loss: 1.2353... 0.0766 sec/batch
Epoch: 19/20... Training Step: 17883... Training loss: 1.2809... 0.0757 sec/batch
Epoch: 19/20... Training Step: 17884... Training loss: 1.2403... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17885... Training loss: 1.2752... 0.0747 sec/batch
Epoch: 19/20... Training Step: 17886... Training loss: 1.3214... 0.0765 sec/batch
Epoch: 19/20... Training Step: 17887... Training loss: 1.3299... 0.0759 sec/batch
Epoch: 19/20... Training Step: 17888... Training loss: 1.3076... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17889... Training loss: 1.2762... 0.0760 sec/batch
Epoch: 19/20... Training Step: 17890... Training loss: 1.3363... 0.0762 sec/batch
Epoch: 19/20... Training Step: 17891... Training loss: 1.3816... 0.0783 sec/batch
Epoch: 19/20... Training Step: 17892... Training loss: 1.3134... 0.0760 sec/batch
Epoch: 19/20... Training Step: 17893... Training loss: 1.2806... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17894... Training loss: 1.3376... 0.0768 sec/batch
Epoch: 19/20... Training Step: 17895... Training loss: 1.3260... 0.0758 sec/batch
Epoch: 19/20... Training Step: 17896... Training loss: 1.2715... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17897... Training loss: 1.2949... 0.0747 sec/batch
Epoch: 19/20... Training Step: 17898... Training loss: 1.2561... 0.0757 sec/batch
Epoch: 19/20... Training Step: 17899... Training loss: 1.2311... 0.0775 sec/batch
Epoch: 19/20... Training Step: 17900... Training loss: 1.2496... 0.0770 sec/batch
Epoch: 19/20... Training Step: 17901... Training loss: 1.3267... 0.0769 sec/batch
Epoch: 19/20... Training Step: 17902... Training loss: 1.2555... 0.0779 sec/batch
Epoch: 19/20... Training Step: 17903... Training loss: 1.3162... 0.0773 sec/batch
Epoch: 19/20... Training Step: 17904... Training loss: 1.2603... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17905... Training loss: 1.3293... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17906... Training loss: 1.3231... 0.0771 sec/batch
Epoch: 19/20... Training Step: 17907... Training loss: 1.2385... 0.0744 sec/batch
Epoch: 19/20... Training Step: 17908... Training loss: 1.3221... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17909... Training loss: 1.2354... 0.0779 sec/batch
Epoch: 19/20... Training Step: 17910... Training loss: 1.2282... 0.0776 sec/batch
Epoch: 19/20... Training Step: 17911... Training loss: 1.2511... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17912... Training loss: 1.2484... 0.0767 sec/batch
Epoch: 19/20... Training Step: 17913... Training loss: 1.3066... 0.0751 sec/batch
Epoch: 19/20... Training Step: 17914... Training loss: 1.3352... 0.0759 sec/batch
Epoch: 19/20... Training Step: 17915... Training loss: 1.2872... 0.0755 sec/batch
Epoch: 19/20... Training Step: 17916... Training loss: 1.3221... 0.0753 sec/batch
Epoch: 19/20... Training Step: 17917... Training loss: 1.2018... 0.0776 sec/batch
Epoch: 19/20... Training Step: 17918... Training loss: 1.2272... 0.0784 sec/batch
Epoch: 19/20... Training Step: 17919... Training loss: 1.2827... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17920... Training loss: 1.2528... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17921... Training loss: 1.3172... 0.0753 sec/batch
Epoch: 19/20... Training Step: 17922... Training loss: 1.2835... 0.0752 sec/batch
Epoch: 19/20... Training Step: 17923... Training loss: 1.2521... 0.0765 sec/batch
Epoch: 19/20... Training Step: 17924... Training loss: 1.2789... 0.0758 sec/batch
Epoch: 19/20... Training Step: 17925... Training loss: 1.3231... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17926... Training loss: 1.2801... 0.0756 sec/batch
Epoch: 19/20... Training Step: 17927... Training loss: 1.2938... 0.0766 sec/batch
Epoch: 19/20... Training Step: 17928... Training loss: 1.3364... 0.0778 sec/batch
Epoch: 19/20... Training Step: 17929... Training loss: 1.3028... 0.0768 sec/batch
Epoch: 19/20... Training Step: 17930... Training loss: 1.2285... 0.0768 sec/batch
Epoch: 19/20... Training Step: 17931... Training loss: 1.2707... 0.0781 sec/batch
Epoch: 19/20... Training Step: 17932... Training loss: 1.2352... 0.0786 sec/batch
Epoch: 19/20... Training Step: 17933... Training loss: 1.2589... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17934... Training loss: 1.2804... 0.0750 sec/batch
Epoch: 19/20... Training Step: 17935... Training loss: 1.3002... 0.0750 sec/batch
Epoch: 19/20... Training Step: 17936... Training loss: 1.3507... 0.0755 sec/batch
Epoch: 19/20... Training Step: 17937... Training loss: 1.2838... 0.0764 sec/batch
Epoch: 19/20... Training Step: 17938... Training loss: 1.2413... 0.0775 sec/batch
Epoch: 19/20... Training Step: 17939... Training loss: 1.3057... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17940... Training loss: 1.3003... 0.0771 sec/batch
Epoch: 19/20... Training Step: 17941... Training loss: 1.2677... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17942... Training loss: 1.3080... 0.0776 sec/batch
Epoch: 19/20... Training Step: 17943... Training loss: 1.2858... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17944... Training loss: 1.2869... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17945... Training loss: 1.3264... 0.0768 sec/batch
Epoch: 19/20... Training Step: 17946... Training loss: 1.3254... 0.0769 sec/batch
Epoch: 19/20... Training Step: 17947... Training loss: 1.3180... 0.0761 sec/batch
Epoch: 19/20... Training Step: 17948... Training loss: 1.2763... 0.0775 sec/batch
Epoch: 19/20... Training Step: 17949... Training loss: 1.2278... 0.0758 sec/batch
Epoch: 19/20... Training Step: 17950... Training loss: 1.2206... 0.0778 sec/batch
Epoch: 19/20... Training Step: 17951... Training loss: 1.1988... 0.0762 sec/batch
Epoch: 19/20... Training Step: 17952... Training loss: 1.2794... 0.0770 sec/batch
Epoch: 19/20... Training Step: 17953... Training loss: 1.2415... 0.0764 sec/batch
Epoch: 19/20... Training Step: 17954... Training loss: 1.2299... 0.0814 sec/batch
Epoch: 19/20... Training Step: 17955... Training loss: 1.2354... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17956... Training loss: 1.2552... 0.0760 sec/batch
Epoch: 19/20... Training Step: 17957... Training loss: 1.2447... 0.0762 sec/batch
Epoch: 19/20... Training Step: 17958... Training loss: 1.2569... 0.0747 sec/batch
Epoch: 19/20... Training Step: 17959... Training loss: 1.3006... 0.0770 sec/batch
Epoch: 19/20... Training Step: 17960... Training loss: 1.2941... 0.0776 sec/batch
Epoch: 19/20... Training Step: 17961... Training loss: 1.2969... 0.0755 sec/batch
Epoch: 19/20... Training Step: 17962... Training loss: 1.2581... 0.0772 sec/batch
Epoch: 19/20... Training Step: 17963... Training loss: 1.2262... 0.0773 sec/batch
Epoch: 19/20... Training Step: 17964... Training loss: 1.3105... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17965... Training loss: 1.2867... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17966... Training loss: 1.3073... 0.0765 sec/batch
Epoch: 19/20... Training Step: 17967... Training loss: 1.2899... 0.0733 sec/batch
Epoch: 19/20... Training Step: 17968... Training loss: 1.2717... 0.0785 sec/batch
Epoch: 19/20... Training Step: 17969... Training loss: 1.3348... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17970... Training loss: 1.3068... 0.0764 sec/batch
Epoch: 19/20... Training Step: 17971... Training loss: 1.2291... 0.0764 sec/batch
Epoch: 19/20... Training Step: 17972... Training loss: 1.3812... 0.0755 sec/batch
Epoch: 19/20... Training Step: 17973... Training loss: 1.3141... 0.0749 sec/batch
Epoch: 19/20... Training Step: 17974... Training loss: 1.3739... 0.0780 sec/batch
Epoch: 19/20... Training Step: 17975... Training loss: 1.2889... 0.0773 sec/batch
Epoch: 19/20... Training Step: 17976... Training loss: 1.2990... 0.0754 sec/batch
Epoch: 19/20... Training Step: 17977... Training loss: 1.3882... 0.0771 sec/batch
Epoch: 19/20... Training Step: 17978... Training loss: 1.2560... 0.0747 sec/batch
Epoch: 19/20... Training Step: 17979... Training loss: 1.3077... 0.0755 sec/batch
Epoch: 19/20... Training Step: 17980... Training loss: 1.2775... 0.0801 sec/batch
Epoch: 19/20... Training Step: 17981... Training loss: 1.3350... 0.0774 sec/batch
Epoch: 19/20... Training Step: 17982... Training loss: 1.3477... 0.0799 sec/batch
Epoch: 19/20... Training Step: 17983... Training loss: 1.3579... 0.0758 sec/batch
Epoch: 19/20... Training Step: 17984... Training loss: 1.3003... 0.0752 sec/batch
Epoch: 19/20... Training Step: 17985... Training loss: 1.2356... 0.0751 sec/batch
Epoch: 19/20... Training Step: 17986... Training loss: 1.2985... 0.0763 sec/batch
Epoch: 19/20... Training Step: 17987... Training loss: 1.2457... 0.0777 sec/batch
Epoch: 19/20... Training Step: 17988... Training loss: 1.3422... 0.0756 sec/batch
Epoch: 19/20... Training Step: 17989... Training loss: 1.3118... 0.0771 sec/batch
Epoch: 19/20... Training Step: 17990... Training loss: 1.2690... 0.0768 sec/batch
Epoch: 19/20... Training Step: 17991... Training loss: 1.3240... 0.0767 sec/batch
Epoch: 19/20... Training Step: 17992... Training loss: 1.3265... 0.0748 sec/batch
Epoch: 19/20... Training Step: 17993... Training loss: 1.3345... 0.0758 sec/batch
Epoch: 19/20... Training Step: 17994... Training loss: 1.3705... 0.0748 sec/batch
Epoch: 19/20... Training Step: 17995... Training loss: 1.3286... 0.0766 sec/batch
Epoch: 19/20... Training Step: 17996... Training loss: 1.2896... 0.0751 sec/batch
Epoch: 19/20... Training Step: 17997... Training loss: 1.3138... 0.0748 sec/batch
Epoch: 19/20... Training Step: 17998... Training loss: 1.2582... 0.0745 sec/batch
Epoch: 19/20... Training Step: 17999... Training loss: 1.2868... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18000... Training loss: 1.2793... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18001... Training loss: 1.1604... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18002... Training loss: 1.3298... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18003... Training loss: 1.2884... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18004... Training loss: 1.2339... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18005... Training loss: 1.2455... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18006... Training loss: 1.2423... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18007... Training loss: 1.1596... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18008... Training loss: 1.2876... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18009... Training loss: 1.2032... 0.0806 sec/batch
Epoch: 19/20... Training Step: 18010... Training loss: 1.2664... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18011... Training loss: 1.2446... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18012... Training loss: 1.2425... 0.0789 sec/batch
Epoch: 19/20... Training Step: 18013... Training loss: 1.2662... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18014... Training loss: 1.2760... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18015... Training loss: 1.2774... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18016... Training loss: 1.3114... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18017... Training loss: 1.2549... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18018... Training loss: 1.2814... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18019... Training loss: 1.2561... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18020... Training loss: 1.3387... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18021... Training loss: 1.2908... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18022... Training loss: 1.3726... 0.0786 sec/batch
Epoch: 19/20... Training Step: 18023... Training loss: 1.3200... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18024... Training loss: 1.2773... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18025... Training loss: 1.2990... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18026... Training loss: 1.2848... 0.0737 sec/batch
Epoch: 19/20... Training Step: 18027... Training loss: 1.3022... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18028... Training loss: 1.2732... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18029... Training loss: 1.3047... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18030... Training loss: 1.3375... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18031... Training loss: 1.2467... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18032... Training loss: 1.3114... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18033... Training loss: 1.2967... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18034... Training loss: 1.3914... 0.0814 sec/batch
Epoch: 19/20... Training Step: 18035... Training loss: 1.3124... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18036... Training loss: 1.2494... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18037... Training loss: 1.2906... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18038... Training loss: 1.3558... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18039... Training loss: 1.2822... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18040... Training loss: 1.3043... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18041... Training loss: 1.3111... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18042... Training loss: 1.3182... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18043... Training loss: 1.2977... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18044... Training loss: 1.2737... 0.0738 sec/batch
Epoch: 19/20... Training Step: 18045... Training loss: 1.2918... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18046... Training loss: 1.3487... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18047... Training loss: 1.3522... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18048... Training loss: 1.3156... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18049... Training loss: 1.3372... 0.0790 sec/batch
Epoch: 19/20... Training Step: 18050... Training loss: 1.3631... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18051... Training loss: 1.2945... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18052... Training loss: 1.2537... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18053... Training loss: 1.3335... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18054... Training loss: 1.3091... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18055... Training loss: 1.3228... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18056... Training loss: 1.2710... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18057... Training loss: 1.2843... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18058... Training loss: 1.3072... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18059... Training loss: 1.3042... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18060... Training loss: 1.2862... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18061... Training loss: 1.2440... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18062... Training loss: 1.3463... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18063... Training loss: 1.2859... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18064... Training loss: 1.2724... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18065... Training loss: 1.3283... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18066... Training loss: 1.2879... 0.0743 sec/batch
Epoch: 19/20... Training Step: 18067... Training loss: 1.2332... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18068... Training loss: 1.3226... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18069... Training loss: 1.2679... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18070... Training loss: 1.3292... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18071... Training loss: 1.2684... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18072... Training loss: 1.2714... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18073... Training loss: 1.2799... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18074... Training loss: 1.2968... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18075... Training loss: 1.2880... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18076... Training loss: 1.2265... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18077... Training loss: 1.2401... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18078... Training loss: 1.4110... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18079... Training loss: 1.3025... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18080... Training loss: 1.2946... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18081... Training loss: 1.3353... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18082... Training loss: 1.2632... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18083... Training loss: 1.3248... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18084... Training loss: 1.3393... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18085... Training loss: 1.2151... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18086... Training loss: 1.2158... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18087... Training loss: 1.2728... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18088... Training loss: 1.2998... 0.0797 sec/batch
Epoch: 19/20... Training Step: 18089... Training loss: 1.2835... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18090... Training loss: 1.3658... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18091... Training loss: 1.2851... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18092... Training loss: 1.3070... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18093... Training loss: 1.2712... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18094... Training loss: 1.3122... 0.0789 sec/batch
Epoch: 19/20... Training Step: 18095... Training loss: 1.3214... 0.0790 sec/batch
Epoch: 19/20... Training Step: 18096... Training loss: 1.2538... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18097... Training loss: 1.2787... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18098... Training loss: 1.2864... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18099... Training loss: 1.3346... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18100... Training loss: 1.3272... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18101... Training loss: 1.2820... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18102... Training loss: 1.3230... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18103... Training loss: 1.2426... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18104... Training loss: 1.1913... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18105... Training loss: 1.2679... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18106... Training loss: 1.3163... 0.0827 sec/batch
Epoch: 19/20... Training Step: 18107... Training loss: 1.2599... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18108... Training loss: 1.2609... 0.0797 sec/batch
Epoch: 19/20... Training Step: 18109... Training loss: 1.2562... 0.0743 sec/batch
Epoch: 19/20... Training Step: 18110... Training loss: 1.2549... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18111... Training loss: 1.2267... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18112... Training loss: 1.2512... 0.0797 sec/batch
Epoch: 19/20... Training Step: 18113... Training loss: 1.3504... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18114... Training loss: 1.2628... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18115... Training loss: 1.2805... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18116... Training loss: 1.2965... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18117... Training loss: 1.3224... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18118... Training loss: 1.3238... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18119... Training loss: 1.3193... 0.0738 sec/batch
Epoch: 19/20... Training Step: 18120... Training loss: 1.3164... 0.0787 sec/batch
Epoch: 19/20... Training Step: 18121... Training loss: 1.2948... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18122... Training loss: 1.1924... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18123... Training loss: 1.2479... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18124... Training loss: 1.2684... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18125... Training loss: 1.2238... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18126... Training loss: 1.2504... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18127... Training loss: 1.2817... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18128... Training loss: 1.2648... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18129... Training loss: 1.1833... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18130... Training loss: 1.2216... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18131... Training loss: 1.2665... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18132... Training loss: 1.2947... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18133... Training loss: 1.2789... 0.0820 sec/batch
Epoch: 19/20... Training Step: 18134... Training loss: 1.1773... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18135... Training loss: 1.2382... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18136... Training loss: 1.3385... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18137... Training loss: 1.2598... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18138... Training loss: 1.2526... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18139... Training loss: 1.3075... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18140... Training loss: 1.2433... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18141... Training loss: 1.1636... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18142... Training loss: 1.2381... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18143... Training loss: 1.2829... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18144... Training loss: 1.2736... 0.0781 sec/batch
