notebook.community

Edit and run

Language Translation

In this project, you’re going to take a peek into the realm of neural network machine translation. You’ll be training a sequence to sequence model on a dataset of English and French sentences that can translate new sentences from English to French.

Get the Data

Since translating the whole language of English to French will take lots of time to train, we have provided you with a small portion of the English corpus.





In [1]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper
import problem_unittests as tests

source_path = 'data/small_vocab_en'
target_path = 'data/small_vocab_fr'
source_text = helper.load_data(source_path)
target_text = helper.load_data(target_path)

Explore the Data

Play around with view_sentence_range to view different parts of the data.





In [2]:



    


view_sentence_range = (0, 10)

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

print('Dataset Stats')
print('Roughly the number of unique words: {}'.format(len({word: None for word in source_text.split()})))

sentences = source_text.split('\n')
word_counts = [len(sentence.split()) for sentence in sentences]
print('Number of sentences: {}'.format(len(sentences)))
print('Average number of words in a sentence: {}'.format(np.average(word_counts)))

print()
print('English sentences {} to {}:'.format(*view_sentence_range))
print('\n'.join(source_text.split('\n')[view_sentence_range[0]:view_sentence_range[1]]))
print()
print('French sentences {} to {}:'.format(*view_sentence_range))
print('\n'.join(target_text.split('\n')[view_sentence_range[0]:view_sentence_range[1]]))



    


















    






Dataset Stats
Roughly the number of unique words: 227
Number of sentences: 137861
Average number of words in a sentence: 13.225277634719028

English sentences 0 to 10:
new jersey is sometimes quiet during autumn , and it is snowy in april .
the united states is usually chilly during july , and it is usually freezing in november .
california is usually quiet during march , and it is usually hot in june .
the united states is sometimes mild during june , and it is cold in september .
your least liked fruit is the grape , but my least liked is the apple .
his favorite fruit is the orange , but my favorite is the grape .
paris is relaxing during december , but it is usually chilly in july .
new jersey is busy during spring , and it is never hot in march .
our least liked fruit is the lemon , but my least liked is the grape .
the united states is sometimes busy during january , and it is sometimes warm in november .

French sentences 0 to 10:
new jersey est parfois calme pendant l' automne , et il est neigeux en avril .
les états-unis est généralement froid en juillet , et il gèle habituellement en novembre .
california est généralement calme en mars , et il est généralement chaud en juin .
les états-unis est parfois légère en juin , et il fait froid en septembre .
votre moins aimé fruit est le raisin , mais mon moins aimé est la pomme .
son fruit préféré est l'orange , mais mon préféré est le raisin .
paris est relaxant en décembre , mais il est généralement froid en juillet .
new jersey est occupé au printemps , et il est jamais chaude en mars .
notre fruit est moins aimé le citron , mais mon moins aimé est le raisin .
les états-unis est parfois occupé en janvier , et il est parfois chaud en novembre .

Implement Preprocessing Function

Text to Word Ids

As you did with other RNNs, you must turn the text into a number so the computer can understand it. In the function text_to_ids(), you'll turn source_text and target_text from words to ids. However, you need to add the <EOS> word id at the end of each sentence from target_text. This will help the neural network predict when the sentence should end.

You can get the <EOS> word id by doing:

target_vocab_to_int['<EOS>']

You can get other word ids using source_vocab_to_int and target_vocab_to_int.





In [3]:



    


def text_to_ids(source_text, target_text, source_vocab_to_int, target_vocab_to_int):
    """
    Convert source and target text to proper word ids
    :param source_text: String that contains all the source text.
    :param target_text: String that contains all the target text.
    :param source_vocab_to_int: Dictionary to go from the source words to an id
    :param target_vocab_to_int: Dictionary to go from the target words to an id
    :return: A tuple of lists (source_id_text, target_id_text)
    """
    # TODO: Implement Function
#     print(target_vocab_to_int["EOS"])
#     print(target_vocab_to_int)
#     print(target_vocab_to_int['<EOS>'])
    source_id_text = [ [source_vocab_to_int.get(word,source_vocab_to_int['<UNK>']) for word in sentence.split()]for sentence in source_text.split('\n')  ]
    target_id_text = [ [target_vocab_to_int.get(word,target_vocab_to_int['<UNK>']) for word in sentence.split()] + [target_vocab_to_int['<EOS>']] for sentence in target_text.split('\n')  ]
#     print(target_id_text)
    return source_id_text, target_id_text

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_text_to_ids(text_to_ids)



    


















    






Tests Passed

Preprocess all the data and save it

Running the code cell below will preprocess all the data and save it to file.





In [4]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
helper.preprocess_and_save_data(source_path, target_path, text_to_ids)

Check Point

This is your first checkpoint. If you ever decide to come back to this notebook or have to restart the notebook, you can start from here. The preprocessed data has been saved to disk.





In [5]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np
import helper

(source_int_text, target_int_text), (source_vocab_to_int, target_vocab_to_int), _ = helper.load_preprocess()

Check the Version of TensorFlow and Access to GPU

This will check to make sure you have the correct version of TensorFlow and access to a GPU





In [6]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) in [LooseVersion('1.0.0'), LooseVersion('1.0.1')], 'This project requires TensorFlow version 1.0  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))



    


















    






TensorFlow Version: 1.0.0
Default GPU Device: /gpu:0

Build the Neural Network

You'll build the components necessary to build a Sequence-to-Sequence model by implementing the following functions below:

  • model_inputs
  • process_decoding_input
  • encoding_layer
  • decoding_layer_train
  • decoding_layer_infer
  • decoding_layer
  • seq2seq_model

Input

Implement the model_inputs() function to create TF Placeholders for the Neural Network. It should create the following placeholders:

  • Input text placeholder named "input" using the TF Placeholder name parameter with rank 2.
  • Targets placeholder with rank 2.
  • Learning rate placeholder with rank 0.
  • Keep probability placeholder named "keep_prob" using the TF Placeholder name parameter with rank 0.

Return the placeholders in the following the tuple (Input, Targets, Learing Rate, Keep Probability)





In [7]:



    


def model_inputs():
    """
    Create TF Placeholders for input, targets, and learning rate.
    :return: Tuple (input, targets, learning rate, keep probability)
    """
    # TODO: Implement Function
    inputs = tf.placeholder(tf.int32,[None,None],name='input')
    targets = tf.placeholder(tf.int32,[None,None])
    lr = tf.placeholder(tf.float32)
    keep_prob =  tf.placeholder(tf.float32,None,name='keep_prob')
    return (inputs, targets, lr, keep_prob)

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)



    


















    






Tests Passed

Process Decoding Input

Implement process_decoding_input using TensorFlow to remove the last word id from each batch in target_data and concat the GO ID to the beginning of each batch.





In [8]:



    


def process_decoding_input(target_data, target_vocab_to_int, batch_size):
    """
    Preprocess target data for decoding
    :param target_data: Target Placeholder
    :param target_vocab_to_int: Dictionary to go from the target words to an id
    :param batch_size: Batch Size
    :return: Preprocessed target data
    """
    # target_data has shape [batch_size, seq_length]
    # TODO: Implement Function
#     batch_size,_ = target_data.shape.as_list()
#     target_shape = tf.shape(target_data)
#     print("target_shape is:",target_shape)
#     print(target_data)
#     batch_size = target_shape[0]
#     print(batch_size)
    
    ending = tf.strided_slice(target_data, [0, 0], [batch_size, -1], [1, 1])
#     print(ending)
    dec_input = tf.concat([tf.fill([batch_size, 1], target_vocab_to_int['<GO>']), ending], 1)
#     shape_input = tf.reshape(dec_input,target_shape)
#     print(dec_input)
#     print(shape_input)
#     print(shape_input)
#     return shape_input
    return dec_input  
"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_process_decoding_input(process_decoding_input)



    


















    






Tests Passed

Encoding

Implement encoding_layer() to create a Encoder RNN layer using tf.nn.dynamic_rnn().





In [9]:



    


def encoding_layer(rnn_inputs, rnn_size, num_layers, keep_prob):
    """
    Create encoding layer
    :param rnn_inputs: Inputs for the RNN
    :param rnn_size: RNN Size
    :param num_layers: Number of layers
    :param keep_prob: Dropout keep probability
    :return: RNN state
    """
    # TODO: Implement Function
    cell = tf.contrib.rnn.BasicLSTMCell(rnn_size)
    drop = tf.contrib.rnn.DropoutWrapper(cell, output_keep_prob=keep_prob)
    enc_cell = tf.contrib.rnn.MultiRNNCell([drop] * num_layers)
    _, enc_state = tf.nn.dynamic_rnn(enc_cell, rnn_inputs, dtype=tf.float32)
    return enc_state

"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_encoding_layer(encoding_layer)



    


















    






Tests Passed

Decoding - Training

Create training logits using tf.contrib.seq2seq.simple_decoder_fn_train() and tf.contrib.seq2seq.dynamic_rnn_decoder(). Apply the output_fn to the tf.contrib.seq2seq.dynamic_rnn_decoder() outputs.





In [10]:



    


def decoding_layer_train(encoder_state, dec_cell, dec_embed_input, sequence_length, decoding_scope,
                         output_fn, keep_prob):
    """
    Create a decoding layer for training
    :param encoder_state: Encoder State
    :param dec_cell: Decoder RNN Cell
    :param dec_embed_input: Decoder embedded input
    :param sequence_length: Sequence Length
    :param decoding_scope: TenorFlow Variable Scope for decoding
    :param output_fn: Function to apply the output layer
    :param keep_prob: Dropout keep probability
    :return: Train Logits
    """
    # TODO: Implement Function
    train_decoder_fn = tf.contrib.seq2seq.simple_decoder_fn_train(encoder_state)
#     drop = tf.contrib.rnn.DropoutWrapper(dec_cell, output_keep_prob=keep_prob)
    #  (outputs, final_state, final_context_state)
    train_pred, _, _ = tf.contrib.seq2seq.dynamic_rnn_decoder(
        dec_cell, train_decoder_fn, dec_embed_input, sequence_length, scope=decoding_scope)
    
    # Apply output function
    train_logits =  output_fn(train_pred)
    return train_logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_decoding_layer_train(decoding_layer_train)



    


















    






Tests Passed

Decoding - Inference

Create inference logits using tf.contrib.seq2seq.simple_decoder_fn_inference() and tf.contrib.seq2seq.dynamic_rnn_decoder().





In [11]:



    


def decoding_layer_infer(encoder_state, dec_cell, dec_embeddings, start_of_sequence_id, end_of_sequence_id,
                         maximum_length, vocab_size, decoding_scope, output_fn, keep_prob):
    """
    Create a decoding layer for inference
    :param encoder_state: Encoder state
    :param dec_cell: Decoder RNN Cell
    :param dec_embeddings: Decoder embeddings
    :param start_of_sequence_id: GO ID
    :param end_of_sequence_id: EOS Id
    :param maximum_length: The maximum allowed time steps to decode
    :param vocab_size: Size of vocabulary
    :param decoding_scope: TensorFlow Variable Scope for decoding
    :param output_fn: Function to apply the output layer
    :param keep_prob: Dropout keep probability
    :return: Inference Logits
    """
    # TODO: Implement Function
     # Inference Decoder
    infer_decoder_fn = tf.contrib.seq2seq.simple_decoder_fn_inference(
        output_fn, encoder_state, dec_embeddings, start_of_sequence_id, end_of_sequence_id, 
        maximum_length, vocab_size)
#     drop = tf.contrib.rnn.DropoutWrapper(dec_cell, output_keep_prob=keep_prob)
    inference_logits, _, _ = tf.contrib.seq2seq.dynamic_rnn_decoder(dec_cell, infer_decoder_fn, scope=decoding_scope)
    return inference_logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_decoding_layer_infer(decoding_layer_infer)



    


















    






Tests Passed

Build the Decoding Layer

Implement decoding_layer() to create a Decoder RNN layer.

  • Create RNN cell for decoding using rnn_size and num_layers.
  • Create the output fuction using lambda to transform it's input, logits, to class logits.
  • Use the your decoding_layer_train(encoder_state, dec_cell, dec_embed_input, sequence_length, decoding_scope, output_fn, keep_prob) function to get the training logits.
  • Use your decoding_layer_infer(encoder_state, dec_cell, dec_embeddings, start_of_sequence_id, end_of_sequence_id, maximum_length, vocab_size, decoding_scope, output_fn, keep_prob) function to get the inference logits.

Note: You'll need to use tf.variable_scope to share variables between training and inference.





In [30]:



    


def decoding_layer(dec_embed_input, dec_embeddings, encoder_state, vocab_size, sequence_length, rnn_size,
                   num_layers, target_vocab_to_int, keep_prob):
    """
    Create decoding layer
    :param dec_embed_input: Decoder embedded input
    :param dec_embeddings: Decoder embeddings
    :param encoder_state: The encoded state
    :param vocab_size: Size of vocabulary
    :param sequence_length: Sequence Length
    :param rnn_size: RNN Size
    :param num_layers: Number of layers
    :param target_vocab_to_int: Dictionary to go from the target words to an id
    :param keep_prob: Dropout keep probability
    :return: Tuple of (Training Logits, Inference Logits)
    """
    # TODO: Implement Function
    cell = tf.contrib.rnn.BasicLSTMCell(rnn_size)
    drop = tf.contrib.rnn.DropoutWrapper(cell, output_keep_prob=keep_prob)
    dec_cell = tf.contrib.rnn.MultiRNNCell([drop] * num_layers)

#     print(sequence_length)

  
    maximum_length = sequence_length

    with tf.variable_scope("decoding") as decoding_scope:
        output_fn = lambda x: tf.contrib.layers.fully_connected(x, vocab_size, None, scope=decoding_scope,
                                                               weights_initializer=tf.truncated_normal_initializer(mean=0.0, stddev=0.1),
                                                       biases_initializer=tf.zeros_initializer())
        train_logits = decoding_layer_train(encoder_state,dec_cell,dec_embed_input,
                                            sequence_length,decoding_scope,output_fn,keep_prob)
        decoding_scope.reuse_variables()
        inference_logits = decoding_layer_infer(encoder_state, dec_cell, dec_embeddings, 
                                                target_vocab_to_int['<GO>'], target_vocab_to_int["<EOS>"],
                         maximum_length, vocab_size, decoding_scope, output_fn, keep_prob)
    return train_logits, inference_logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_decoding_layer(decoding_layer)



    


















    






Tests Passed

Build the Neural Network

Apply the functions you implemented above to:

  • Apply embedding to the input data for the encoder.
  • Encode the input using your encoding_layer(rnn_inputs, rnn_size, num_layers, keep_prob).
  • Process target data using your process_decoding_input(target_data, target_vocab_to_int, batch_size) function.
  • Apply embedding to the target data for the decoder.
  • Decode the encoded input using your decoding_layer(dec_embed_input, dec_embeddings, encoder_state, vocab_size, sequence_length, rnn_size, num_layers, target_vocab_to_int, keep_prob).




In [31]:



    


def seq2seq_model(input_data, target_data, keep_prob, batch_size, sequence_length, source_vocab_size, target_vocab_size,
                  enc_embedding_size, dec_embedding_size, rnn_size, num_layers, target_vocab_to_int):
    """
    Build the Sequence-to-Sequence part of the neural network
    :param input_data: Input placeholder
    :param target_data: Target placeholder
    :param keep_prob: Dropout keep probability placeholder
    :param batch_size: Batch Size
    :param sequence_length: Sequence Length
    :param source_vocab_size: Source vocabulary size
    :param target_vocab_size: Target vocabulary size
    :param enc_embedding_size: Decoder embedding size
    :param dec_embedding_size: Encoder embedding size
    :param rnn_size: RNN Size
    :param num_layers: Number of layers
    :param target_vocab_to_int: Dictionary to go from the target words to an id
    :return: Tuple of (Training Logits, Inference Logits)
    """
    # TODO: Implement Function
    # Encoder Embedding
    enc_embeddings = tf.Variable(tf.truncated_normal([source_vocab_size, enc_embedding_size],mean=0.0, stddev=0.1))
    enc_embed_input = tf.nn.embedding_lookup(enc_embeddings, input_data)
    # decoder Embedding
    dec_embeddings = tf.Variable(tf.truncated_normal([target_vocab_size, dec_embedding_size],mean=0.0, stddev=0.1))
    dec_input = process_decoding_input(target_data, target_vocab_to_int, batch_size)
    dec_embed_input = tf.nn.embedding_lookup(dec_embeddings, dec_input)
    
    
    enc_state = encoding_layer(enc_embed_input,rnn_size,num_layers,keep_prob)
    

    return decoding_layer(dec_embed_input,dec_embeddings,enc_state,target_vocab_size,sequence_length,
                             rnn_size, num_layers, target_vocab_to_int, keep_prob)


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_seq2seq_model(seq2seq_model)



    


















    






Tests Passed

Neural Network Training

Hyperparameters

Tune the following parameters:

  • Set epochs to the number of epochs.
  • Set batch_size to the batch size.
  • Set rnn_size to the size of the RNNs.
  • Set num_layers to the number of layers.
  • Set encoding_embedding_size to the size of the embedding for the encoder.
  • Set decoding_embedding_size to the size of the embedding for the decoder.
  • Set learning_rate to the learning rate.
  • Set keep_probability to the Dropout keep probability




In [77]:



    


# Number of Epochs
epochs = 10
# Batch Size
batch_size = 1000
# RNN Size
rnn_size = 128
# Number of Layers
num_layers = 2
# Embedding Size
encoding_embedding_size = 100
decoding_embedding_size = 100
# Learning Rate
learning_rate = 0.01
# Dropout Keep Probability
keep_probability = 0.5

Build the Graph

Build the graph using the neural network you implemented.