Epoch: 19/20... Training Step: 18145... Training loss: 1.2629... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18146... Training loss: 1.2507... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18147... Training loss: 1.2512... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18148... Training loss: 1.2896... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18149... Training loss: 1.2338... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18150... Training loss: 1.2305... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18151... Training loss: 1.2761... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18152... Training loss: 1.2931... 0.0743 sec/batch
Epoch: 19/20... Training Step: 18153... Training loss: 1.2523... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18154... Training loss: 1.3466... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18155... Training loss: 1.2617... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18156... Training loss: 1.3109... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18157... Training loss: 1.3541... 0.0794 sec/batch
Epoch: 19/20... Training Step: 18158... Training loss: 1.2250... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18159... Training loss: 1.2541... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18160... Training loss: 1.3115... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18161... Training loss: 1.2783... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18162... Training loss: 1.3203... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18163... Training loss: 1.2446... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18164... Training loss: 1.2252... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18165... Training loss: 1.3576... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18166... Training loss: 1.2562... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18167... Training loss: 1.3523... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18168... Training loss: 1.2837... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18169... Training loss: 1.2508... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18170... Training loss: 1.2863... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18171... Training loss: 1.2846... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18172... Training loss: 1.2541... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18173... Training loss: 1.2449... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18174... Training loss: 1.2116... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18175... Training loss: 1.2742... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18176... Training loss: 1.2927... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18177... Training loss: 1.2991... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18178... Training loss: 1.2299... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18179... Training loss: 1.2536... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18180... Training loss: 1.2549... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18181... Training loss: 1.3022... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18182... Training loss: 1.2701... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18183... Training loss: 1.2791... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18184... Training loss: 1.2791... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18185... Training loss: 1.2726... 0.0781 sec/batch
Epoch: 19/20... Training Step: 18186... Training loss: 1.2945... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18187... Training loss: 1.3205... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18188... Training loss: 1.2198... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18189... Training loss: 1.3546... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18190... Training loss: 1.3593... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18191... Training loss: 1.3534... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18192... Training loss: 1.3034... 0.0845 sec/batch
Epoch: 19/20... Training Step: 18193... Training loss: 1.2619... 0.0850 sec/batch
Epoch: 19/20... Training Step: 18194... Training loss: 1.2734... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18195... Training loss: 1.2709... 0.0825 sec/batch
Epoch: 19/20... Training Step: 18196... Training loss: 1.2703... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18197... Training loss: 1.2634... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18198... Training loss: 1.2394... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18199... Training loss: 1.2372... 0.0786 sec/batch
Epoch: 19/20... Training Step: 18200... Training loss: 1.2800... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18201... Training loss: 1.2585... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18202... Training loss: 1.3144... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18203... Training loss: 1.3053... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18204... Training loss: 1.3228... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18205... Training loss: 1.3198... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18206... Training loss: 1.3264... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18207... Training loss: 1.3179... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18208... Training loss: 1.2664... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18209... Training loss: 1.2557... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18210... Training loss: 1.2695... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18211... Training loss: 1.2814... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18212... Training loss: 1.2950... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18213... Training loss: 1.2465... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18214... Training loss: 1.2820... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18215... Training loss: 1.2951... 0.0803 sec/batch
Epoch: 19/20... Training Step: 18216... Training loss: 1.2881... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18217... Training loss: 1.3182... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18218... Training loss: 1.3440... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18219... Training loss: 1.2705... 0.0818 sec/batch
Epoch: 19/20... Training Step: 18220... Training loss: 1.2888... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18221... Training loss: 1.2962... 0.0807 sec/batch
Epoch: 19/20... Training Step: 18222... Training loss: 1.2443... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18223... Training loss: 1.2579... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18224... Training loss: 1.2204... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18225... Training loss: 1.2344... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18226... Training loss: 1.2347... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18227... Training loss: 1.3295... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18228... Training loss: 1.2390... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18229... Training loss: 1.2658... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18230... Training loss: 1.2186... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18231... Training loss: 1.2562... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18232... Training loss: 1.2389... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18233... Training loss: 1.2662... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18234... Training loss: 1.2909... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18235... Training loss: 1.2478... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18236... Training loss: 1.3032... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18237... Training loss: 1.2837... 0.0835 sec/batch
Epoch: 19/20... Training Step: 18238... Training loss: 1.2614... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18239... Training loss: 1.3026... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18240... Training loss: 1.2931... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18241... Training loss: 1.3025... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18242... Training loss: 1.2939... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18243... Training loss: 1.3005... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18244... Training loss: 1.3346... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18245... Training loss: 1.3185... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18246... Training loss: 1.2318... 0.0813 sec/batch
Epoch: 19/20... Training Step: 18247... Training loss: 1.3139... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18248... Training loss: 1.2499... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18249... Training loss: 1.2839... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18250... Training loss: 1.2712... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18251... Training loss: 1.2875... 0.0739 sec/batch
Epoch: 19/20... Training Step: 18252... Training loss: 1.2539... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18253... Training loss: 1.2689... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18254... Training loss: 1.3204... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18255... Training loss: 1.2904... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18256... Training loss: 1.2627... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18257... Training loss: 1.2394... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18258... Training loss: 1.2967... 0.0860 sec/batch
Epoch: 19/20... Training Step: 18259... Training loss: 1.2508... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18260... Training loss: 1.3093... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18261... Training loss: 1.2506... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18262... Training loss: 1.2718... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18263... Training loss: 1.2110... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18264... Training loss: 1.3548... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18265... Training loss: 1.2136... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18266... Training loss: 1.2170... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18267... Training loss: 1.1974... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18268... Training loss: 1.2590... 0.0730 sec/batch
Epoch: 19/20... Training Step: 18269... Training loss: 1.2693... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18270... Training loss: 1.2092... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18271... Training loss: 1.2269... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18272... Training loss: 1.3017... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18273... Training loss: 1.2205... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18274... Training loss: 1.2229... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18275... Training loss: 1.3167... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18276... Training loss: 1.3532... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18277... Training loss: 1.2114... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18278... Training loss: 1.3030... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18279... Training loss: 1.2724... 0.0833 sec/batch
Epoch: 19/20... Training Step: 18280... Training loss: 1.2942... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18281... Training loss: 1.2549... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18282... Training loss: 1.2841... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18283... Training loss: 1.3715... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18284... Training loss: 1.3682... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18285... Training loss: 1.3396... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18286... Training loss: 1.3696... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18287... Training loss: 1.2887... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18288... Training loss: 1.3066... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18289... Training loss: 1.3139... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18290... Training loss: 1.3209... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18291... Training loss: 1.3187... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18292... Training loss: 1.2482... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18293... Training loss: 1.3230... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18294... Training loss: 1.3607... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18295... Training loss: 1.2970... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18296... Training loss: 1.3230... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18297... Training loss: 1.2747... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18298... Training loss: 1.3125... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18299... Training loss: 1.2928... 0.0796 sec/batch
Epoch: 19/20... Training Step: 18300... Training loss: 1.3204... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18301... Training loss: 1.2785... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18302... Training loss: 1.3486... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18303... Training loss: 1.2954... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18304... Training loss: 1.3104... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18305... Training loss: 1.3583... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18306... Training loss: 1.3721... 0.0787 sec/batch
Epoch: 19/20... Training Step: 18307... Training loss: 1.2791... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18308... Training loss: 1.2872... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18309... Training loss: 1.3388... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18310... Training loss: 1.2981... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18311... Training loss: 1.3193... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18312... Training loss: 1.2782... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18313... Training loss: 1.2828... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18314... Training loss: 1.2875... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18315... Training loss: 1.2761... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18316... Training loss: 1.3018... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18317... Training loss: 1.2937... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18318... Training loss: 1.2869... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18319... Training loss: 1.2358... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18320... Training loss: 1.3096... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18321... Training loss: 1.2645... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18322... Training loss: 1.2598... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18323... Training loss: 1.2136... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18324... Training loss: 1.2786... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18325... Training loss: 1.2745... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18326... Training loss: 1.3611... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18327... Training loss: 1.3112... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18328... Training loss: 1.2955... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18329... Training loss: 1.2839... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18330... Training loss: 1.2693... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18331... Training loss: 1.1929... 0.0787 sec/batch
Epoch: 19/20... Training Step: 18332... Training loss: 1.2517... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18333... Training loss: 1.2929... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18334... Training loss: 1.2825... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18335... Training loss: 1.2238... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18336... Training loss: 1.2430... 0.0818 sec/batch
Epoch: 19/20... Training Step: 18337... Training loss: 1.2618... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18338... Training loss: 1.2915... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18339... Training loss: 1.2574... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18340... Training loss: 1.2712... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18341... Training loss: 1.1915... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18342... Training loss: 1.2571... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18343... Training loss: 1.3060... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18344... Training loss: 1.2645... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18345... Training loss: 1.2276... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18346... Training loss: 1.2191... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18347... Training loss: 1.2469... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18348... Training loss: 1.2426... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18349... Training loss: 1.2549... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18350... Training loss: 1.2587... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18351... Training loss: 1.2301... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18352... Training loss: 1.2518... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18353... Training loss: 1.2590... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18354... Training loss: 1.2682... 0.0796 sec/batch
Epoch: 19/20... Training Step: 18355... Training loss: 1.2501... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18356... Training loss: 1.2871... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18357... Training loss: 1.2455... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18358... Training loss: 1.2742... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18359... Training loss: 1.3005... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18360... Training loss: 1.2489... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18361... Training loss: 1.2813... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18362... Training loss: 1.2425... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18363... Training loss: 1.2611... 0.0786 sec/batch
Epoch: 19/20... Training Step: 18364... Training loss: 1.2229... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18365... Training loss: 1.2349... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18366... Training loss: 1.2255... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18367... Training loss: 1.3088... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18368... Training loss: 1.2967... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18369... Training loss: 1.3007... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18370... Training loss: 1.3221... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18371... Training loss: 1.2744... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18372... Training loss: 1.2791... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18373... Training loss: 1.2958... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18374... Training loss: 1.2737... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18375... Training loss: 1.2613... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18376... Training loss: 1.2293... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18377... Training loss: 1.2858... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18378... Training loss: 1.2309... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18379... Training loss: 1.3572... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18380... Training loss: 1.2658... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18381... Training loss: 1.2520... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18382... Training loss: 1.2400... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18383... Training loss: 1.2569... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18384... Training loss: 1.2329... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18385... Training loss: 1.2896... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18386... Training loss: 1.3295... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18387... Training loss: 1.3773... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18388... Training loss: 1.3065... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18389... Training loss: 1.3327... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18390... Training loss: 1.3490... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18391... Training loss: 1.2886... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18392... Training loss: 1.2430... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18393... Training loss: 1.2368... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18394... Training loss: 1.2502... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18395... Training loss: 1.2507... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18396... Training loss: 1.2616... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18397... Training loss: 1.3598... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18398... Training loss: 1.2651... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18399... Training loss: 1.2551... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18400... Training loss: 1.2772... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18401... Training loss: 1.2018... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18402... Training loss: 1.2537... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18403... Training loss: 1.3260... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18404... Training loss: 1.2810... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18405... Training loss: 1.2489... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18406... Training loss: 1.2479... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18407... Training loss: 1.2649... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18408... Training loss: 1.2192... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18409... Training loss: 1.2177... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18410... Training loss: 1.2629... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18411... Training loss: 1.2807... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18412... Training loss: 1.2635... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18413... Training loss: 1.2312... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18414... Training loss: 1.2569... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18415... Training loss: 1.2870... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18416... Training loss: 1.1803... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18417... Training loss: 1.2458... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18418... Training loss: 1.2604... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18419... Training loss: 1.2470... 0.0798 sec/batch
Epoch: 19/20... Training Step: 18420... Training loss: 1.2578... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18421... Training loss: 1.2722... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18422... Training loss: 1.2906... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18423... Training loss: 1.2465... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18424... Training loss: 1.1965... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18425... Training loss: 1.1901... 0.0825 sec/batch
Epoch: 19/20... Training Step: 18426... Training loss: 1.2234... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18427... Training loss: 1.2005... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18428... Training loss: 1.2370... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18429... Training loss: 1.3039... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18430... Training loss: 1.3047... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18431... Training loss: 1.3050... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18432... Training loss: 1.2914... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18433... Training loss: 1.2502... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18434... Training loss: 1.3780... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18435... Training loss: 1.3130... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18436... Training loss: 1.2534... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18437... Training loss: 1.2057... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18438... Training loss: 1.2334... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18439... Training loss: 1.3167... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18440... Training loss: 1.2225... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18441... Training loss: 1.2556... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18442... Training loss: 1.2961... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18443... Training loss: 1.3085... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18444... Training loss: 1.2904... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18445... Training loss: 1.2808... 0.0801 sec/batch
Epoch: 19/20... Training Step: 18446... Training loss: 1.2450... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18447... Training loss: 1.2786... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18448... Training loss: 1.2941... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18449... Training loss: 1.2111... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18450... Training loss: 1.2473... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18451... Training loss: 1.1700... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18452... Training loss: 1.3040... 0.0801 sec/batch
Epoch: 19/20... Training Step: 18453... Training loss: 1.1607... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18454... Training loss: 1.1761... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18455... Training loss: 1.2570... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18456... Training loss: 1.2559... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18457... Training loss: 1.2986... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18458... Training loss: 1.2505... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18459... Training loss: 1.2411... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18460... Training loss: 1.2833... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18461... Training loss: 1.2067... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18462... Training loss: 1.2380... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18463... Training loss: 1.3216... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18464... Training loss: 1.2869... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18465... Training loss: 1.2663... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18466... Training loss: 1.3255... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18467... Training loss: 1.1898... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18468... Training loss: 1.3049... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18469... Training loss: 1.2781... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18470... Training loss: 1.2273... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18471... Training loss: 1.3231... 0.0789 sec/batch
Epoch: 19/20... Training Step: 18472... Training loss: 1.2712... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18473... Training loss: 1.2961... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18474... Training loss: 1.3050... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18475... Training loss: 1.2124... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18476... Training loss: 1.2330... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18477... Training loss: 1.2367... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18478... Training loss: 1.2090... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18479... Training loss: 1.3028... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18480... Training loss: 1.2876... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18481... Training loss: 1.2170... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18482... Training loss: 1.2431... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18483... Training loss: 1.2433... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18484... Training loss: 1.2700... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18485... Training loss: 1.1079... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18486... Training loss: 1.2684... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18487... Training loss: 1.2649... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18488... Training loss: 1.2565... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18489... Training loss: 1.2152... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18490... Training loss: 1.2671... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18491... Training loss: 1.2964... 0.0799 sec/batch
Epoch: 19/20... Training Step: 18492... Training loss: 1.2324... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18493... Training loss: 1.3339... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18494... Training loss: 1.2468... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18495... Training loss: 1.3017... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18496... Training loss: 1.3641... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18497... Training loss: 1.3681... 0.0781 sec/batch
Epoch: 19/20... Training Step: 18498... Training loss: 1.3500... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18499... Training loss: 1.2546... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18500... Training loss: 1.2837... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18501... Training loss: 1.2409... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18502... Training loss: 1.2860... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18503... Training loss: 1.2940... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18504... Training loss: 1.2010... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18505... Training loss: 1.2188... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18506... Training loss: 1.2659... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18507... Training loss: 1.2598... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18508... Training loss: 1.3185... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18509... Training loss: 1.2523... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18510... Training loss: 1.3182... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18511... Training loss: 1.3498... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18512... Training loss: 1.3019... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18513... Training loss: 1.1950... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18514... Training loss: 1.2494... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18515... Training loss: 1.2394... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18516... Training loss: 1.3611... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18517... Training loss: 1.3081... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18518... Training loss: 1.3493... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18519... Training loss: 1.2825... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18520... Training loss: 1.3524... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18521... Training loss: 1.3832... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18522... Training loss: 1.2374... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18523... Training loss: 1.3829... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18524... Training loss: 1.3056... 0.0801 sec/batch
Epoch: 19/20... Training Step: 18525... Training loss: 1.2866... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18526... Training loss: 1.2568... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18527... Training loss: 1.3101... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18528... Training loss: 1.3338... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18529... Training loss: 1.2738... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18530... Training loss: 1.2762... 0.0792 sec/batch
Epoch: 19/20... Training Step: 18531... Training loss: 1.2353... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18532... Training loss: 1.2594... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18533... Training loss: 1.3427... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18534... Training loss: 1.2783... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18535... Training loss: 1.2220... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18536... Training loss: 1.3379... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18537... Training loss: 1.3470... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18538... Training loss: 1.3169... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18539... Training loss: 1.3154... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18540... Training loss: 1.2546... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18541... Training loss: 1.3426... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18542... Training loss: 1.3278... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18543... Training loss: 1.2745... 0.0787 sec/batch
Epoch: 19/20... Training Step: 18544... Training loss: 1.2825... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18545... Training loss: 1.2305... 0.0822 sec/batch
Epoch: 19/20... Training Step: 18546... Training loss: 1.3136... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18547... Training loss: 1.2894... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18548... Training loss: 1.2864... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18549... Training loss: 1.2779... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18550... Training loss: 1.2122... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18551... Training loss: 1.2948... 0.0795 sec/batch
Epoch: 19/20... Training Step: 18552... Training loss: 1.2757... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18553... Training loss: 1.3607... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18554... Training loss: 1.2733... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18555... Training loss: 1.2802... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18556... Training loss: 1.2914... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18557... Training loss: 1.3120... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18558... Training loss: 1.2633... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18559... Training loss: 1.2864... 0.0738 sec/batch
Epoch: 19/20... Training Step: 18560... Training loss: 1.2035... 0.0836 sec/batch
Epoch: 19/20... Training Step: 18561... Training loss: 1.2213... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18562... Training loss: 1.2226... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18563... Training loss: 1.2653... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18564... Training loss: 1.3011... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18565... Training loss: 1.3105... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18566... Training loss: 1.2673... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18567... Training loss: 1.2407... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18568... Training loss: 1.3127... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18569... Training loss: 1.2451... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18570... Training loss: 1.2599... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18571... Training loss: 1.2911... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18572... Training loss: 1.3083... 0.0850 sec/batch
Epoch: 19/20... Training Step: 18573... Training loss: 1.3338... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18574... Training loss: 1.3003... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18575... Training loss: 1.3090... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18576... Training loss: 1.3832... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18577... Training loss: 1.3124... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18578... Training loss: 1.2853... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18579... Training loss: 1.2698... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18580... Training loss: 1.3248... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18581... Training loss: 1.3279... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18582... Training loss: 1.3674... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18583... Training loss: 1.2589... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18584... Training loss: 1.3422... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18585... Training loss: 1.3285... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18586... Training loss: 1.2529... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18587... Training loss: 1.2662... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18588... Training loss: 1.3075... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18589... Training loss: 1.2821... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18590... Training loss: 1.3174... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18591... Training loss: 1.2871... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18592... Training loss: 1.2751... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18593... Training loss: 1.2757... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18594... Training loss: 1.3491... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18595... Training loss: 1.2510... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18596... Training loss: 1.2894... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18597... Training loss: 1.2290... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18598... Training loss: 1.2733... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18599... Training loss: 1.2857... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18600... Training loss: 1.3222... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18601... Training loss: 1.3232... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18602... Training loss: 1.2707... 0.0741 sec/batch
Epoch: 19/20... Training Step: 18603... Training loss: 1.2752... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18604... Training loss: 1.2852... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18605... Training loss: 1.2254... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18606... Training loss: 1.2472... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18607... Training loss: 1.3785... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18608... Training loss: 1.2683... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18609... Training loss: 1.2990... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18610... Training loss: 1.3177... 0.0815 sec/batch
Epoch: 19/20... Training Step: 18611... Training loss: 1.2063... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18612... Training loss: 1.2617... 0.0885 sec/batch
Epoch: 19/20... Training Step: 18613... Training loss: 1.2981... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18614... Training loss: 1.3059... 0.0793 sec/batch
Epoch: 19/20... Training Step: 18615... Training loss: 1.3195... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18616... Training loss: 1.3639... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18617... Training loss: 1.2881... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18618... Training loss: 1.2264... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18619... Training loss: 1.2986... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18620... Training loss: 1.2790... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18621... Training loss: 1.3259... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18622... Training loss: 1.3249... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18623... Training loss: 1.2795... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18624... Training loss: 1.2316... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18625... Training loss: 1.2430... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18626... Training loss: 1.3001... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18627... Training loss: 1.2993... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18628... Training loss: 1.3628... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18629... Training loss: 1.3215... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18630... Training loss: 1.2966... 0.0787 sec/batch
Epoch: 19/20... Training Step: 18631... Training loss: 1.2866... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18632... Training loss: 1.1874... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18633... Training loss: 1.2447... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18634... Training loss: 1.2862... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18635... Training loss: 1.2454... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18636... Training loss: 1.2636... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18637... Training loss: 1.2756... 0.0792 sec/batch
Epoch: 19/20... Training Step: 18638... Training loss: 1.3023... 0.0801 sec/batch
Epoch: 19/20... Training Step: 18639... Training loss: 1.2972... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18640... Training loss: 1.2856... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18641... Training loss: 1.2819... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18642... Training loss: 1.2994... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18643... Training loss: 1.2809... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18644... Training loss: 1.3076... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18645... Training loss: 1.3016... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18646... Training loss: 1.3192... 0.0851 sec/batch
Epoch: 19/20... Training Step: 18647... Training loss: 1.2781... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18648... Training loss: 1.2761... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18649... Training loss: 1.2622... 0.0794 sec/batch
Epoch: 19/20... Training Step: 18650... Training loss: 1.2558... 0.0742 sec/batch
Epoch: 19/20... Training Step: 18651... Training loss: 1.1936... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18652... Training loss: 1.2686... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18653... Training loss: 1.2408... 0.0739 sec/batch
Epoch: 19/20... Training Step: 18654... Training loss: 1.2824... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18655... Training loss: 1.2326... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18656... Training loss: 1.2654... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18657... Training loss: 1.3108... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18658... Training loss: 1.3650... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18659... Training loss: 1.3531... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18660... Training loss: 1.2795... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18661... Training loss: 1.2333... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18662... Training loss: 1.2894... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18663... Training loss: 1.2694... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18664... Training loss: 1.3013... 0.0823 sec/batch