In [78]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
save_path = 'checkpoints/dev'
(source_int_text, target_int_text), (source_vocab_to_int, target_vocab_to_int), _ = helper.load_preprocess()
# input_shape [batch_size, seq_len]
max_source_sentence_length = max([len(sentence) for sentence in source_int_text])

train_graph = tf.Graph()
with train_graph.as_default():
    input_data, targets, lr, keep_prob = model_inputs()
    sequence_length = tf.placeholder_with_default(max_source_sentence_length, None, name='sequence_length')
    input_shape = tf.shape(input_data)
    
    train_logits, inference_logits = seq2seq_model(
        tf.reverse(input_data, [-1]), targets, keep_prob, batch_size, sequence_length, len(source_vocab_to_int), len(target_vocab_to_int),
        encoding_embedding_size, decoding_embedding_size, rnn_size, num_layers, target_vocab_to_int)

    tf.identity(inference_logits, 'logits')
    with tf.name_scope("optimization"):
        # Loss function
        cost = tf.contrib.seq2seq.sequence_loss(
            train_logits,
            targets,
            tf.ones([input_shape[0], sequence_length]))

        # Optimizer
        optimizer = tf.train.AdamOptimizer(lr)

        # Gradient Clipping
        gradients = optimizer.compute_gradients(cost)
        capped_gradients = [(tf.clip_by_value(grad, -1., 1.), var) for grad, var in gradients if grad is not None]
        train_op = optimizer.apply_gradients(capped_gradients)

Train

Train the neural network on the preprocessed data. If you have a hard time getting a good loss, check the forums to see if anyone is having the same problem.





In [79]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import time

def get_accuracy(target, logits):
    """
    Calculate accuracy
    """
    max_seq = max(target.shape[1], logits.shape[1])
    if max_seq - target.shape[1]:
        target = np.pad(
            target,
            [(0,0),(0,max_seq - target.shape[1])],
            'constant')
    if max_seq - logits.shape[1]:
        logits = np.pad(
            logits,
            [(0,0),(0,max_seq - logits.shape[1]), (0,0)],
            'constant')

    return np.mean(np.equal(target, np.argmax(logits, 2)))

train_source = source_int_text[batch_size:]
train_target = target_int_text[batch_size:]

valid_source = helper.pad_sentence_batch(source_int_text[:batch_size])
valid_target = helper.pad_sentence_batch(target_int_text[:batch_size])

with tf.Session(graph=train_graph) as sess:
    sess.run(tf.global_variables_initializer())

    for epoch_i in range(epochs):
        for batch_i, (source_batch, target_batch) in enumerate(
                helper.batch_data(train_source, train_target, batch_size)):
            start_time = time.time()
            
#             print(source_batch)
#             print(target_batch.shape[1])
        
            _, loss = sess.run(
                [train_op, cost],
                {input_data: source_batch,
                 targets: target_batch,
                 lr: learning_rate,
                 sequence_length: target_batch.shape[1],
                 keep_prob: keep_probability})
            
            batch_train_logits = sess.run(
                inference_logits,
                {input_data: source_batch, keep_prob: 1.0})
            batch_valid_logits = sess.run(
                inference_logits,
                {input_data: valid_source, keep_prob: 1.0})
                
            train_acc = get_accuracy(target_batch, batch_train_logits)
            valid_acc = get_accuracy(np.array(valid_target), batch_valid_logits)
            end_time = time.time()
            print('Epoch {:>3} Batch {:>4}/{} - Train Accuracy: {:>6.3f}, Validation Accuracy: {:>6.3f}, Loss: {:>6.3f}'
                  .format(epoch_i, batch_i, len(source_int_text) // batch_size, train_acc, valid_acc, loss))

    # Save Model
    saver = tf.train.Saver()
    saver.save(sess, save_path)
    print('Model Trained and Saved')
    

#  slice index 0 of dimension 0 out of bounds.
#   encoding_embedding_size, decoding_embedding_size, rnn_size, num_layers, target_vocab_to_int)
#   File "<ipython-input-15-8888a89f5a5f>", line 33, in seq2seq_model
#     rnn_size, num_layers, target_vocab_to_int, keep_prob)
#   File "<ipython-input-14-e26eb46fd08a>", line 25, in decoding_layer
#     sequence_length,decoding_scope,output_fn,keep_prob)
#   File "<ipython-input-12-e3037e3ad4b7>", line 19, in decoding_layer_train



    


















    