Epoch: 19/20... Training Step: 18665... Training loss: 1.3277... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18666... Training loss: 1.3290... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18667... Training loss: 1.2663... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18668... Training loss: 1.3468... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18669... Training loss: 1.2881... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18670... Training loss: 1.3212... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18671... Training loss: 1.3108... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18672... Training loss: 1.3176... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18673... Training loss: 1.3316... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18674... Training loss: 1.3877... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18675... Training loss: 1.3097... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18676... Training loss: 1.2895... 0.0788 sec/batch
Epoch: 19/20... Training Step: 18677... Training loss: 1.2626... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18678... Training loss: 1.2349... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18679... Training loss: 1.2914... 0.0802 sec/batch
Epoch: 19/20... Training Step: 18680... Training loss: 1.2696... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18681... Training loss: 1.2939... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18682... Training loss: 1.3107... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18683... Training loss: 1.3101... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18684... Training loss: 1.3244... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18685... Training loss: 1.3308... 0.0945 sec/batch
Epoch: 19/20... Training Step: 18686... Training loss: 1.2526... 0.0781 sec/batch
Epoch: 19/20... Training Step: 18687... Training loss: 1.3148... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18688... Training loss: 1.2958... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18689... Training loss: 1.3327... 0.0753 sec/batch
Epoch: 19/20... Training Step: 18690... Training loss: 1.2874... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18691... Training loss: 1.2242... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18692... Training loss: 1.2555... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18693... Training loss: 1.3062... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18694... Training loss: 1.2911... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18695... Training loss: 1.3133... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18696... Training loss: 1.2873... 0.0775 sec/batch
Epoch: 19/20... Training Step: 18697... Training loss: 1.2892... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18698... Training loss: 1.4197... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18699... Training loss: 1.3229... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18700... Training loss: 1.3062... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18701... Training loss: 1.3233... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18702... Training loss: 1.2153... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18703... Training loss: 1.2980... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18704... Training loss: 1.2772... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18705... Training loss: 1.2852... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18706... Training loss: 1.2975... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18707... Training loss: 1.2810... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18708... Training loss: 1.3643... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18709... Training loss: 1.3845... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18710... Training loss: 1.2838... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18711... Training loss: 1.2968... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18712... Training loss: 1.2891... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18713... Training loss: 1.2923... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18714... Training loss: 1.2703... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18715... Training loss: 1.3051... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18716... Training loss: 1.3366... 0.0747 sec/batch
Epoch: 19/20... Training Step: 18717... Training loss: 1.2945... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18718... Training loss: 1.2475... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18719... Training loss: 1.2493... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18720... Training loss: 1.2976... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18721... Training loss: 1.2272... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18722... Training loss: 1.2824... 0.0828 sec/batch
Epoch: 19/20... Training Step: 18723... Training loss: 1.2918... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18724... Training loss: 1.3149... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18725... Training loss: 1.3160... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18726... Training loss: 1.3548... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18727... Training loss: 1.3240... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18728... Training loss: 1.3062... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18729... Training loss: 1.2854... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18730... Training loss: 1.3249... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18731... Training loss: 1.3032... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18732... Training loss: 1.2929... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18733... Training loss: 1.2796... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18734... Training loss: 1.3075... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18735... Training loss: 1.3073... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18736... Training loss: 1.2772... 0.0859 sec/batch
Epoch: 19/20... Training Step: 18737... Training loss: 1.2923... 0.0749 sec/batch
Epoch: 19/20... Training Step: 18738... Training loss: 1.2317... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18739... Training loss: 1.3063... 0.0804 sec/batch
Epoch: 19/20... Training Step: 18740... Training loss: 1.2947... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18741... Training loss: 1.3031... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18742... Training loss: 1.3530... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18743... Training loss: 1.3035... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18744... Training loss: 1.3224... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18745... Training loss: 1.3236... 0.0789 sec/batch
Epoch: 19/20... Training Step: 18746... Training loss: 1.3624... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18747... Training loss: 1.3171... 0.0776 sec/batch
Epoch: 19/20... Training Step: 18748... Training loss: 1.2134... 0.0782 sec/batch
Epoch: 19/20... Training Step: 18749... Training loss: 1.2773... 0.0734 sec/batch
Epoch: 19/20... Training Step: 18750... Training loss: 1.2161... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18751... Training loss: 1.3310... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18752... Training loss: 1.2557... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18753... Training loss: 1.3252... 0.0746 sec/batch
Epoch: 19/20... Training Step: 18754... Training loss: 1.3024... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18755... Training loss: 1.2536... 0.0740 sec/batch
Epoch: 19/20... Training Step: 18756... Training loss: 1.2247... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18757... Training loss: 1.2298... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18758... Training loss: 1.2683... 0.0770 sec/batch
Epoch: 19/20... Training Step: 18759... Training loss: 1.2544... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18760... Training loss: 1.2468... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18761... Training loss: 1.2050... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18762... Training loss: 1.3269... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18763... Training loss: 1.2410... 0.0821 sec/batch
Epoch: 19/20... Training Step: 18764... Training loss: 1.2695... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18765... Training loss: 1.2134... 0.0839 sec/batch
Epoch: 19/20... Training Step: 18766... Training loss: 1.1990... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18767... Training loss: 1.2149... 0.0778 sec/batch
Epoch: 19/20... Training Step: 18768... Training loss: 1.2755... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18769... Training loss: 1.3745... 0.0779 sec/batch
Epoch: 19/20... Training Step: 18770... Training loss: 1.2643... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18771... Training loss: 1.2608... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18772... Training loss: 1.2686... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18773... Training loss: 1.2813... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18774... Training loss: 1.3199... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18775... Training loss: 1.2571... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18776... Training loss: 1.3004... 0.0745 sec/batch
Epoch: 19/20... Training Step: 18777... Training loss: 1.3930... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18778... Training loss: 1.3425... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18779... Training loss: 1.2788... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18780... Training loss: 1.2459... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18781... Training loss: 1.2891... 0.0764 sec/batch
Epoch: 19/20... Training Step: 18782... Training loss: 1.3287... 0.0767 sec/batch
Epoch: 19/20... Training Step: 18783... Training loss: 1.2870... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18784... Training loss: 1.2710... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18785... Training loss: 1.3471... 0.0739 sec/batch
Epoch: 19/20... Training Step: 18786... Training loss: 1.3445... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18787... Training loss: 1.3045... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18788... Training loss: 1.3982... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18789... Training loss: 1.2795... 0.0738 sec/batch
Epoch: 19/20... Training Step: 18790... Training loss: 1.3113... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18791... Training loss: 1.2936... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18792... Training loss: 1.3241... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18793... Training loss: 1.2787... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18794... Training loss: 1.2546... 0.0771 sec/batch
Epoch: 19/20... Training Step: 18795... Training loss: 1.2766... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18796... Training loss: 1.3451... 0.0772 sec/batch
Epoch: 19/20... Training Step: 18797... Training loss: 1.2982... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18798... Training loss: 1.3243... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18799... Training loss: 1.3441... 0.0794 sec/batch
Epoch: 19/20... Training Step: 18800... Training loss: 1.3031... 0.0781 sec/batch
Epoch: 19/20... Training Step: 18801... Training loss: 1.3412... 0.0755 sec/batch
Epoch: 19/20... Training Step: 18802... Training loss: 1.2934... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18803... Training loss: 1.2838... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18804... Training loss: 1.2546... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18805... Training loss: 1.2696... 0.0759 sec/batch
Epoch: 19/20... Training Step: 18806... Training loss: 1.2045... 0.0774 sec/batch
Epoch: 19/20... Training Step: 18807... Training loss: 1.2792... 0.0783 sec/batch
Epoch: 19/20... Training Step: 18808... Training loss: 1.2489... 0.0762 sec/batch
Epoch: 19/20... Training Step: 18809... Training loss: 1.2756... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18810... Training loss: 1.3108... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18811... Training loss: 1.3080... 0.0756 sec/batch
Epoch: 19/20... Training Step: 18812... Training loss: 1.2869... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18813... Training loss: 1.2690... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18814... Training loss: 1.2955... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18815... Training loss: 1.2533... 0.0744 sec/batch
Epoch: 19/20... Training Step: 18816... Training loss: 1.2591... 0.0754 sec/batch
Epoch: 19/20... Training Step: 18817... Training loss: 1.3015... 0.0785 sec/batch
Epoch: 19/20... Training Step: 18818... Training loss: 1.2873... 0.0768 sec/batch
Epoch: 19/20... Training Step: 18819... Training loss: 1.2886... 0.0766 sec/batch
Epoch: 19/20... Training Step: 18820... Training loss: 1.2711... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18821... Training loss: 1.3649... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18822... Training loss: 1.3331... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18823... Training loss: 1.3338... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18824... Training loss: 1.3274... 0.0763 sec/batch
Epoch: 19/20... Training Step: 18825... Training loss: 1.3202... 0.0750 sec/batch
Epoch: 19/20... Training Step: 18826... Training loss: 1.4601... 0.0773 sec/batch
Epoch: 19/20... Training Step: 18827... Training loss: 1.2821... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18828... Training loss: 1.3703... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18829... Training loss: 1.4034... 0.0757 sec/batch
Epoch: 19/20... Training Step: 18830... Training loss: 1.2555... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18831... Training loss: 1.3243... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18832... Training loss: 1.3568... 0.0758 sec/batch
Epoch: 19/20... Training Step: 18833... Training loss: 1.3776... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18834... Training loss: 1.3332... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18835... Training loss: 1.3461... 0.0780 sec/batch
Epoch: 19/20... Training Step: 18836... Training loss: 1.2571... 0.0769 sec/batch
Epoch: 19/20... Training Step: 18837... Training loss: 1.2438... 0.0748 sec/batch
Epoch: 19/20... Training Step: 18838... Training loss: 1.3896... 0.0765 sec/batch
Epoch: 19/20... Training Step: 18839... Training loss: 1.3050... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18840... Training loss: 1.3062... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18841... Training loss: 1.3105... 0.0751 sec/batch
Epoch: 19/20... Training Step: 18842... Training loss: 1.2434... 0.0784 sec/batch
Epoch: 19/20... Training Step: 18843... Training loss: 1.2411... 0.0790 sec/batch
Epoch: 19/20... Training Step: 18844... Training loss: 1.2733... 0.0777 sec/batch
Epoch: 19/20... Training Step: 18845... Training loss: 1.2879... 0.0760 sec/batch
Epoch: 19/20... Training Step: 18846... Training loss: 1.1941... 0.0761 sec/batch
Epoch: 19/20... Training Step: 18847... Training loss: 1.2562... 0.0752 sec/batch
Epoch: 19/20... Training Step: 18848... Training loss: 1.3443... 0.0748 sec/batch
Epoch: 20/20... Training Step: 18849... Training loss: 1.3803... 0.0739 sec/batch
Epoch: 20/20... Training Step: 18850... Training loss: 1.2451... 0.0738 sec/batch
Epoch: 20/20... Training Step: 18851... Training loss: 1.3387... 0.0768 sec/batch
Epoch: 20/20... Training Step: 18852... Training loss: 1.2986... 0.0785 sec/batch
Epoch: 20/20... Training Step: 18853... Training loss: 1.2316... 0.0749 sec/batch
Epoch: 20/20... Training Step: 18854... Training loss: 1.1580... 0.0739 sec/batch
Epoch: 20/20... Training Step: 18855... Training loss: 1.2549... 0.0776 sec/batch
Epoch: 20/20... Training Step: 18856... Training loss: 1.2978... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18857... Training loss: 1.3136... 0.0744 sec/batch
Epoch: 20/20... Training Step: 18858... Training loss: 1.2296... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18859... Training loss: 1.3415... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18860... Training loss: 1.2626... 0.0742 sec/batch
Epoch: 20/20... Training Step: 18861... Training loss: 1.2764... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18862... Training loss: 1.2629... 0.0750 sec/batch
Epoch: 20/20... Training Step: 18863... Training loss: 1.3244... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18864... Training loss: 1.2204... 0.0758 sec/batch
Epoch: 20/20... Training Step: 18865... Training loss: 1.2816... 0.0750 sec/batch
Epoch: 20/20... Training Step: 18866... Training loss: 1.3148... 0.0773 sec/batch
Epoch: 20/20... Training Step: 18867... Training loss: 1.2570... 0.0749 sec/batch
Epoch: 20/20... Training Step: 18868... Training loss: 1.2583... 0.0770 sec/batch
Epoch: 20/20... Training Step: 18869... Training loss: 1.2982... 0.0776 sec/batch
Epoch: 20/20... Training Step: 18870... Training loss: 1.2636... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18871... Training loss: 1.2613... 0.0753 sec/batch
Epoch: 20/20... Training Step: 18872... Training loss: 1.2214... 0.0781 sec/batch
Epoch: 20/20... Training Step: 18873... Training loss: 1.3098... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18874... Training loss: 1.2237... 0.0770 sec/batch
Epoch: 20/20... Training Step: 18875... Training loss: 1.2593... 0.0744 sec/batch
Epoch: 20/20... Training Step: 18876... Training loss: 1.2255... 0.0823 sec/batch
Epoch: 20/20... Training Step: 18877... Training loss: 1.2706... 0.0774 sec/batch
Epoch: 20/20... Training Step: 18878... Training loss: 1.3214... 0.0751 sec/batch
Epoch: 20/20... Training Step: 18879... Training loss: 1.3175... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18880... Training loss: 1.3140... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18881... Training loss: 1.2783... 0.0785 sec/batch
Epoch: 20/20... Training Step: 18882... Training loss: 1.3220... 0.0774 sec/batch
Epoch: 20/20... Training Step: 18883... Training loss: 1.3498... 0.0766 sec/batch
Epoch: 20/20... Training Step: 18884... Training loss: 1.3044... 0.0775 sec/batch
Epoch: 20/20... Training Step: 18885... Training loss: 1.2701... 0.0841 sec/batch
Epoch: 20/20... Training Step: 18886... Training loss: 1.3365... 0.0771 sec/batch
Epoch: 20/20... Training Step: 18887... Training loss: 1.2891... 0.0769 sec/batch
Epoch: 20/20... Training Step: 18888... Training loss: 1.2565... 0.0749 sec/batch
Epoch: 20/20... Training Step: 18889... Training loss: 1.2985... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18890... Training loss: 1.2644... 0.0766 sec/batch
Epoch: 20/20... Training Step: 18891... Training loss: 1.2511... 0.0787 sec/batch
Epoch: 20/20... Training Step: 18892... Training loss: 1.2606... 0.0753 sec/batch
Epoch: 20/20... Training Step: 18893... Training loss: 1.2827... 0.0750 sec/batch
Epoch: 20/20... Training Step: 18894... Training loss: 1.2819... 0.0785 sec/batch
Epoch: 20/20... Training Step: 18895... Training loss: 1.3160... 0.0745 sec/batch
Epoch: 20/20... Training Step: 18896... Training loss: 1.2606... 0.0743 sec/batch
Epoch: 20/20... Training Step: 18897... Training loss: 1.3218... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18898... Training loss: 1.3174... 0.0771 sec/batch
Epoch: 20/20... Training Step: 18899... Training loss: 1.2070... 0.0762 sec/batch
Epoch: 20/20... Training Step: 18900... Training loss: 1.3222... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18901... Training loss: 1.2247... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18902... Training loss: 1.2293... 0.0764 sec/batch
Epoch: 20/20... Training Step: 18903... Training loss: 1.2465... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18904... Training loss: 1.2412... 0.0768 sec/batch
Epoch: 20/20... Training Step: 18905... Training loss: 1.3034... 0.0740 sec/batch
Epoch: 20/20... Training Step: 18906... Training loss: 1.3365... 0.0757 sec/batch
Epoch: 20/20... Training Step: 18907... Training loss: 1.2891... 0.0747 sec/batch
Epoch: 20/20... Training Step: 18908... Training loss: 1.3016... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18909... Training loss: 1.2073... 0.0777 sec/batch
Epoch: 20/20... Training Step: 18910... Training loss: 1.2224... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18911... Training loss: 1.2760... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18912... Training loss: 1.2196... 0.0754 sec/batch
Epoch: 20/20... Training Step: 18913... Training loss: 1.3044... 0.0742 sec/batch
Epoch: 20/20... Training Step: 18914... Training loss: 1.2723... 0.0760 sec/batch
Epoch: 20/20... Training Step: 18915... Training loss: 1.2625... 0.0776 sec/batch
Epoch: 20/20... Training Step: 18916... Training loss: 1.2680... 0.0766 sec/batch
Epoch: 20/20... Training Step: 18917... Training loss: 1.3039... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18918... Training loss: 1.2659... 0.0749 sec/batch
Epoch: 20/20... Training Step: 18919... Training loss: 1.2961... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18920... Training loss: 1.3431... 0.0800 sec/batch
Epoch: 20/20... Training Step: 18921... Training loss: 1.3047... 0.0846 sec/batch
Epoch: 20/20... Training Step: 18922... Training loss: 1.2376... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18923... Training loss: 1.2705... 0.0777 sec/batch
Epoch: 20/20... Training Step: 18924... Training loss: 1.2315... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18925... Training loss: 1.2464... 0.0750 sec/batch
Epoch: 20/20... Training Step: 18926... Training loss: 1.2780... 0.0776 sec/batch
Epoch: 20/20... Training Step: 18927... Training loss: 1.2684... 0.0778 sec/batch
Epoch: 20/20... Training Step: 18928... Training loss: 1.3235... 0.0757 sec/batch
Epoch: 20/20... Training Step: 18929... Training loss: 1.2643... 0.0744 sec/batch
Epoch: 20/20... Training Step: 18930... Training loss: 1.2492... 0.0775 sec/batch
Epoch: 20/20... Training Step: 18931... Training loss: 1.2935... 0.0743 sec/batch
Epoch: 20/20... Training Step: 18932... Training loss: 1.2922... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18933... Training loss: 1.2491... 0.0762 sec/batch