Epoch   0 Batch    0/137 - Train Accuracy:  0.290, Validation Accuracy:  0.331, Loss:  5.878
Epoch   0 Batch    1/137 - Train Accuracy:  0.238, Validation Accuracy:  0.310, Loss:  5.580
Epoch   0 Batch    2/137 - Train Accuracy:  0.318, Validation Accuracy:  0.353, Loss:  4.253
Epoch   0 Batch    3/137 - Train Accuracy:  0.310, Validation Accuracy:  0.343, Loss:  4.169
Epoch   0 Batch    4/137 - Train Accuracy:  0.300, Validation Accuracy:  0.343, Loss:  3.712
Epoch   0 Batch    5/137 - Train Accuracy:  0.315, Validation Accuracy:  0.343, Loss:  3.546
Epoch   0 Batch    6/137 - Train Accuracy:  0.342, Validation Accuracy:  0.346, Loss:  3.374
Epoch   0 Batch    7/137 - Train Accuracy:  0.303, Validation Accuracy:  0.344, Loss:  3.397
Epoch   0 Batch    8/137 - Train Accuracy:  0.347, Validation Accuracy:  0.381, Loss:  3.338
Epoch   0 Batch    9/137 - Train Accuracy:  0.372, Validation Accuracy:  0.375, Loss:  3.035
Epoch   0 Batch   10/137 - Train Accuracy:  0.345, Validation Accuracy:  0.374, Loss:  3.298
Epoch   0 Batch   11/137 - Train Accuracy:  0.225, Validation Accuracy:  0.262, Loss:  3.302
Epoch   0 Batch   12/137 - Train Accuracy:  0.350, Validation Accuracy:  0.383, Loss:  3.652
Epoch   0 Batch   13/137 - Train Accuracy:  0.385, Validation Accuracy:  0.385, Loss:  2.983
Epoch   0 Batch   14/137 - Train Accuracy:  0.284, Validation Accuracy:  0.352, Loss:  3.153
Epoch   0 Batch   15/137 - Train Accuracy:  0.321, Validation Accuracy:  0.354, Loss:  3.002
Epoch   0 Batch   16/137 - Train Accuracy:  0.338, Validation Accuracy:  0.369, Loss:  2.932
Epoch   0 Batch   17/137 - Train Accuracy:  0.356, Validation Accuracy:  0.388, Loss:  2.881
Epoch   0 Batch   18/137 - Train Accuracy:  0.362, Validation Accuracy:  0.391, Loss:  2.852
Epoch   0 Batch   19/137 - Train Accuracy:  0.358, Validation Accuracy:  0.388, Loss:  2.821
Epoch   0 Batch   20/137 - Train Accuracy:  0.342, Validation Accuracy:  0.376, Loss:  2.778
Epoch   0 Batch   21/137 - Train Accuracy:  0.376, Validation Accuracy:  0.378, Loss:  2.634
Epoch   0 Batch   22/137 - Train Accuracy:  0.351, Validation Accuracy:  0.382, Loss:  2.728
Epoch   0 Batch   23/137 - Train Accuracy:  0.392, Validation Accuracy:  0.391, Loss:  2.577
Epoch   0 Batch   24/137 - Train Accuracy:  0.365, Validation Accuracy:  0.393, Loss:  2.649
Epoch   0 Batch   25/137 - Train Accuracy:  0.376, Validation Accuracy:  0.406, Loss:  2.639
Epoch   0 Batch   26/137 - Train Accuracy:  0.369, Validation Accuracy:  0.401, Loss:  2.611
Epoch   0 Batch   27/137 - Train Accuracy:  0.386, Validation Accuracy:  0.411, Loss:  2.563
Epoch   0 Batch   28/137 - Train Accuracy:  0.400, Validation Accuracy:  0.426, Loss:  2.539
Epoch   0 Batch   29/137 - Train Accuracy:  0.409, Validation Accuracy:  0.432, Loss:  2.497
Epoch   0 Batch   30/137 - Train Accuracy:  0.433, Validation Accuracy:  0.439, Loss:  2.367
Epoch   0 Batch   31/137 - Train Accuracy:  0.441, Validation Accuracy:  0.442, Loss:  2.349
Epoch   0 Batch   32/137 - Train Accuracy:  0.420, Validation Accuracy:  0.451, Loss:  2.387
Epoch   0 Batch   33/137 - Train Accuracy:  0.447, Validation Accuracy:  0.451, Loss:  2.280
Epoch   0 Batch   34/137 - Train Accuracy:  0.414, Validation Accuracy:  0.455, Loss:  2.365
Epoch   0 Batch   35/137 - Train Accuracy:  0.425, Validation Accuracy:  0.452, Loss:  2.297
Epoch   0 Batch   36/137 - Train Accuracy:  0.452, Validation Accuracy:  0.455, Loss:  2.171
Epoch   0 Batch   37/137 - Train Accuracy:  0.394, Validation Accuracy:  0.423, Loss:  2.244
Epoch   0 Batch   38/137 - Train Accuracy:  0.417, Validation Accuracy:  0.451, Loss:  2.250
Epoch   0 Batch   39/137 - Train Accuracy:  0.431, Validation Accuracy:  0.460, Loss:  2.188
Epoch   0 Batch   40/137 - Train Accuracy:  0.403, Validation Accuracy:  0.432, Loss:  2.141
Epoch   0 Batch   41/137 - Train Accuracy:  0.430, Validation Accuracy:  0.453, Loss:  2.128
Epoch   0 Batch   42/137 - Train Accuracy:  0.428, Validation Accuracy:  0.454, Loss:  2.047
Epoch   0 Batch   43/137 - Train Accuracy:  0.396, Validation Accuracy:  0.430, Loss:  2.075
Epoch   0 Batch   44/137 - Train Accuracy:  0.399, Validation Accuracy:  0.437, Loss:  2.020
Epoch   0 Batch   45/137 - Train Accuracy:  0.409, Validation Accuracy:  0.442, Loss:  1.972
Epoch   0 Batch   46/137 - Train Accuracy:  0.411, Validation Accuracy:  0.442, Loss:  1.909
Epoch   0 Batch   47/137 - Train Accuracy:  0.434, Validation Accuracy:  0.442, Loss:  1.820
Epoch   0 Batch   48/137 - Train Accuracy:  0.416, Validation Accuracy:  0.444, Loss:  1.844
Epoch   0 Batch   49/137 - Train Accuracy:  0.416, Validation Accuracy:  0.445, Loss:  1.815
Epoch   0 Batch   50/137 - Train Accuracy:  0.415, Validation Accuracy:  0.439, Loss:  1.773
Epoch   0 Batch   51/137 - Train Accuracy:  0.378, Validation Accuracy:  0.410, Loss:  1.730
Epoch   0 Batch   52/137 - Train Accuracy:  0.370, Validation Accuracy:  0.401, Loss:  1.720
Epoch   0 Batch   53/137 - Train Accuracy:  0.380, Validation Accuracy:  0.416, Loss:  1.685
Epoch   0 Batch   54/137 - Train Accuracy:  0.412, Validation Accuracy:  0.437, Loss:  1.656
Epoch   0 Batch   55/137 - Train Accuracy:  0.406, Validation Accuracy:  0.441, Loss:  1.628
Epoch   0 Batch   56/137 - Train Accuracy:  0.415, Validation Accuracy:  0.451, Loss:  1.621
Epoch   0 Batch   57/137 - Train Accuracy:  0.436, Validation Accuracy:  0.442, Loss:  1.486
Epoch   0 Batch   58/137 - Train Accuracy:  0.392, Validation Accuracy:  0.427, Loss:  1.528
Epoch   0 Batch   59/137 - Train Accuracy:  0.390, Validation Accuracy:  0.424, Loss:  1.527
Epoch   0 Batch   60/137 - Train Accuracy:  0.405, Validation Accuracy:  0.430, Loss:  1.484
Epoch   0 Batch   61/137 - Train Accuracy:  0.406, Validation Accuracy:  0.438, Loss:  1.463
Epoch   0 Batch   62/137 - Train Accuracy:  0.393, Validation Accuracy:  0.427, Loss:  1.445
Epoch   0 Batch   63/137 - Train Accuracy:  0.377, Validation Accuracy:  0.407, Loss:  1.393
Epoch   0 Batch   64/137 - Train Accuracy:  0.359, Validation Accuracy:  0.390, Loss:  1.389
Epoch   0 Batch   65/137 - Train Accuracy:  0.366, Validation Accuracy:  0.396, Loss:  1.389
Epoch   0 Batch   66/137 - Train Accuracy:  0.316, Validation Accuracy:  0.373, Loss:  1.440
Epoch   0 Batch   67/137 - Train Accuracy:  0.357, Validation Accuracy:  0.394, Loss:  1.336
Epoch   0 Batch   68/137 - Train Accuracy:  0.391, Validation Accuracy:  0.417, Loss:  1.311
Epoch   0 Batch   69/137 - Train Accuracy:  0.413, Validation Accuracy:  0.464, Loss:  1.385
Epoch   0 Batch   70/137 - Train Accuracy:  0.451, Validation Accuracy:  0.471, Loss:  1.266
Epoch   0 Batch   71/137 - Train Accuracy:  0.454, Validation Accuracy:  0.477, Loss:  1.262
Epoch   0 Batch   72/137 - Train Accuracy:  0.456, Validation Accuracy:  0.472, Loss:  1.235
Epoch   0 Batch   73/137 - Train Accuracy:  0.451, Validation Accuracy:  0.476, Loss:  1.246
Epoch   0 Batch   74/137 - Train Accuracy:  0.491, Validation Accuracy:  0.494, Loss:  1.165
Epoch   0 Batch   75/137 - Train Accuracy:  0.447, Validation Accuracy:  0.478, Loss:  1.228
Epoch   0 Batch   76/137 - Train Accuracy:  0.496, Validation Accuracy:  0.485, Loss:  1.147
Epoch   0 Batch   77/137 - Train Accuracy:  0.477, Validation Accuracy:  0.496, Loss:  1.187
Epoch   0 Batch   78/137 - Train Accuracy:  0.456, Validation Accuracy:  0.486, Loss:  1.162
Epoch   0 Batch   79/137 - Train Accuracy:  0.493, Validation Accuracy:  0.484, Loss:  1.089
Epoch   0 Batch   80/137 - Train Accuracy:  0.472, Validation Accuracy:  0.488, Loss:  1.139
Epoch   0 Batch   81/137 - Train Accuracy:  0.483, Validation Accuracy:  0.510, Loss:  1.143
Epoch   0 Batch   82/137 - Train Accuracy:  0.477, Validation Accuracy:  0.502, Loss:  1.118
Epoch   0 Batch   83/137 - Train Accuracy:  0.482, Validation Accuracy:  0.503, Loss:  1.109
Epoch   0 Batch   84/137 - Train Accuracy:  0.499, Validation Accuracy:  0.499, Loss:  1.036
Epoch   0 Batch   85/137 - Train Accuracy:  0.494, Validation Accuracy:  0.516, Loss:  1.098
Epoch   0 Batch   86/137 - Train Accuracy:  0.493, Validation Accuracy:  0.514, Loss:  1.087
Epoch   0 Batch   87/137 - Train Accuracy:  0.500, Validation Accuracy:  0.520, Loss:  1.072
Epoch   0 Batch   88/137 - Train Accuracy:  0.484, Validation Accuracy:  0.519, Loss:  1.062
Epoch   0 Batch   89/137 - Train Accuracy:  0.493, Validation Accuracy:  0.510, Loss:  1.072
Epoch   0 Batch   90/137 - Train Accuracy:  0.503, Validation Accuracy:  0.508, Loss:  1.001
Epoch   0 Batch   91/137 - Train Accuracy:  0.486, Validation Accuracy:  0.512, Loss:  1.036
Epoch   0 Batch   92/137 - Train Accuracy:  0.486, Validation Accuracy:  0.505, Loss:  1.031
Epoch   0 Batch   93/137 - Train Accuracy:  0.558, Validation Accuracy:  0.505, Loss:  0.888
Epoch   0 Batch   94/137 - Train Accuracy:  0.499, Validation Accuracy:  0.512, Loss:  1.004
Epoch   0 Batch   95/137 - Train Accuracy:  0.497, Validation Accuracy:  0.507, Loss:  0.993
Epoch   0 Batch   96/137 - Train Accuracy:  0.490, Validation Accuracy:  0.502, Loss:  0.979
Epoch   0 Batch   97/137 - Train Accuracy:  0.495, Validation Accuracy:  0.511, Loss:  0.994
Epoch   0 Batch   98/137 - Train Accuracy:  0.484, Validation Accuracy:  0.508, Loss:  0.982
Epoch   0 Batch   99/137 - Train Accuracy:  0.496, Validation Accuracy:  0.512, Loss:  0.983
Epoch   0 Batch  100/137 - Train Accuracy:  0.461, Validation Accuracy:  0.511, Loss:  1.014
Epoch   0 Batch  101/137 - Train Accuracy:  0.472, Validation Accuracy:  0.498, Loss:  0.980
Epoch   0 Batch  102/137 - Train Accuracy:  0.504, Validation Accuracy:  0.525, Loss:  0.953
Epoch   0 Batch  103/137 - Train Accuracy:  0.487, Validation Accuracy:  0.509, Loss:  0.964
Epoch   0 Batch  104/137 - Train Accuracy:  0.496, Validation Accuracy:  0.519, Loss:  0.946
Epoch   0 Batch  105/137 - Train Accuracy:  0.524, Validation Accuracy:  0.524, Loss:  0.907
Epoch   0 Batch  106/137 - Train Accuracy:  0.482, Validation Accuracy:  0.524, Loss:  0.976
Epoch   0 Batch  107/137 - Train Accuracy:  0.513, Validation Accuracy:  0.520, Loss:  0.918
Epoch   0 Batch  108/137 - Train Accuracy:  0.508, Validation Accuracy:  0.518, Loss:  0.921
Epoch   0 Batch  109/137 - Train Accuracy:  0.518, Validation Accuracy:  0.511, Loss:  0.863
Epoch   0 Batch  110/137 - Train Accuracy:  0.500, Validation Accuracy:  0.520, Loss:  0.923
Epoch   0 Batch  111/137 - Train Accuracy:  0.498, Validation Accuracy:  0.529, Loss:  0.941
Epoch   0 Batch  112/137 - Train Accuracy:  0.531, Validation Accuracy:  0.539, Loss:  0.860
Epoch   0 Batch  113/137 - Train Accuracy:  0.527, Validation Accuracy:  0.537, Loss:  0.865
Epoch   0 Batch  114/137 - Train Accuracy:  0.517, Validation Accuracy:  0.533, Loss:  0.881
Epoch   0 Batch  115/137 - Train Accuracy:  0.549, Validation Accuracy:  0.543, Loss:  0.839
Epoch   0 Batch  116/137 - Train Accuracy:  0.583, Validation Accuracy:  0.548, Loss:  0.761
Epoch   0 Batch  117/137 - Train Accuracy:  0.510, Validation Accuracy:  0.540, Loss:  0.878
Epoch   0 Batch  118/137 - Train Accuracy:  0.509, Validation Accuracy:  0.550, Loss:  0.909
Epoch   0 Batch  119/137 - Train Accuracy:  0.548, Validation Accuracy:  0.564, Loss:  0.861
Epoch   0 Batch  120/137 - Train Accuracy:  0.529, Validation Accuracy:  0.549, Loss:  0.852
Epoch   0 Batch  121/137 - Train Accuracy:  0.536, Validation Accuracy:  0.551, Loss:  0.838
Epoch   0 Batch  122/137 - Train Accuracy:  0.586, Validation Accuracy:  0.556, Loss:  0.769
Epoch   0 Batch  123/137 - Train Accuracy:  0.541, Validation Accuracy:  0.555, Loss:  0.841
Epoch   0 Batch  124/137 - Train Accuracy:  0.550, Validation Accuracy:  0.552, Loss:  0.812
Epoch   0 Batch  125/137 - Train Accuracy:  0.542, Validation Accuracy:  0.567, Loss:  0.829
Epoch   0 Batch  126/137 - Train Accuracy:  0.563, Validation Accuracy:  0.576, Loss:  0.806
Epoch   0 Batch  127/137 - Train Accuracy:  0.574, Validation Accuracy:  0.570, Loss:  0.768
Epoch   0 Batch  128/137 - Train Accuracy:  0.548, Validation Accuracy:  0.566, Loss:  0.796
Epoch   0 Batch  129/137 - Train Accuracy:  0.553, Validation Accuracy:  0.573, Loss:  0.808
Epoch   0 Batch  130/137 - Train Accuracy:  0.572, Validation Accuracy:  0.573, Loss:  0.771
Epoch   0 Batch  131/137 - Train Accuracy:  0.535, Validation Accuracy:  0.568, Loss:  0.804
Epoch   0 Batch  132/137 - Train Accuracy:  0.561, Validation Accuracy:  0.571, Loss:  0.802
Epoch   0 Batch  133/137 - Train Accuracy:  0.555, Validation Accuracy:  0.576, Loss:  0.784
Epoch   0 Batch  134/137 - Train Accuracy:  0.533, Validation Accuracy:  0.568, Loss:  0.814
Epoch   0 Batch  135/137 - Train Accuracy:  0.522, Validation Accuracy:  0.563, Loss:  0.803
Epoch   1 Batch    0/137 - Train Accuracy:  0.553, Validation Accuracy:  0.571, Loss:  0.769
Epoch   1 Batch    1/137 - Train Accuracy:  0.546, Validation Accuracy:  0.581, Loss:  0.795
Epoch   1 Batch    2/137 - Train Accuracy:  0.557, Validation Accuracy:  0.581, Loss:  0.748
Epoch   1 Batch    3/137 - Train Accuracy:  0.580, Validation Accuracy:  0.588, Loss:  0.753
Epoch   1 Batch    4/137 - Train Accuracy:  0.576, Validation Accuracy:  0.585, Loss:  0.760
Epoch   1 Batch    5/137 - Train Accuracy:  0.571, Validation Accuracy:  0.592, Loss:  0.756
Epoch   1 Batch    6/137 - Train Accuracy:  0.594, Validation Accuracy:  0.590, Loss:  0.711
Epoch   1 Batch    7/137 - Train Accuracy:  0.581, Validation Accuracy:  0.597, Loss:  0.738
Epoch   1 Batch    8/137 - Train Accuracy:  0.582, Validation Accuracy:  0.590, Loss:  0.729
Epoch   1 Batch    9/137 - Train Accuracy:  0.604, Validation Accuracy:  0.596, Loss:  0.699
Epoch   1 Batch   10/137 - Train Accuracy:  0.596, Validation Accuracy:  0.601, Loss:  0.732
Epoch   1 Batch   11/137 - Train Accuracy:  0.594, Validation Accuracy:  0.599, Loss:  0.727
Epoch   1 Batch   12/137 - Train Accuracy:  0.584, Validation Accuracy:  0.600, Loss:  0.730
Epoch   1 Batch   13/137 - Train Accuracy:  0.603, Validation Accuracy:  0.602, Loss:  0.689
Epoch   1 Batch   14/137 - Train Accuracy:  0.563, Validation Accuracy:  0.604, Loss:  0.756
Epoch   1 Batch   15/137 - Train Accuracy:  0.591, Validation Accuracy:  0.604, Loss:  0.713
Epoch   1 Batch   16/137 - Train Accuracy:  0.592, Validation Accuracy:  0.603, Loss:  0.703
Epoch   1 Batch   17/137 - Train Accuracy:  0.591, Validation Accuracy:  0.601, Loss:  0.722
Epoch   1 Batch   18/137 - Train Accuracy:  0.593, Validation Accuracy:  0.601, Loss:  0.715
Epoch   1 Batch   19/137 - Train Accuracy:  0.588, Validation Accuracy:  0.602, Loss:  0.704
Epoch   1 Batch   20/137 - Train Accuracy:  0.583, Validation Accuracy:  0.595, Loss:  0.718
Epoch   1 Batch   21/137 - Train Accuracy:  0.612, Validation Accuracy:  0.597, Loss:  0.673
Epoch   1 Batch   22/137 - Train Accuracy:  0.589, Validation Accuracy:  0.597, Loss:  0.708
Epoch   1 Batch   23/137 - Train Accuracy:  0.606, Validation Accuracy:  0.589, Loss:  0.663
Epoch   1 Batch   24/137 - Train Accuracy:  0.587, Validation Accuracy:  0.599, Loss:  0.688