Epoch: 20/20... Training Step: 18934... Training loss: 1.3377... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18935... Training loss: 1.2920... 0.0758 sec/batch
Epoch: 20/20... Training Step: 18936... Training loss: 1.2853... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18937... Training loss: 1.3240... 0.0745 sec/batch
Epoch: 20/20... Training Step: 18938... Training loss: 1.3317... 0.0759 sec/batch
Epoch: 20/20... Training Step: 18939... Training loss: 1.3590... 0.0815 sec/batch
Epoch: 20/20... Training Step: 18940... Training loss: 1.2610... 0.0750 sec/batch
Epoch: 20/20... Training Step: 18941... Training loss: 1.2432... 0.0748 sec/batch
Epoch: 20/20... Training Step: 18942... Training loss: 1.2067... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18943... Training loss: 1.1790... 0.0753 sec/batch
Epoch: 20/20... Training Step: 18944... Training loss: 1.2606... 0.0760 sec/batch
Epoch: 20/20... Training Step: 18945... Training loss: 1.2502... 0.0771 sec/batch
Epoch: 20/20... Training Step: 18946... Training loss: 1.2334... 0.0775 sec/batch
Epoch: 20/20... Training Step: 18947... Training loss: 1.2422... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18948... Training loss: 1.2435... 0.0772 sec/batch
Epoch: 20/20... Training Step: 18949... Training loss: 1.2282... 0.0747 sec/batch
Epoch: 20/20... Training Step: 18950... Training loss: 1.2387... 0.0756 sec/batch
Epoch: 20/20... Training Step: 18951... Training loss: 1.2901... 0.0740 sec/batch
Epoch: 20/20... Training Step: 18952... Training loss: 1.2944... 0.0763 sec/batch
Epoch: 20/20... Training Step: 18953... Training loss: 1.3044... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18954... Training loss: 1.2656... 0.0767 sec/batch
Epoch: 20/20... Training Step: 18955... Training loss: 1.2055... 0.0769 sec/batch
Epoch: 20/20... Training Step: 18956... Training loss: 1.3183... 0.0770 sec/batch
Epoch: 20/20... Training Step: 18957... Training loss: 1.2664... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18958... Training loss: 1.2934... 0.0779 sec/batch
Epoch: 20/20... Training Step: 18959... Training loss: 1.2720... 0.0782 sec/batch
Epoch: 20/20... Training Step: 18960... Training loss: 1.2822... 0.0758 sec/batch
Epoch: 20/20... Training Step: 18961... Training loss: 1.3274... 0.0768 sec/batch
Epoch: 20/20... Training Step: 18962... Training loss: 1.2975... 0.0770 sec/batch
Epoch: 20/20... Training Step: 18963... Training loss: 1.2289... 0.0781 sec/batch
Epoch: 20/20... Training Step: 18964... Training loss: 1.3762... 0.0783 sec/batch
Epoch: 20/20... Training Step: 18965... Training loss: 1.3123... 0.0742 sec/batch
Epoch: 20/20... Training Step: 18966... Training loss: 1.3867... 0.0779 sec/batch
Epoch: 20/20... Training Step: 18967... Training loss: 1.2770... 0.0744 sec/batch
Epoch: 20/20... Training Step: 18968... Training loss: 1.2963... 0.0770 sec/batch
Epoch: 20/20... Training Step: 18969... Training loss: 1.3931... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18970... Training loss: 1.2691... 0.0766 sec/batch
Epoch: 20/20... Training Step: 18971... Training loss: 1.3066... 0.0769 sec/batch
Epoch: 20/20... Training Step: 18972... Training loss: 1.2779... 0.0845 sec/batch
Epoch: 20/20... Training Step: 18973... Training loss: 1.3370... 0.0781 sec/batch
Epoch: 20/20... Training Step: 18974... Training loss: 1.3295... 0.0768 sec/batch
Epoch: 20/20... Training Step: 18975... Training loss: 1.3418... 0.0785 sec/batch
Epoch: 20/20... Training Step: 18976... Training loss: 1.2928... 0.0737 sec/batch
Epoch: 20/20... Training Step: 18977... Training loss: 1.2112... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18978... Training loss: 1.3089... 0.0757 sec/batch
Epoch: 20/20... Training Step: 18979... Training loss: 1.2570... 0.0757 sec/batch
Epoch: 20/20... Training Step: 18980... Training loss: 1.3271... 0.0748 sec/batch
Epoch: 20/20... Training Step: 18981... Training loss: 1.3029... 0.0777 sec/batch
Epoch: 20/20... Training Step: 18982... Training loss: 1.2733... 0.0751 sec/batch
Epoch: 20/20... Training Step: 18983... Training loss: 1.3000... 0.0774 sec/batch
Epoch: 20/20... Training Step: 18984... Training loss: 1.3014... 0.0762 sec/batch
Epoch: 20/20... Training Step: 18985... Training loss: 1.3196... 0.0761 sec/batch
Epoch: 20/20... Training Step: 18986... Training loss: 1.3513... 0.0776 sec/batch
Epoch: 20/20... Training Step: 18987... Training loss: 1.3163... 0.0762 sec/batch
Epoch: 20/20... Training Step: 18988... Training loss: 1.2845... 0.0778 sec/batch
Epoch: 20/20... Training Step: 18989... Training loss: 1.2925... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18990... Training loss: 1.2611... 0.0813 sec/batch
Epoch: 20/20... Training Step: 18991... Training loss: 1.2651... 0.0754 sec/batch
Epoch: 20/20... Training Step: 18992... Training loss: 1.2801... 0.0785 sec/batch
Epoch: 20/20... Training Step: 18993... Training loss: 1.1637... 0.0753 sec/batch
Epoch: 20/20... Training Step: 18994... Training loss: 1.3017... 0.0765 sec/batch
Epoch: 20/20... Training Step: 18995... Training loss: 1.2610... 0.0755 sec/batch
Epoch: 20/20... Training Step: 18996... Training loss: 1.2294... 0.0775 sec/batch
Epoch: 20/20... Training Step: 18997... Training loss: 1.2563... 0.0745 sec/batch
Epoch: 20/20... Training Step: 18998... Training loss: 1.2213... 0.0780 sec/batch
Epoch: 20/20... Training Step: 18999... Training loss: 1.1713... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19000... Training loss: 1.2713... 0.0790 sec/batch
Epoch: 20/20... Training Step: 19001... Training loss: 1.2120... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19002... Training loss: 1.2587... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19003... Training loss: 1.2546... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19004... Training loss: 1.2501... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19005... Training loss: 1.2583... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19006... Training loss: 1.2787... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19007... Training loss: 1.2735... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19008... Training loss: 1.3212... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19009... Training loss: 1.2532... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19010... Training loss: 1.2827... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19011... Training loss: 1.2751... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19012... Training loss: 1.3386... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19013... Training loss: 1.2817... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19014... Training loss: 1.3724... 0.0799 sec/batch
Epoch: 20/20... Training Step: 19015... Training loss: 1.3241... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19016... Training loss: 1.2677... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19017... Training loss: 1.2940... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19018... Training loss: 1.2770... 0.0739 sec/batch
Epoch: 20/20... Training Step: 19019... Training loss: 1.3094... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19020... Training loss: 1.2649... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19021... Training loss: 1.2756... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19022... Training loss: 1.3281... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19023... Training loss: 1.2431... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19024... Training loss: 1.3204... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19025... Training loss: 1.2843... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19026... Training loss: 1.3860... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19027... Training loss: 1.3196... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19028... Training loss: 1.2519... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19029... Training loss: 1.2939... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19030... Training loss: 1.3632... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19031... Training loss: 1.2811... 0.0789 sec/batch
Epoch: 20/20... Training Step: 19032... Training loss: 1.2855... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19033... Training loss: 1.3067... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19034... Training loss: 1.3593... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19035... Training loss: 1.2918... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19036... Training loss: 1.2757... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19037... Training loss: 1.2865... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19038... Training loss: 1.3421... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19039... Training loss: 1.3362... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19040... Training loss: 1.3228... 0.0822 sec/batch
Epoch: 20/20... Training Step: 19041... Training loss: 1.3357... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19042... Training loss: 1.3833... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19043... Training loss: 1.2896... 0.0824 sec/batch
Epoch: 20/20... Training Step: 19044... Training loss: 1.2325... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19045... Training loss: 1.3520... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19046... Training loss: 1.3067... 0.0847 sec/batch
Epoch: 20/20... Training Step: 19047... Training loss: 1.3041... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19048... Training loss: 1.2908... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19049... Training loss: 1.2657... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19050... Training loss: 1.2855... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19051... Training loss: 1.2771... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19052... Training loss: 1.2722... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19053... Training loss: 1.2591... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19054... Training loss: 1.3367... 0.0787 sec/batch
Epoch: 20/20... Training Step: 19055... Training loss: 1.2756... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19056... Training loss: 1.2792... 0.0794 sec/batch
Epoch: 20/20... Training Step: 19057... Training loss: 1.3207... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19058... Training loss: 1.2996... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19059... Training loss: 1.2324... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19060... Training loss: 1.3180... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19061... Training loss: 1.2829... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19062... Training loss: 1.3353... 0.0742 sec/batch
Epoch: 20/20... Training Step: 19063... Training loss: 1.2670... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19064... Training loss: 1.2843... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19065... Training loss: 1.2503... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19066... Training loss: 1.2983... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19067... Training loss: 1.2892... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19068... Training loss: 1.2419... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19069... Training loss: 1.2300... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19070... Training loss: 1.3709... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19071... Training loss: 1.2981... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19072... Training loss: 1.3054... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19073... Training loss: 1.3117... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19074... Training loss: 1.2664... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19075... Training loss: 1.2928... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19076... Training loss: 1.3197... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19077... Training loss: 1.2217... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19078... Training loss: 1.2230... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19079... Training loss: 1.2793... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19080... Training loss: 1.2850... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19081... Training loss: 1.2758... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19082... Training loss: 1.3573... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19083... Training loss: 1.2713... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19084... Training loss: 1.2828... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19085... Training loss: 1.2539... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19086... Training loss: 1.2925... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19087... Training loss: 1.3215... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19088... Training loss: 1.2861... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19089... Training loss: 1.2753... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19090... Training loss: 1.2811... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19091... Training loss: 1.3042... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19092... Training loss: 1.3168... 0.0806 sec/batch
Epoch: 20/20... Training Step: 19093... Training loss: 1.2679... 0.0830 sec/batch
Epoch: 20/20... Training Step: 19094... Training loss: 1.3114... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19095... Training loss: 1.2392... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19096... Training loss: 1.2081... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19097... Training loss: 1.2647... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19098... Training loss: 1.3071... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19099... Training loss: 1.2518... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19100... Training loss: 1.2693... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19101... Training loss: 1.2513... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19102... Training loss: 1.2486... 0.0731 sec/batch
Epoch: 20/20... Training Step: 19103... Training loss: 1.2365... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19104... Training loss: 1.2423... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19105... Training loss: 1.3716... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19106... Training loss: 1.2519... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19107... Training loss: 1.2899... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19108... Training loss: 1.2751... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19109... Training loss: 1.3199... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19110... Training loss: 1.3194... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19111... Training loss: 1.3103... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19112... Training loss: 1.2961... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19113... Training loss: 1.3038... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19114... Training loss: 1.1681... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19115... Training loss: 1.2345... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19116... Training loss: 1.2982... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19117... Training loss: 1.2188... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19118... Training loss: 1.2544... 0.0804 sec/batch
Epoch: 20/20... Training Step: 19119... Training loss: 1.2852... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19120... Training loss: 1.2510... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19121... Training loss: 1.2034... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19122... Training loss: 1.2165... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19123... Training loss: 1.2679... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19124... Training loss: 1.2848... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19125... Training loss: 1.2695... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19126... Training loss: 1.1681... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19127... Training loss: 1.2419... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19128... Training loss: 1.3248... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19129... Training loss: 1.2586... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19130... Training loss: 1.2480... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19131... Training loss: 1.2749... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19132... Training loss: 1.2490... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19133... Training loss: 1.1495... 0.0805 sec/batch
Epoch: 20/20... Training Step: 19134... Training loss: 1.2272... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19135... Training loss: 1.2885... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19136... Training loss: 1.2580... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19137... Training loss: 1.2355... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19138... Training loss: 1.2107... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19139... Training loss: 1.2640... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19140... Training loss: 1.2675... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19141... Training loss: 1.2126... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19142... Training loss: 1.2440... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19143... Training loss: 1.2782... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19144... Training loss: 1.2858... 0.0796 sec/batch
Epoch: 20/20... Training Step: 19145... Training loss: 1.2619... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19146... Training loss: 1.3380... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19147... Training loss: 1.2559... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19148... Training loss: 1.2977... 0.0788 sec/batch
Epoch: 20/20... Training Step: 19149... Training loss: 1.3146... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19150... Training loss: 1.2260... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19151... Training loss: 1.2487... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19152... Training loss: 1.2862... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19153... Training loss: 1.2540... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19154... Training loss: 1.3313... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19155... Training loss: 1.2497... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19156... Training loss: 1.2147... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19157... Training loss: 1.3225... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19158... Training loss: 1.2461... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19159... Training loss: 1.3371... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19160... Training loss: 1.2735... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19161... Training loss: 1.2458... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19162... Training loss: 1.2651... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19163... Training loss: 1.2826... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19164... Training loss: 1.2471... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19165... Training loss: 1.2427... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19166... Training loss: 1.2253... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19167... Training loss: 1.2659... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19168... Training loss: 1.2915... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19169... Training loss: 1.2797... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19170... Training loss: 1.2241... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19171... Training loss: 1.2444... 0.0816 sec/batch
Epoch: 20/20... Training Step: 19172... Training loss: 1.2640... 0.0821 sec/batch
Epoch: 20/20... Training Step: 19173... Training loss: 1.3000... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19174... Training loss: 1.2664... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19175... Training loss: 1.2600... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19176... Training loss: 1.2688... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19177... Training loss: 1.2638... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19178... Training loss: 1.2970... 0.0788 sec/batch
Epoch: 20/20... Training Step: 19179... Training loss: 1.3103... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19180... Training loss: 1.2452... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19181... Training loss: 1.3784... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19182... Training loss: 1.3631... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19183... Training loss: 1.3598... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19184... Training loss: 1.2955... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19185... Training loss: 1.2506... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19186... Training loss: 1.2511... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19187... Training loss: 1.2758... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19188... Training loss: 1.2545... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19189... Training loss: 1.2740... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19190... Training loss: 1.2375... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19191... Training loss: 1.2261... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19192... Training loss: 1.2833... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19193... Training loss: 1.2773... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19194... Training loss: 1.2756... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19195... Training loss: 1.2828... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19196... Training loss: 1.3023... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19197... Training loss: 1.3398... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19198... Training loss: 1.3138... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19199... Training loss: 1.3120... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19200... Training loss: 1.2989... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19201... Training loss: 1.2867... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19202... Training loss: 1.2619... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19203... Training loss: 1.2741... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19204... Training loss: 1.2801... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19205... Training loss: 1.2491... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19206... Training loss: 1.2981... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19207... Training loss: 1.2867... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19208... Training loss: 1.2671... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19209... Training loss: 1.2797... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19210... Training loss: 1.3178... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19211... Training loss: 1.2802... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19212... Training loss: 1.2662... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19213... Training loss: 1.2674... 0.0799 sec/batch