Epoch   1 Batch   25/137 - Train Accuracy:  0.597, Validation Accuracy:  0.604, Loss:  0.688
Epoch   1 Batch   26/137 - Train Accuracy:  0.586, Validation Accuracy:  0.605, Loss:  0.685
Epoch   1 Batch   27/137 - Train Accuracy:  0.609, Validation Accuracy:  0.610, Loss:  0.684
Epoch   1 Batch   28/137 - Train Accuracy:  0.605, Validation Accuracy:  0.607, Loss:  0.690
Epoch   1 Batch   29/137 - Train Accuracy:  0.602, Validation Accuracy:  0.604, Loss:  0.684
Epoch   1 Batch   30/137 - Train Accuracy:  0.623, Validation Accuracy:  0.611, Loss:  0.643
Epoch   1 Batch   31/137 - Train Accuracy:  0.603, Validation Accuracy:  0.610, Loss:  0.660
Epoch   1 Batch   32/137 - Train Accuracy:  0.589, Validation Accuracy:  0.604, Loss:  0.676
Epoch   1 Batch   33/137 - Train Accuracy:  0.603, Validation Accuracy:  0.598, Loss:  0.649
Epoch   1 Batch   34/137 - Train Accuracy:  0.589, Validation Accuracy:  0.611, Loss:  0.688
Epoch   1 Batch   35/137 - Train Accuracy:  0.607, Validation Accuracy:  0.616, Loss:  0.682
Epoch   1 Batch   36/137 - Train Accuracy:  0.620, Validation Accuracy:  0.615, Loss:  0.656
Epoch   1 Batch   37/137 - Train Accuracy:  0.608, Validation Accuracy:  0.611, Loss:  0.666
Epoch   1 Batch   38/137 - Train Accuracy:  0.601, Validation Accuracy:  0.619, Loss:  0.678
Epoch   1 Batch   39/137 - Train Accuracy:  0.612, Validation Accuracy:  0.613, Loss:  0.666
Epoch   1 Batch   40/137 - Train Accuracy:  0.610, Validation Accuracy:  0.622, Loss:  0.660
Epoch   1 Batch   41/137 - Train Accuracy:  0.623, Validation Accuracy:  0.625, Loss:  0.656
Epoch   1 Batch   42/137 - Train Accuracy:  0.610, Validation Accuracy:  0.617, Loss:  0.660
Epoch   1 Batch   43/137 - Train Accuracy:  0.591, Validation Accuracy:  0.617, Loss:  0.656
Epoch   1 Batch   44/137 - Train Accuracy:  0.608, Validation Accuracy:  0.625, Loss:  0.668
Epoch   1 Batch   45/137 - Train Accuracy:  0.616, Validation Accuracy:  0.627, Loss:  0.661
Epoch   1 Batch   46/137 - Train Accuracy:  0.627, Validation Accuracy:  0.629, Loss:  0.632
Epoch   1 Batch   47/137 - Train Accuracy:  0.626, Validation Accuracy:  0.631, Loss:  0.625
Epoch   1 Batch   48/137 - Train Accuracy:  0.609, Validation Accuracy:  0.628, Loss:  0.655
Epoch   1 Batch   49/137 - Train Accuracy:  0.609, Validation Accuracy:  0.634, Loss:  0.644
Epoch   1 Batch   50/137 - Train Accuracy:  0.623, Validation Accuracy:  0.633, Loss:  0.641
Epoch   1 Batch   51/137 - Train Accuracy:  0.613, Validation Accuracy:  0.627, Loss:  0.637
Epoch   1 Batch   52/137 - Train Accuracy:  0.599, Validation Accuracy:  0.621, Loss:  0.642
Epoch   1 Batch   53/137 - Train Accuracy:  0.603, Validation Accuracy:  0.627, Loss:  0.645
Epoch   1 Batch   54/137 - Train Accuracy:  0.614, Validation Accuracy:  0.631, Loss:  0.636
Epoch   1 Batch   55/137 - Train Accuracy:  0.592, Validation Accuracy:  0.626, Loss:  0.637
Epoch   1 Batch   56/137 - Train Accuracy:  0.607, Validation Accuracy:  0.630, Loss:  0.630
Epoch   1 Batch   57/137 - Train Accuracy:  0.638, Validation Accuracy:  0.634, Loss:  0.606
Epoch   1 Batch   58/137 - Train Accuracy:  0.625, Validation Accuracy:  0.635, Loss:  0.628
Epoch   1 Batch   59/137 - Train Accuracy:  0.624, Validation Accuracy:  0.634, Loss:  0.633
Epoch   1 Batch   60/137 - Train Accuracy:  0.633, Validation Accuracy:  0.639, Loss:  0.622
Epoch   1 Batch   61/137 - Train Accuracy:  0.635, Validation Accuracy:  0.641, Loss:  0.613
Epoch   1 Batch   62/137 - Train Accuracy:  0.625, Validation Accuracy:  0.641, Loss:  0.615
Epoch   1 Batch   63/137 - Train Accuracy:  0.620, Validation Accuracy:  0.631, Loss:  0.618
Epoch   1 Batch   64/137 - Train Accuracy:  0.625, Validation Accuracy:  0.634, Loss:  0.636
Epoch   1 Batch   65/137 - Train Accuracy:  0.629, Validation Accuracy:  0.634, Loss:  0.647
Epoch   1 Batch   66/137 - Train Accuracy:  0.591, Validation Accuracy:  0.628, Loss:  0.650
Epoch   1 Batch   67/137 - Train Accuracy:  0.613, Validation Accuracy:  0.624, Loss:  0.623
Epoch   1 Batch   68/137 - Train Accuracy:  0.615, Validation Accuracy:  0.625, Loss:  0.617
Epoch   1 Batch   69/137 - Train Accuracy:  0.590, Validation Accuracy:  0.628, Loss:  0.663
Epoch   1 Batch   70/137 - Train Accuracy:  0.627, Validation Accuracy:  0.634, Loss:  0.612
Epoch   1 Batch   71/137 - Train Accuracy:  0.639, Validation Accuracy:  0.638, Loss:  0.623
Epoch   1 Batch   72/137 - Train Accuracy:  0.632, Validation Accuracy:  0.635, Loss:  0.604
Epoch   1 Batch   73/137 - Train Accuracy:  0.621, Validation Accuracy:  0.630, Loss:  0.602
Epoch   1 Batch   74/137 - Train Accuracy:  0.644, Validation Accuracy:  0.638, Loss:  0.579
Epoch   1 Batch   75/137 - Train Accuracy:  0.620, Validation Accuracy:  0.639, Loss:  0.620
Epoch   1 Batch   76/137 - Train Accuracy:  0.654, Validation Accuracy:  0.643, Loss:  0.587
Epoch   1 Batch   77/137 - Train Accuracy:  0.632, Validation Accuracy:  0.641, Loss:  0.608
Epoch   1 Batch   78/137 - Train Accuracy:  0.626, Validation Accuracy:  0.637, Loss:  0.597
Epoch   1 Batch   79/137 - Train Accuracy:  0.649, Validation Accuracy:  0.642, Loss:  0.576
Epoch   1 Batch   80/137 - Train Accuracy:  0.631, Validation Accuracy:  0.642, Loss:  0.604
Epoch   1 Batch   81/137 - Train Accuracy:  0.624, Validation Accuracy:  0.637, Loss:  0.603
Epoch   1 Batch   82/137 - Train Accuracy:  0.623, Validation Accuracy:  0.632, Loss:  0.604
Epoch   1 Batch   83/137 - Train Accuracy:  0.641, Validation Accuracy:  0.643, Loss:  0.593
Epoch   1 Batch   84/137 - Train Accuracy:  0.650, Validation Accuracy:  0.648, Loss:  0.563
Epoch   1 Batch   85/137 - Train Accuracy:  0.631, Validation Accuracy:  0.649, Loss:  0.602
Epoch   1 Batch   86/137 - Train Accuracy:  0.635, Validation Accuracy:  0.650, Loss:  0.596
Epoch   1 Batch   87/137 - Train Accuracy:  0.639, Validation Accuracy:  0.650, Loss:  0.600
Epoch   1 Batch   88/137 - Train Accuracy:  0.640, Validation Accuracy:  0.653, Loss:  0.585
Epoch   1 Batch   89/137 - Train Accuracy:  0.634, Validation Accuracy:  0.644, Loss:  0.600
Epoch   1 Batch   90/137 - Train Accuracy:  0.654, Validation Accuracy:  0.647, Loss:  0.556
Epoch   1 Batch   91/137 - Train Accuracy:  0.629, Validation Accuracy:  0.646, Loss:  0.593
Epoch   1 Batch   92/137 - Train Accuracy:  0.628, Validation Accuracy:  0.645, Loss:  0.586
Epoch   1 Batch   93/137 - Train Accuracy:  0.676, Validation Accuracy:  0.636, Loss:  0.506
Epoch   1 Batch   94/137 - Train Accuracy:  0.640, Validation Accuracy:  0.639, Loss:  0.576
Epoch   1 Batch   95/137 - Train Accuracy:  0.639, Validation Accuracy:  0.647, Loss:  0.575
Epoch   1 Batch   96/137 - Train Accuracy:  0.646, Validation Accuracy:  0.655, Loss:  0.564
Epoch   1 Batch   97/137 - Train Accuracy:  0.646, Validation Accuracy:  0.653, Loss:  0.571
Epoch   1 Batch   98/137 - Train Accuracy:  0.648, Validation Accuracy:  0.657, Loss:  0.577
Epoch   1 Batch   99/137 - Train Accuracy:  0.654, Validation Accuracy:  0.665, Loss:  0.574
Epoch   1 Batch  100/137 - Train Accuracy:  0.618, Validation Accuracy:  0.663, Loss:  0.604
Epoch   1 Batch  101/137 - Train Accuracy:  0.646, Validation Accuracy:  0.664, Loss:  0.584
Epoch   1 Batch  102/137 - Train Accuracy:  0.658, Validation Accuracy:  0.660, Loss:  0.566
Epoch   1 Batch  103/137 - Train Accuracy:  0.653, Validation Accuracy:  0.662, Loss:  0.575
Epoch   1 Batch  104/137 - Train Accuracy:  0.640, Validation Accuracy:  0.660, Loss:  0.564
Epoch   1 Batch  105/137 - Train Accuracy:  0.665, Validation Accuracy:  0.662, Loss:  0.548
Epoch   1 Batch  106/137 - Train Accuracy:  0.621, Validation Accuracy:  0.645, Loss:  0.586
Epoch   1 Batch  107/137 - Train Accuracy:  0.643, Validation Accuracy:  0.648, Loss:  0.567
Epoch   1 Batch  108/137 - Train Accuracy:  0.651, Validation Accuracy:  0.650, Loss:  0.568
Epoch   1 Batch  109/137 - Train Accuracy:  0.673, Validation Accuracy:  0.656, Loss:  0.538
Epoch   1 Batch  110/137 - Train Accuracy:  0.637, Validation Accuracy:  0.654, Loss:  0.567
Epoch   1 Batch  111/137 - Train Accuracy:  0.637, Validation Accuracy:  0.661, Loss:  0.584
Epoch   1 Batch  112/137 - Train Accuracy:  0.670, Validation Accuracy:  0.666, Loss:  0.533
Epoch   1 Batch  113/137 - Train Accuracy:  0.672, Validation Accuracy:  0.667, Loss:  0.546
Epoch   1 Batch  114/137 - Train Accuracy:  0.663, Validation Accuracy:  0.663, Loss:  0.552
Epoch   1 Batch  115/137 - Train Accuracy:  0.671, Validation Accuracy:  0.666, Loss:  0.525
Epoch   1 Batch  116/137 - Train Accuracy:  0.709, Validation Accuracy:  0.668, Loss:  0.486
Epoch   1 Batch  117/137 - Train Accuracy:  0.646, Validation Accuracy:  0.666, Loss:  0.554
Epoch   1 Batch  118/137 - Train Accuracy:  0.635, Validation Accuracy:  0.663, Loss:  0.582
Epoch   1 Batch  119/137 - Train Accuracy:  0.658, Validation Accuracy:  0.663, Loss:  0.551
Epoch   1 Batch  120/137 - Train Accuracy:  0.640, Validation Accuracy:  0.651, Loss:  0.547
Epoch   1 Batch  121/137 - Train Accuracy:  0.641, Validation Accuracy:  0.648, Loss:  0.573
Epoch   1 Batch  122/137 - Train Accuracy:  0.676, Validation Accuracy:  0.657, Loss:  0.537
Epoch   1 Batch  123/137 - Train Accuracy:  0.633, Validation Accuracy:  0.658, Loss:  0.569
Epoch   1 Batch  124/137 - Train Accuracy:  0.650, Validation Accuracy:  0.650, Loss:  0.548
Epoch   1 Batch  125/137 - Train Accuracy:  0.640, Validation Accuracy:  0.658, Loss:  0.561
Epoch   1 Batch  126/137 - Train Accuracy:  0.657, Validation Accuracy:  0.665, Loss:  0.550
Epoch   1 Batch  127/137 - Train Accuracy:  0.681, Validation Accuracy:  0.666, Loss:  0.521
Epoch   1 Batch  128/137 - Train Accuracy:  0.664, Validation Accuracy:  0.667, Loss:  0.548
Epoch   1 Batch  129/137 - Train Accuracy:  0.647, Validation Accuracy:  0.672, Loss:  0.555
Epoch   1 Batch  130/137 - Train Accuracy:  0.668, Validation Accuracy:  0.673, Loss:  0.530
Epoch   1 Batch  131/137 - Train Accuracy:  0.643, Validation Accuracy:  0.674, Loss:  0.552
Epoch   1 Batch  132/137 - Train Accuracy:  0.664, Validation Accuracy:  0.672, Loss:  0.547
Epoch   1 Batch  133/137 - Train Accuracy:  0.651, Validation Accuracy:  0.668, Loss:  0.540
Epoch   1 Batch  134/137 - Train Accuracy:  0.642, Validation Accuracy:  0.661, Loss:  0.560
Epoch   1 Batch  135/137 - Train Accuracy:  0.632, Validation Accuracy:  0.659, Loss:  0.555
Epoch   2 Batch    0/137 - Train Accuracy:  0.644, Validation Accuracy:  0.660, Loss:  0.537
Epoch   2 Batch    1/137 - Train Accuracy:  0.640, Validation Accuracy:  0.665, Loss:  0.561
Epoch   2 Batch    2/137 - Train Accuracy:  0.655, Validation Accuracy:  0.667, Loss:  0.524
Epoch   2 Batch    3/137 - Train Accuracy:  0.660, Validation Accuracy:  0.671, Loss:  0.530
Epoch   2 Batch    4/137 - Train Accuracy:  0.656, Validation Accuracy:  0.669, Loss:  0.539
Epoch   2 Batch    5/137 - Train Accuracy:  0.665, Validation Accuracy:  0.675, Loss:  0.527
Epoch   2 Batch    6/137 - Train Accuracy:  0.671, Validation Accuracy:  0.677, Loss:  0.507
Epoch   2 Batch    7/137 - Train Accuracy:  0.659, Validation Accuracy:  0.674, Loss:  0.525
Epoch   2 Batch    8/137 - Train Accuracy:  0.667, Validation Accuracy:  0.677, Loss:  0.519
Epoch   2 Batch    9/137 - Train Accuracy:  0.681, Validation Accuracy:  0.674, Loss:  0.495
Epoch   2 Batch   10/137 - Train Accuracy:  0.665, Validation Accuracy:  0.676, Loss:  0.522
Epoch   2 Batch   11/137 - Train Accuracy:  0.670, Validation Accuracy:  0.676, Loss:  0.513
Epoch   2 Batch   12/137 - Train Accuracy:  0.663, Validation Accuracy:  0.675, Loss:  0.520
Epoch   2 Batch   13/137 - Train Accuracy:  0.679, Validation Accuracy:  0.676, Loss:  0.494
Epoch   2 Batch   14/137 - Train Accuracy:  0.638, Validation Accuracy:  0.677, Loss:  0.545
Epoch   2 Batch   15/137 - Train Accuracy:  0.663, Validation Accuracy:  0.675, Loss:  0.514
Epoch   2 Batch   16/137 - Train Accuracy:  0.668, Validation Accuracy:  0.677, Loss:  0.505
Epoch   2 Batch   17/137 - Train Accuracy:  0.664, Validation Accuracy:  0.680, Loss:  0.519
Epoch   2 Batch   18/137 - Train Accuracy:  0.671, Validation Accuracy:  0.677, Loss:  0.509
Epoch   2 Batch   19/137 - Train Accuracy:  0.660, Validation Accuracy:  0.678, Loss:  0.510
Epoch   2 Batch   20/137 - Train Accuracy:  0.661, Validation Accuracy:  0.678, Loss:  0.521
Epoch   2 Batch   21/137 - Train Accuracy:  0.684, Validation Accuracy:  0.675, Loss:  0.477
Epoch   2 Batch   22/137 - Train Accuracy:  0.669, Validation Accuracy:  0.673, Loss:  0.512
Epoch   2 Batch   23/137 - Train Accuracy:  0.691, Validation Accuracy:  0.675, Loss:  0.479
Epoch   2 Batch   24/137 - Train Accuracy:  0.659, Validation Accuracy:  0.676, Loss:  0.493
Epoch   2 Batch   25/137 - Train Accuracy:  0.659, Validation Accuracy:  0.678, Loss:  0.503
Epoch   2 Batch   26/137 - Train Accuracy:  0.662, Validation Accuracy:  0.676, Loss:  0.497
Epoch   2 Batch   27/137 - Train Accuracy:  0.679, Validation Accuracy:  0.679, Loss:  0.498
Epoch   2 Batch   28/137 - Train Accuracy:  0.681, Validation Accuracy:  0.686, Loss:  0.504
Epoch   2 Batch   29/137 - Train Accuracy:  0.676, Validation Accuracy:  0.684, Loss:  0.496
Epoch   2 Batch   30/137 - Train Accuracy:  0.677, Validation Accuracy:  0.680, Loss:  0.468
Epoch   2 Batch   31/137 - Train Accuracy:  0.675, Validation Accuracy:  0.677, Loss:  0.485
Epoch   2 Batch   32/137 - Train Accuracy:  0.667, Validation Accuracy:  0.678, Loss:  0.494
Epoch   2 Batch   33/137 - Train Accuracy:  0.664, Validation Accuracy:  0.676, Loss:  0.481
Epoch   2 Batch   34/137 - Train Accuracy:  0.646, Validation Accuracy:  0.674, Loss:  0.510
Epoch   2 Batch   35/137 - Train Accuracy:  0.670, Validation Accuracy:  0.679, Loss:  0.497
Epoch   2 Batch   36/137 - Train Accuracy:  0.680, Validation Accuracy:  0.683, Loss:  0.487
Epoch   2 Batch   37/137 - Train Accuracy:  0.684, Validation Accuracy:  0.686, Loss:  0.494
Epoch   2 Batch   38/137 - Train Accuracy:  0.667, Validation Accuracy:  0.687, Loss:  0.499
Epoch   2 Batch   39/137 - Train Accuracy:  0.683, Validation Accuracy:  0.682, Loss:  0.489
Epoch   2 Batch   40/137 - Train Accuracy:  0.673, Validation Accuracy:  0.679, Loss:  0.486
Epoch   2 Batch   41/137 - Train Accuracy:  0.688, Validation Accuracy:  0.680, Loss:  0.485
Epoch   2 Batch   42/137 - Train Accuracy:  0.679, Validation Accuracy:  0.685, Loss:  0.500
Epoch   2 Batch   43/137 - Train Accuracy:  0.662, Validation Accuracy:  0.685, Loss:  0.491
Epoch   2 Batch   44/137 - Train Accuracy:  0.665, Validation Accuracy:  0.687, Loss:  0.488
Epoch   2 Batch   45/137 - Train Accuracy:  0.678, Validation Accuracy:  0.691, Loss:  0.492
Epoch   2 Batch   46/137 - Train Accuracy:  0.682, Validation Accuracy:  0.688, Loss:  0.477
Epoch   2 Batch   47/137 - Train Accuracy:  0.683, Validation Accuracy:  0.689, Loss:  0.466
Epoch   2 Batch   48/137 - Train Accuracy:  0.678, Validation Accuracy:  0.689, Loss:  0.490