Epoch: 20/20... Training Step: 19214... Training loss: 1.2477... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19215... Training loss: 1.2574... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19216... Training loss: 1.2108... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19217... Training loss: 1.2401... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19218... Training loss: 1.2311... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19219... Training loss: 1.3308... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19220... Training loss: 1.2262... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19221... Training loss: 1.2471... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19222... Training loss: 1.2049... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19223... Training loss: 1.2466... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19224... Training loss: 1.2402... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19225... Training loss: 1.2574... 0.0787 sec/batch
Epoch: 20/20... Training Step: 19226... Training loss: 1.2863... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19227... Training loss: 1.2455... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19228... Training loss: 1.2822... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19229... Training loss: 1.2835... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19230... Training loss: 1.2385... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19231... Training loss: 1.3165... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19232... Training loss: 1.2887... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19233... Training loss: 1.2952... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19234... Training loss: 1.2876... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19235... Training loss: 1.3070... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19236... Training loss: 1.2742... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19237... Training loss: 1.3311... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19238... Training loss: 1.2613... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19239... Training loss: 1.3005... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19240... Training loss: 1.2811... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19241... Training loss: 1.3055... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19242... Training loss: 1.2650... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19243... Training loss: 1.2795... 0.0787 sec/batch
Epoch: 20/20... Training Step: 19244... Training loss: 1.2442... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19245... Training loss: 1.2479... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19246... Training loss: 1.3503... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19247... Training loss: 1.2931... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19248... Training loss: 1.2482... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19249... Training loss: 1.2423... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19250... Training loss: 1.2857... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19251... Training loss: 1.2302... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19252... Training loss: 1.3074... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19253... Training loss: 1.2639... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19254... Training loss: 1.2737... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19255... Training loss: 1.1988... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19256... Training loss: 1.3422... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19257... Training loss: 1.2250... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19258... Training loss: 1.2248... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19259... Training loss: 1.1792... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19260... Training loss: 1.2443... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19261... Training loss: 1.2719... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19262... Training loss: 1.1993... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19263... Training loss: 1.2269... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19264... Training loss: 1.3142... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19265... Training loss: 1.2010... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19266... Training loss: 1.2376... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19267... Training loss: 1.2962... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19268... Training loss: 1.3281... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19269... Training loss: 1.2060... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19270... Training loss: 1.2956... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19271... Training loss: 1.2607... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19272... Training loss: 1.2918... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19273... Training loss: 1.2386... 0.0808 sec/batch
Epoch: 20/20... Training Step: 19274... Training loss: 1.2834... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19275... Training loss: 1.3221... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19276... Training loss: 1.3799... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19277... Training loss: 1.3231... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19278... Training loss: 1.3599... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19279... Training loss: 1.2814... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19280... Training loss: 1.3022... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19281... Training loss: 1.3105... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19282... Training loss: 1.3105... 0.0823 sec/batch
Epoch: 20/20... Training Step: 19283... Training loss: 1.3210... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19284... Training loss: 1.2540... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19285... Training loss: 1.3254... 0.0790 sec/batch
Epoch: 20/20... Training Step: 19286... Training loss: 1.3617... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19287... Training loss: 1.3092... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19288... Training loss: 1.3281... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19289... Training loss: 1.2632... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19290... Training loss: 1.3213... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19291... Training loss: 1.2878... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19292... Training loss: 1.3068... 0.0742 sec/batch
Epoch: 20/20... Training Step: 19293... Training loss: 1.2757... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19294... Training loss: 1.3547... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19295... Training loss: 1.2785... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19296... Training loss: 1.2875... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19297... Training loss: 1.3526... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19298... Training loss: 1.3851... 0.0736 sec/batch
Epoch: 20/20... Training Step: 19299... Training loss: 1.2882... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19300... Training loss: 1.2709... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19301... Training loss: 1.3217... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19302... Training loss: 1.2945... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19303... Training loss: 1.3056... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19304... Training loss: 1.2713... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19305... Training loss: 1.2675... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19306... Training loss: 1.2627... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19307... Training loss: 1.2496... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19308... Training loss: 1.2823... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19309... Training loss: 1.3004... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19310... Training loss: 1.2769... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19311... Training loss: 1.2491... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19312... Training loss: 1.3044... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19313... Training loss: 1.2536... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19314... Training loss: 1.2365... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19315... Training loss: 1.1876... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19316... Training loss: 1.2538... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19317... Training loss: 1.2866... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19318... Training loss: 1.3525... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19319... Training loss: 1.3060... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19320... Training loss: 1.2724... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19321... Training loss: 1.2889... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19322... Training loss: 1.2721... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19323... Training loss: 1.1835... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19324... Training loss: 1.2697... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19325... Training loss: 1.2769... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19326... Training loss: 1.2843... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19327... Training loss: 1.2010... 0.0841 sec/batch
Epoch: 20/20... Training Step: 19328... Training loss: 1.2468... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19329... Training loss: 1.2357... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19330... Training loss: 1.2781... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19331... Training loss: 1.2614... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19332... Training loss: 1.2737... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19333... Training loss: 1.2066... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19334... Training loss: 1.2407... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19335... Training loss: 1.2933... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19336... Training loss: 1.2410... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19337... Training loss: 1.2357... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19338... Training loss: 1.2220... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19339... Training loss: 1.2701... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19340... Training loss: 1.2480... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19341... Training loss: 1.2696... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19342... Training loss: 1.2521... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19343... Training loss: 1.2389... 0.0800 sec/batch
Epoch: 20/20... Training Step: 19344... Training loss: 1.2153... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19345... Training loss: 1.2662... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19346... Training loss: 1.2568... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19347... Training loss: 1.2409... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19348... Training loss: 1.2711... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19349... Training loss: 1.2398... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19350... Training loss: 1.2479... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19351... Training loss: 1.2818... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19352... Training loss: 1.2517... 0.0742 sec/batch
Epoch: 20/20... Training Step: 19353... Training loss: 1.2579... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19354... Training loss: 1.2149... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19355... Training loss: 1.2554... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19356... Training loss: 1.2490... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19357... Training loss: 1.2491... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19358... Training loss: 1.2270... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19359... Training loss: 1.2868... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19360... Training loss: 1.2592... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19361... Training loss: 1.3301... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19362... Training loss: 1.3349... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19363... Training loss: 1.2485... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19364... Training loss: 1.2718... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19365... Training loss: 1.2712... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19366... Training loss: 1.2737... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19367... Training loss: 1.2891... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19368... Training loss: 1.2496... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19369... Training loss: 1.2479... 0.0806 sec/batch
Epoch: 20/20... Training Step: 19370... Training loss: 1.2483... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19371... Training loss: 1.3565... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19372... Training loss: 1.2681... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19373... Training loss: 1.2609... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19374... Training loss: 1.2549... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19375... Training loss: 1.2607... 0.0788 sec/batch
Epoch: 20/20... Training Step: 19376... Training loss: 1.2287... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19377... Training loss: 1.2578... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19378... Training loss: 1.3157... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19379... Training loss: 1.3762... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19380... Training loss: 1.3035... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19381... Training loss: 1.3288... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19382... Training loss: 1.3079... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19383... Training loss: 1.2691... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19384... Training loss: 1.2347... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19385... Training loss: 1.2274... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19386... Training loss: 1.2396... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19387... Training loss: 1.2581... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19388... Training loss: 1.2529... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19389... Training loss: 1.3574... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19390... Training loss: 1.2593... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19391... Training loss: 1.2386... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19392... Training loss: 1.2621... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19393... Training loss: 1.2155... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19394... Training loss: 1.2561... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19395... Training loss: 1.3196... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19396... Training loss: 1.2932... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19397... Training loss: 1.2164... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19398... Training loss: 1.2214... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19399... Training loss: 1.2583... 0.0795 sec/batch
Epoch: 20/20... Training Step: 19400... Training loss: 1.2046... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19401... Training loss: 1.2190... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19402... Training loss: 1.2606... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19403... Training loss: 1.2909... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19404... Training loss: 1.2196... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19405... Training loss: 1.2365... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19406... Training loss: 1.2549... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19407... Training loss: 1.2838... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19408... Training loss: 1.1758... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19409... Training loss: 1.2476... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19410... Training loss: 1.2679... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19411... Training loss: 1.2481... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19412... Training loss: 1.2474... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19413... Training loss: 1.2569... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19414... Training loss: 1.2894... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19415... Training loss: 1.2530... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19416... Training loss: 1.2038... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19417... Training loss: 1.1658... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19418... Training loss: 1.2085... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19419... Training loss: 1.1947... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19420... Training loss: 1.2357... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19421... Training loss: 1.3131... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19422... Training loss: 1.3172... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19423... Training loss: 1.2976... 0.0788 sec/batch
Epoch: 20/20... Training Step: 19424... Training loss: 1.2718... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19425... Training loss: 1.2561... 0.0825 sec/batch
Epoch: 20/20... Training Step: 19426... Training loss: 1.3766... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19427... Training loss: 1.3143... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19428... Training loss: 1.2348... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19429... Training loss: 1.2080... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19430... Training loss: 1.2239... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19431... Training loss: 1.3170... 0.0803 sec/batch
Epoch: 20/20... Training Step: 19432... Training loss: 1.2001... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19433... Training loss: 1.2482... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19434... Training loss: 1.2862... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19435... Training loss: 1.2806... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19436... Training loss: 1.3062... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19437... Training loss: 1.2388... 0.0810 sec/batch
Epoch: 20/20... Training Step: 19438... Training loss: 1.2353... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19439... Training loss: 1.2654... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19440... Training loss: 1.2698... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19441... Training loss: 1.2228... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19442... Training loss: 1.2327... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19443... Training loss: 1.1909... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19444... Training loss: 1.2927... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19445... Training loss: 1.1640... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19446... Training loss: 1.1819... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19447... Training loss: 1.2247... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19448... Training loss: 1.2266... 0.0789 sec/batch
Epoch: 20/20... Training Step: 19449... Training loss: 1.2828... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19450... Training loss: 1.2535... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19451... Training loss: 1.2386... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19452... Training loss: 1.2667... 0.0942 sec/batch
Epoch: 20/20... Training Step: 19453... Training loss: 1.2089... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19454... Training loss: 1.2141... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19455... Training loss: 1.3087... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19456... Training loss: 1.2987... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19457... Training loss: 1.2598... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19458... Training loss: 1.3015... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19459... Training loss: 1.1773... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19460... Training loss: 1.3143... 0.0739 sec/batch