Epoch   2 Batch   49/137 - Train Accuracy:  0.671, Validation Accuracy:  0.693, Loss:  0.479
Epoch   2 Batch   50/137 - Train Accuracy:  0.684, Validation Accuracy:  0.691, Loss:  0.481
Epoch   2 Batch   51/137 - Train Accuracy:  0.689, Validation Accuracy:  0.690, Loss:  0.476
Epoch   2 Batch   52/137 - Train Accuracy:  0.677, Validation Accuracy:  0.691, Loss:  0.479
Epoch   2 Batch   53/137 - Train Accuracy:  0.661, Validation Accuracy:  0.685, Loss:  0.480
Epoch   2 Batch   54/137 - Train Accuracy:  0.670, Validation Accuracy:  0.684, Loss:  0.476
Epoch   2 Batch   55/137 - Train Accuracy:  0.657, Validation Accuracy:  0.685, Loss:  0.481
Epoch   2 Batch   56/137 - Train Accuracy:  0.673, Validation Accuracy:  0.693, Loss:  0.476
Epoch   2 Batch   57/137 - Train Accuracy:  0.691, Validation Accuracy:  0.697, Loss:  0.464
Epoch   2 Batch   58/137 - Train Accuracy:  0.684, Validation Accuracy:  0.692, Loss:  0.479
Epoch   2 Batch   59/137 - Train Accuracy:  0.673, Validation Accuracy:  0.690, Loss:  0.478
Epoch   2 Batch   60/137 - Train Accuracy:  0.697, Validation Accuracy:  0.696, Loss:  0.466
Epoch   2 Batch   61/137 - Train Accuracy:  0.682, Validation Accuracy:  0.695, Loss:  0.466
Epoch   2 Batch   62/137 - Train Accuracy:  0.689, Validation Accuracy:  0.693, Loss:  0.465
Epoch   2 Batch   63/137 - Train Accuracy:  0.678, Validation Accuracy:  0.692, Loss:  0.469
Epoch   2 Batch   64/137 - Train Accuracy:  0.685, Validation Accuracy:  0.694, Loss:  0.462
Epoch   2 Batch   65/137 - Train Accuracy:  0.683, Validation Accuracy:  0.696, Loss:  0.476
Epoch   2 Batch   66/137 - Train Accuracy:  0.668, Validation Accuracy:  0.696, Loss:  0.492
Epoch   2 Batch   67/137 - Train Accuracy:  0.681, Validation Accuracy:  0.698, Loss:  0.465
Epoch   2 Batch   68/137 - Train Accuracy:  0.684, Validation Accuracy:  0.697, Loss:  0.466
Epoch   2 Batch   69/137 - Train Accuracy:  0.656, Validation Accuracy:  0.695, Loss:  0.490
Epoch   2 Batch   70/137 - Train Accuracy:  0.698, Validation Accuracy:  0.694, Loss:  0.459
Epoch   2 Batch   71/137 - Train Accuracy:  0.691, Validation Accuracy:  0.699, Loss:  0.474
Epoch   2 Batch   72/137 - Train Accuracy:  0.695, Validation Accuracy:  0.700, Loss:  0.457
Epoch   2 Batch   73/137 - Train Accuracy:  0.700, Validation Accuracy:  0.705, Loss:  0.453
Epoch   2 Batch   74/137 - Train Accuracy:  0.712, Validation Accuracy:  0.709, Loss:  0.437
Epoch   2 Batch   75/137 - Train Accuracy:  0.692, Validation Accuracy:  0.708, Loss:  0.468
Epoch   2 Batch   76/137 - Train Accuracy:  0.716, Validation Accuracy:  0.706, Loss:  0.439
Epoch   2 Batch   77/137 - Train Accuracy:  0.695, Validation Accuracy:  0.707, Loss:  0.458
Epoch   2 Batch   78/137 - Train Accuracy:  0.689, Validation Accuracy:  0.706, Loss:  0.451
Epoch   2 Batch   79/137 - Train Accuracy:  0.715, Validation Accuracy:  0.705, Loss:  0.431
Epoch   2 Batch   80/137 - Train Accuracy:  0.702, Validation Accuracy:  0.708, Loss:  0.456
Epoch   2 Batch   81/137 - Train Accuracy:  0.700, Validation Accuracy:  0.717, Loss:  0.452
Epoch   2 Batch   82/137 - Train Accuracy:  0.698, Validation Accuracy:  0.721, Loss:  0.454
Epoch   2 Batch   83/137 - Train Accuracy:  0.710, Validation Accuracy:  0.722, Loss:  0.449
Epoch   2 Batch   84/137 - Train Accuracy:  0.720, Validation Accuracy:  0.724, Loss:  0.422
Epoch   2 Batch   85/137 - Train Accuracy:  0.702, Validation Accuracy:  0.723, Loss:  0.453
Epoch   2 Batch   86/137 - Train Accuracy:  0.707, Validation Accuracy:  0.722, Loss:  0.443
Epoch   2 Batch   87/137 - Train Accuracy:  0.711, Validation Accuracy:  0.728, Loss:  0.451
Epoch   2 Batch   88/137 - Train Accuracy:  0.713, Validation Accuracy:  0.722, Loss:  0.435
Epoch   2 Batch   89/137 - Train Accuracy:  0.714, Validation Accuracy:  0.720, Loss:  0.453
Epoch   2 Batch   90/137 - Train Accuracy:  0.730, Validation Accuracy:  0.724, Loss:  0.420
Epoch   2 Batch   91/137 - Train Accuracy:  0.694, Validation Accuracy:  0.723, Loss:  0.448
Epoch   2 Batch   92/137 - Train Accuracy:  0.698, Validation Accuracy:  0.719, Loss:  0.441
Epoch   2 Batch   93/137 - Train Accuracy:  0.752, Validation Accuracy:  0.718, Loss:  0.382
Epoch   2 Batch   94/137 - Train Accuracy:  0.733, Validation Accuracy:  0.722, Loss:  0.432
Epoch   2 Batch   95/137 - Train Accuracy:  0.725, Validation Accuracy:  0.728, Loss:  0.425
Epoch   2 Batch   96/137 - Train Accuracy:  0.730, Validation Accuracy:  0.730, Loss:  0.425
Epoch   2 Batch   97/137 - Train Accuracy:  0.727, Validation Accuracy:  0.726, Loss:  0.431
Epoch   2 Batch   98/137 - Train Accuracy:  0.727, Validation Accuracy:  0.728, Loss:  0.428
Epoch   2 Batch   99/137 - Train Accuracy:  0.723, Validation Accuracy:  0.732, Loss:  0.432
Epoch   2 Batch  100/137 - Train Accuracy:  0.705, Validation Accuracy:  0.736, Loss:  0.455
Epoch   2 Batch  101/137 - Train Accuracy:  0.724, Validation Accuracy:  0.736, Loss:  0.430
Epoch   2 Batch  102/137 - Train Accuracy:  0.734, Validation Accuracy:  0.735, Loss:  0.422
Epoch   2 Batch  103/137 - Train Accuracy:  0.735, Validation Accuracy:  0.732, Loss:  0.428
Epoch   2 Batch  104/137 - Train Accuracy:  0.719, Validation Accuracy:  0.733, Loss:  0.430
Epoch   2 Batch  105/137 - Train Accuracy:  0.736, Validation Accuracy:  0.732, Loss:  0.408
Epoch   2 Batch  106/137 - Train Accuracy:  0.723, Validation Accuracy:  0.735, Loss:  0.442
Epoch   2 Batch  107/137 - Train Accuracy:  0.738, Validation Accuracy:  0.732, Loss:  0.413
Epoch   2 Batch  108/137 - Train Accuracy:  0.730, Validation Accuracy:  0.734, Loss:  0.429
Epoch   2 Batch  109/137 - Train Accuracy:  0.753, Validation Accuracy:  0.739, Loss:  0.400
Epoch   2 Batch  110/137 - Train Accuracy:  0.725, Validation Accuracy:  0.740, Loss:  0.425
Epoch   2 Batch  111/137 - Train Accuracy:  0.726, Validation Accuracy:  0.735, Loss:  0.436
Epoch   2 Batch  112/137 - Train Accuracy:  0.738, Validation Accuracy:  0.733, Loss:  0.401
Epoch   2 Batch  113/137 - Train Accuracy:  0.751, Validation Accuracy:  0.730, Loss:  0.408
Epoch   2 Batch  114/137 - Train Accuracy:  0.741, Validation Accuracy:  0.733, Loss:  0.421
Epoch   2 Batch  115/137 - Train Accuracy:  0.745, Validation Accuracy:  0.735, Loss:  0.395
Epoch   2 Batch  116/137 - Train Accuracy:  0.781, Validation Accuracy:  0.736, Loss:  0.365
Epoch   2 Batch  117/137 - Train Accuracy:  0.721, Validation Accuracy:  0.740, Loss:  0.417
Epoch   2 Batch  118/137 - Train Accuracy:  0.724, Validation Accuracy:  0.741, Loss:  0.434
Epoch   2 Batch  119/137 - Train Accuracy:  0.739, Validation Accuracy:  0.740, Loss:  0.414
Epoch   2 Batch  120/137 - Train Accuracy:  0.731, Validation Accuracy:  0.740, Loss:  0.400
Epoch   2 Batch  121/137 - Train Accuracy:  0.744, Validation Accuracy:  0.746, Loss:  0.401
Epoch   2 Batch  122/137 - Train Accuracy:  0.759, Validation Accuracy:  0.747, Loss:  0.371
Epoch   2 Batch  123/137 - Train Accuracy:  0.731, Validation Accuracy:  0.742, Loss:  0.418
Epoch   2 Batch  124/137 - Train Accuracy:  0.756, Validation Accuracy:  0.738, Loss:  0.396
Epoch   2 Batch  125/137 - Train Accuracy:  0.724, Validation Accuracy:  0.736, Loss:  0.408
Epoch   2 Batch  126/137 - Train Accuracy:  0.743, Validation Accuracy:  0.740, Loss:  0.398
Epoch   2 Batch  127/137 - Train Accuracy:  0.761, Validation Accuracy:  0.743, Loss:  0.374
Epoch   2 Batch  128/137 - Train Accuracy:  0.747, Validation Accuracy:  0.744, Loss:  0.401
Epoch   2 Batch  129/137 - Train Accuracy:