Epoch: 20/20... Training Step: 19461... Training loss: 1.2863... 0.0820 sec/batch
Epoch: 20/20... Training Step: 19462... Training loss: 1.2426... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19463... Training loss: 1.3169... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19464... Training loss: 1.2667... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19465... Training loss: 1.2788... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19466... Training loss: 1.2893... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19467... Training loss: 1.2301... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19468... Training loss: 1.2397... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19469... Training loss: 1.2190... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19470... Training loss: 1.2102... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19471... Training loss: 1.2882... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19472... Training loss: 1.2705... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19473... Training loss: 1.2232... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19474... Training loss: 1.2461... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19475... Training loss: 1.2419... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19476... Training loss: 1.2642... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19477... Training loss: 1.1085... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19478... Training loss: 1.2817... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19479... Training loss: 1.2741... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19480... Training loss: 1.2478... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19481... Training loss: 1.2034... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19482... Training loss: 1.2543... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19483... Training loss: 1.2977... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19484... Training loss: 1.2579... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19485... Training loss: 1.3336... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19486... Training loss: 1.2169... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19487... Training loss: 1.2834... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19488... Training loss: 1.3703... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19489... Training loss: 1.3631... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19490... Training loss: 1.3357... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19491... Training loss: 1.2628... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19492... Training loss: 1.2602... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19493... Training loss: 1.2588... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19494... Training loss: 1.2626... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19495... Training loss: 1.2854... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19496... Training loss: 1.2180... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19497... Training loss: 1.2161... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19498... Training loss: 1.2555... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19499... Training loss: 1.2416... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19500... Training loss: 1.3040... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19501... Training loss: 1.2596... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19502... Training loss: 1.3328... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19503... Training loss: 1.3396... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19504... Training loss: 1.3029... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19505... Training loss: 1.2055... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19506... Training loss: 1.2527... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19507... Training loss: 1.2590... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19508... Training loss: 1.3387... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19509... Training loss: 1.2982... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19510... Training loss: 1.3432... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19511... Training loss: 1.2887... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19512... Training loss: 1.3498... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19513... Training loss: 1.3506... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19514... Training loss: 1.2418... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19515... Training loss: 1.3550... 0.0835 sec/batch
Epoch: 20/20... Training Step: 19516... Training loss: 1.3065... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19517... Training loss: 1.2859... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19518... Training loss: 1.2538... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19519... Training loss: 1.2943... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19520... Training loss: 1.3519... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19521... Training loss: 1.2552... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19522... Training loss: 1.2867... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19523... Training loss: 1.2215... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19524... Training loss: 1.2301... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19525... Training loss: 1.3385... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19526... Training loss: 1.2514... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19527... Training loss: 1.1938... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19528... Training loss: 1.2985... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19529... Training loss: 1.3336... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19530... Training loss: 1.2966... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19531... Training loss: 1.3135... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19532... Training loss: 1.2695... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19533... Training loss: 1.3362... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19534... Training loss: 1.3221... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19535... Training loss: 1.2797... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19536... Training loss: 1.2699... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19537... Training loss: 1.2247... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19538... Training loss: 1.3246... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19539... Training loss: 1.2660... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19540... Training loss: 1.2785... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19541... Training loss: 1.2552... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19542... Training loss: 1.2057... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19543... Training loss: 1.2678... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19544... Training loss: 1.2960... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19545... Training loss: 1.3599... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19546... Training loss: 1.2888... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19547... Training loss: 1.2569... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19548... Training loss: 1.2658... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19549... Training loss: 1.2948... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19550... Training loss: 1.2717... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19551... Training loss: 1.2779... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19552... Training loss: 1.2029... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19553... Training loss: 1.2083... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19554... Training loss: 1.1975... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19555... Training loss: 1.2599... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19556... Training loss: 1.2886... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19557... Training loss: 1.2784... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19558... Training loss: 1.2622... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19559... Training loss: 1.2362... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19560... Training loss: 1.3330... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19561... Training loss: 1.2133... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19562... Training loss: 1.2650... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19563... Training loss: 1.2756... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19564... Training loss: 1.3156... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19565... Training loss: 1.3037... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19566... Training loss: 1.3011... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19567... Training loss: 1.3164... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19568... Training loss: 1.3512... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19569... Training loss: 1.3058... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19570... Training loss: 1.2747... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19571... Training loss: 1.2580... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19572... Training loss: 1.3078... 0.0788 sec/batch
Epoch: 20/20... Training Step: 19573... Training loss: 1.3418... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19574... Training loss: 1.3442... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19575... Training loss: 1.2560... 0.0794 sec/batch
Epoch: 20/20... Training Step: 19576... Training loss: 1.3160... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19577... Training loss: 1.3196... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19578... Training loss: 1.2414... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19579... Training loss: 1.2727... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19580... Training loss: 1.3003... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19581... Training loss: 1.2720... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19582... Training loss: 1.2997... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19583... Training loss: 1.2972... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19584... Training loss: 1.2734... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19585... Training loss: 1.2614... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19586... Training loss: 1.3390... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19587... Training loss: 1.2519... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19588... Training loss: 1.2617... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19589... Training loss: 1.2422... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19590... Training loss: 1.2446... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19591... Training loss: 1.2807... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19592... Training loss: 1.3257... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19593... Training loss: 1.3125... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19594... Training loss: 1.2742... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19595... Training loss: 1.2788... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19596... Training loss: 1.2754... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19597... Training loss: 1.2047... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19598... Training loss: 1.2420... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19599... Training loss: 1.3709... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19600... Training loss: 1.2378... 0.0784 sec/batch
Epoch: 20/20... Training Step: 19601... Training loss: 1.3101... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19602... Training loss: 1.2988... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19603... Training loss: 1.2060... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19604... Training loss: 1.2784... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19605... Training loss: 1.2853... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19606... Training loss: 1.3172... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19607... Training loss: 1.3123... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19608... Training loss: 1.3576... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19609... Training loss: 1.2901... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19610... Training loss: 1.2034... 0.0799 sec/batch
Epoch: 20/20... Training Step: 19611... Training loss: 1.2655... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19612... Training loss: 1.2871... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19613... Training loss: 1.3022... 0.0780 sec/batch
Epoch: 20/20... Training Step: 19614... Training loss: 1.3066... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19615... Training loss: 1.2876... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19616... Training loss: 1.2500... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19617... Training loss: 1.2418... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19618... Training loss: 1.2776... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19619... Training loss: 1.2935... 0.0833 sec/batch
Epoch: 20/20... Training Step: 19620... Training loss: 1.3533... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19621... Training loss: 1.2983... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19622... Training loss: 1.2813... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19623... Training loss: 1.2732... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19624... Training loss: 1.1838... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19625... Training loss: 1.2146... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19626... Training loss: 1.2673... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19627... Training loss: 1.2471... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19628... Training loss: 1.2484... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19629... Training loss: 1.2950... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19630... Training loss: 1.2664... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19631... Training loss: 1.2913... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19632... Training loss: 1.2872... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19633... Training loss: 1.2813... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19634... Training loss: 1.2845... 0.0854 sec/batch
Epoch: 20/20... Training Step: 19635... Training loss: 1.2599... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19636... Training loss: 1.2920... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19637... Training loss: 1.2971... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19638... Training loss: 1.2976... 0.0740 sec/batch
Epoch: 20/20... Training Step: 19639... Training loss: 1.2486... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19640... Training loss: 1.2609... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19641... Training loss: 1.3016... 0.0782 sec/batch
Epoch: 20/20... Training Step: 19642... Training loss: 1.2237... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19643... Training loss: 1.1950... 0.0741 sec/batch
Epoch: 20/20... Training Step: 19644... Training loss: 1.2652... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19645... Training loss: 1.2481... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19646... Training loss: 1.2673... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19647... Training loss: 1.2347... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19648... Training loss: 1.2541... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19649... Training loss: 1.2817... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19650... Training loss: 1.3161... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19651... Training loss: 1.3443... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19652... Training loss: 1.2813... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19653... Training loss: 1.2370... 0.0758 sec/batch
Epoch: 20/20... Training Step: 19654... Training loss: 1.2804... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19655... Training loss: 1.2430... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19656... Training loss: 1.3267... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19657... Training loss: 1.3122... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19658... Training loss: 1.3220... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19659... Training loss: 1.2690... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19660... Training loss: 1.3291... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19661... Training loss: 1.2703... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19662... Training loss: 1.3275... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19663... Training loss: 1.3220... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19664... Training loss: 1.3073... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19665... Training loss: 1.3359... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19666... Training loss: 1.3716... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19667... Training loss: 1.3013... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19668... Training loss: 1.2605... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19669... Training loss: 1.2508... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19670... Training loss: 1.2293... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19671... Training loss: 1.2916... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19672... Training loss: 1.2561... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19673... Training loss: 1.2667... 0.0829 sec/batch
Epoch: 20/20... Training Step: 19674... Training loss: 1.2780... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19675... Training loss: 1.2952... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19676... Training loss: 1.3289... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19677... Training loss: 1.3153... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19678... Training loss: 1.2559... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19679... Training loss: 1.3183... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19680... Training loss: 1.2901... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19681... Training loss: 1.3369... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19682... Training loss: 1.2962... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19683... Training loss: 1.2263... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19684... Training loss: 1.2580... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19685... Training loss: 1.2852... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19686... Training loss: 1.3017... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19687... Training loss: 1.2958... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19688... Training loss: 1.2762... 0.0803 sec/batch
Epoch: 20/20... Training Step: 19689... Training loss: 1.2844... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19690... Training loss: 1.3890... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19691... Training loss: 1.3126... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19692... Training loss: 1.3152... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19693... Training loss: 1.3085... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19694... Training loss: 1.2506... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19695... Training loss: 1.3101... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19696... Training loss: 1.2716... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19697... Training loss: 1.2743... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19698... Training loss: 1.2900... 0.0739 sec/batch
Epoch: 20/20... Training Step: 19699... Training loss: 1.2509... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19700... Training loss: 1.3513... 0.0778 sec/batch
Epoch: 20/20... Training Step: 19701... Training loss: 1.3755... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19702... Training loss: 1.2773... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19703... Training loss: 1.2837... 0.0771 sec/batch
Epoch: 20/20... Training Step: 19704... Training loss: 1.2656... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19705... Training loss: 1.2769... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19706... Training loss: 1.2622... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19707... Training loss: 1.2943... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19708... Training loss: 1.3090... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19709... Training loss: 1.2790... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19710... Training loss: 1.2224... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19711... Training loss: 1.2393... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19712... Training loss: 1.2833... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19713... Training loss: 1.2358... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19714... Training loss: 1.2717... 0.0827 sec/batch
Epoch: 20/20... Training Step: 19715... Training loss: 1.2723... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19716... Training loss: 1.3083... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19717... Training loss: 1.3091... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19718... Training loss: 1.3504... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19719... Training loss: 1.3197... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19720... Training loss: 1.2947... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19721... Training loss: 1.2804... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19722... Training loss: 1.3115... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19723... Training loss: 1.2854... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19724... Training loss: 1.3216... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19725... Training loss: 1.2950... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19726... Training loss: 1.3345... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19727... Training loss: 1.2847... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19728... Training loss: 1.2727... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19729... Training loss: 1.2900... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19730... Training loss: 1.2071... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19731... Training loss: 1.3215... 0.0772 sec/batch
Epoch: 20/20... Training Step: 19732... Training loss: 1.2706... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19733... Training loss: 1.3010... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19734... Training loss: 1.3353... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19735... Training loss: 1.2944... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19736... Training loss: 1.3180... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19737... Training loss: 1.3190... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19738... Training loss: 1.3779... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19739... Training loss: 1.3347... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19740... Training loss: 1.2240... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19741... Training loss: 1.2887... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19742... Training loss: 1.2505... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19743... Training loss: 1.3431... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19744... Training loss: 1.2348... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19745... Training loss: 1.2922... 0.0781 sec/batch
Epoch: 20/20... Training Step: 19746... Training loss: 1.3161... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19747... Training loss: 1.2380... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19748... Training loss: 1.2285... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19749... Training loss: 1.2336... 0.0735 sec/batch
Epoch: 20/20... Training Step: 19750... Training loss: 1.2506... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19751... Training loss: 1.2339... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19752... Training loss: 1.2182... 0.0761 sec/batch
Epoch: 20/20... Training Step: 19753... Training loss: 1.2268... 0.0775 sec/batch
Epoch: 20/20... Training Step: 19754... Training loss: 1.3044... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19755... Training loss: 1.2294... 0.0755 sec/batch
Epoch: 20/20... Training Step: 19756... Training loss: 1.2701... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19757... Training loss: 1.2275... 0.0796 sec/batch
Epoch: 20/20... Training Step: 19758... Training loss: 1.1899... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19759... Training loss: 1.2202... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19760... Training loss: 1.2979... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19761... Training loss: 1.3590... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19762... Training loss: 1.2792... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19763... Training loss: 1.2686... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19764... Training loss: 1.2656... 0.0783 sec/batch
Epoch: 20/20... Training Step: 19765... Training loss: 1.2486... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19766... Training loss: 1.3381... 0.0792 sec/batch
Epoch: 20/20... Training Step: 19767... Training loss: 1.2670... 0.0785 sec/batch
Epoch: 20/20... Training Step: 19768... Training loss: 1.3035... 0.0751 sec/batch
Epoch: 20/20... Training Step: 19769... Training loss: 1.3747... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19770... Training loss: 1.3191... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19771... Training loss: 1.2903... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19772... Training loss: 1.2301... 0.0754 sec/batch
Epoch: 20/20... Training Step: 19773... Training loss: 1.2825... 0.0767 sec/batch
Epoch: 20/20... Training Step: 19774... Training loss: 1.3179... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19775... Training loss: 1.2942... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19776... Training loss: 1.2737... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19777... Training loss: 1.3245... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19778... Training loss: 1.3378... 0.0769 sec/batch
Epoch: 20/20... Training Step: 19779... Training loss: 1.3006... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19780... Training loss: 1.3816... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19781... Training loss: 1.2629... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19782... Training loss: 1.3091... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19783... Training loss: 1.2890... 0.0766 sec/batch
Epoch: 20/20... Training Step: 19784... Training loss: 1.3265... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19785... Training loss: 1.2646... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19786... Training loss: 1.2639... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19787... Training loss: 1.2878... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19788... Training loss: 1.3651... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19789... Training loss: 1.2826... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19790... Training loss: 1.2985... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19791... Training loss: 1.3397... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19792... Training loss: 1.3125... 0.0752 sec/batch
Epoch: 20/20... Training Step: 19793... Training loss: 1.3266... 0.0770 sec/batch
Epoch: 20/20... Training Step: 19794... Training loss: 1.2890... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19795... Training loss: 1.2635... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19796... Training loss: 1.2631... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19797... Training loss: 1.2842... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19798... Training loss: 1.2163... 0.0742 sec/batch
Epoch: 20/20... Training Step: 19799... Training loss: 1.2881... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19800... Training loss: 1.2311... 0.0764 sec/batch
Epoch: 20/20... Training Step: 19801... Training loss: 1.2409... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19802... Training loss: 1.2970... 0.0753 sec/batch
Epoch: 20/20... Training Step: 19803... Training loss: 1.3036... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19804... Training loss: 1.2811... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19805... Training loss: 1.2557... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19806... Training loss: 1.3024... 0.0743 sec/batch
Epoch: 20/20... Training Step: 19807... Training loss: 1.2544... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19808... Training loss: 1.2465... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19809... Training loss: 1.2800... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19810... Training loss: 1.2742... 0.0748 sec/batch
Epoch: 20/20... Training Step: 19811... Training loss: 1.2655... 0.0750 sec/batch
Epoch: 20/20... Training Step: 19812... Training loss: 1.2543... 0.0749 sec/batch
Epoch: 20/20... Training Step: 19813... Training loss: 1.3442... 0.0763 sec/batch
Epoch: 20/20... Training Step: 19814... Training loss: 1.3038... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19815... Training loss: 1.3371... 0.0747 sec/batch
Epoch: 20/20... Training Step: 19816... Training loss: 1.3386... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19817... Training loss: 1.3033... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19818... Training loss: 1.4497... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19819... Training loss: 1.2744... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19820... Training loss: 1.3663... 0.0762 sec/batch
Epoch: 20/20... Training Step: 19821... Training loss: 1.3898... 0.0744 sec/batch
Epoch: 20/20... Training Step: 19822... Training loss: 1.2413... 0.0856 sec/batch
Epoch: 20/20... Training Step: 19823... Training loss: 1.3249... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19824... Training loss: 1.3240... 0.0746 sec/batch
Epoch: 20/20... Training Step: 19825... Training loss: 1.3428... 0.0779 sec/batch
Epoch: 20/20... Training Step: 19826... Training loss: 1.3388... 0.0768 sec/batch
Epoch: 20/20... Training Step: 19827... Training loss: 1.3196... 0.0760 sec/batch
Epoch: 20/20... Training Step: 19828... Training loss: 1.2420... 0.0773 sec/batch
Epoch: 20/20... Training Step: 19829... Training loss: 1.2324... 0.0745 sec/batch
Epoch: 20/20... Training Step: 19830... Training loss: 1.3946... 0.0776 sec/batch
Epoch: 20/20... Training Step: 19831... Training loss: 1.2814... 0.0787 sec/batch
Epoch: 20/20... Training Step: 19832... Training loss: 1.3085... 0.0757 sec/batch
Epoch: 20/20... Training Step: 19833... Training loss: 1.3146... 0.0786 sec/batch
Epoch: 20/20... Training Step: 19834... Training loss: 1.2476... 0.0774 sec/batch
Epoch: 20/20... Training Step: 19835... Training loss: 1.2415... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19836... Training loss: 1.2848... 0.0759 sec/batch
Epoch: 20/20... Training Step: 19837... Training loss: 1.2969... 0.0765 sec/batch
Epoch: 20/20... Training Step: 19838... Training loss: 1.1900... 0.0777 sec/batch
Epoch: 20/20... Training Step: 19839... Training loss: 1.2355... 0.0756 sec/batch
Epoch: 20/20... Training Step: 19840... Training loss: 1.3409... 0.0748 sec/batch
Read up on saving and loading checkpoints here: https://www.tensorflow.org/programmers_guide/variables
In [48]:
tf.train.get_checkpoint_state('checkpoints')
Out[48]:
model_checkpoint_path: "checkpoints/i19840_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i1000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i1200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i1400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i1600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i1800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i2000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i2200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i2400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i2600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i2800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i3000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i3200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i3400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i3600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i3800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i4000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i4200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i4400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i4600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i4800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i5000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i5200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i5400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i5600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i5800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i6000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i6200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i6400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i6600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i6800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i7000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i7200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i7400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i7600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i7800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i8000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i8200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i8400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i8600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i8800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i9000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i9200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i9400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i9600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i9800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i10000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i10200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i10400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i10600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i10800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i11000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i11200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i11400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i11600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i11800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i12000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i12200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i12400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i12600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i12800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i13000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i13200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i13400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i13600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i13800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i14000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i14200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i14400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i14600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i14800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i15000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i15200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i15400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i15600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i15800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i16000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i16200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i16400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i16600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i16800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i17000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i17200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i17400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i17600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i17800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i18000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i18200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i18400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i18600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i18800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19000_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19200_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19400_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19600_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19800_l256.ckpt"
all_model_checkpoint_paths: "checkpoints/i19840_l256.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 [49]:
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 [57]:
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), top_n=1)
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 [58]:
tf.train.latest_checkpoint('checkpoints')
Out[58]:
'checkpoints/i19840_l256.ckpt'
In [63]:
checkpoint = 'checkpoints/i200_l256.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Anna ")
print(samp)
INFO:tensorflow:Restoring parameters from checkpoints/i200_l256.ckpt
Anna tetir iti s io see at oretee, wet ood titisde thar wang ao he an on hes, ant tar oad ote tinde hh sorised whe th ather ha sar on te han an hir te te are the sant tate an tot the tio ter the tho setang we hisd oo io sestat ot aod oo tera herte at hire hh tared ar on ore wese toe har hat ot ot tite se an ho ter and wot atere ang in toe whit ti tee sond won ther hos on hart tirtes at arterat hong ood tirer tores he teer he tont an on tond toe we ate wo asete te hingt on onrinn an tor oro ho she sot hert ar as he ther the ar te ase an teres
saretir ar ater ote hhes the sethand oa hhas atte tet ar here wh thet tha the hori he an oriti thand tir tho she tond ar oot ar one an tint at tot as on tin an it asd toe tar toe therat itin wo titethes hhe an the and
tios wos hon an at oo ind tha hare ot tie ta sar atis to te wan oe he the ther teis tasd ater at het ant tee an ot at eor he ared, har we oto ha tarin he want is te tar the hartis irise, tar he tio it has astert har hor tete ar oathe aot ha
In [64]:
checkpoint = 'checkpoints/i600_l256.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Anna ")
print(samp)
INFO:tensorflow:Restoring parameters from checkpoints/i600_l256.ckpt
Anna to
serent of tho hame wint on hore, at his houd hersedes of she hadene to aster of
ithel so sores has at ther.
"Ot thas as the perice to ate., she selant to sele and somine at the comnet to ang he mosh of her, and thing to see hereresed heassing and. Ind so he mict he tour hit of thound.
The sat the
shrowhing to her some her, the tore and steress home the heas at the porenting on the tor tink, and she te cerle allo the serant will hers he houre sitt to cone that her and thing to tor to hin thet sald, thit were anere an then hem sis it a shand the san onesting hor, wastele.
"
"Whan he hand so that at her sold to his.
The congtel that at that there sish ot the paringeress, that. "I so wis of is thit soure ho shat on as how.
"Yer, sho chith, and tome to tall to an of time, werens on
that hor her. I said her tore, and what he saide the ther the pesctess on angises to to and ald he her wis shich, he here her sead onder of that thit his
inther anders thate
sithar of the
theming ot her the
In [65]:
checkpoint = 'checkpoints/i1200_l256.ckpt'
samp = sample(checkpoint, 1000, lstm_size, len(vocab), prime="Anna ")
print(samp)
INFO:tensorflow:Restoring parameters from checkpoints/i1200_l256.ckpt
Anna the said, went a shing a would her.
"I, was so and whith athel the was ant to the sermest the sime. And to him,, that's hould said.
"It's said."
At the
port on the cars of that was to her with the
right on the seis on thre tomess with a posting all the
pease and said to better and shought, sand a word had he was
stand and as and ald hig, had she soud thit he was tere him of his had and
soun a the posich to but and, which a wond a certen the morst of shought, and she that the mading the fort the sace to the souch to the mushitian,
were her had he had bees to beed the ceart was her, the caming the componing the concenst of at a sting their was the mar ous the
prasest onterssands, so dand which had said, the was samine a tending of a mest tithing while had hid hears and was all that was at that he was not a momen ant aroo a tompring troth the seed to the simp on thit a corloms, a talked him for the
ment the pare tare was shat whith so hinder of her all shark at a shat of his bether. He
In [59]:
checkpoint = tf.train.latest_checkpoint('checkpoints')
samp = sample(checkpoint, 2000, lstm_size, len(vocab), prime="Anna ")
print(samp)
INFO:tensorflow:Restoring parameters from checkpoints/i19840_l256.ckpt
Anna Arkadyevna, who had been despered things of his horse, but this was
a letter, that had at once
whispled to her to see her from a stratging and sent to secret of the
prople. "Well, was it was an obdence of a starmish and here,
that is, he can't believe that I
say to him as I don't know that he has not
seen a long while."
"No, it's a short," she thought.
"One must hear herself the more incidematiant thoushes," she said, looking at him. "I don't know that I was taking the munching, and this
communicate with tenderness, but I don't know," she said, was timidly fruming.
Seeing that she carefully with him.
"Why, then I'm this sort of all sorrily."
She spoke of the sense of the same friends. But the morning had several some place to stay the matters of
hope, and was strangely as a smile of state of a little pleasure.
"It's told this are alone?"
"It's not a position of humiliation without her and more than ever she.
And there's no making his mother, and the mistake.... It's to see you to suppose in the
discovery."
The point of spite of the doubts of the place that he sat down on
his feeling with his wife to have him to get over
a serious smile, and when
to take a struck of the solution which is trying
to come in the decision of that words in his hings.
Chapter 18
Sergey Ivanovitch standed at the sort of terror to be suddenly sent on a line when they were sitting to the sacrament in the place, and went off on the repression. And an accord to sank till anyone was the province, that he wasted in the sense of the position of the
words about the prince.
"No, this," she asked.
"I'm not and saying and all any service. To stay anyone with the princess, but you're a peace and all sides."
"Well, this is the middle of
your marshal!
They would be seen him, and how's nothing impossible to make her to think,
and I wouldn be
a morning. I shall not
come to my wife. I cast all to be alone with you and to be to say
it a little sort of
part, and how dear it," he said; "but I ha
Content source: adrianstaniec/deep-learning
Similar notebooks: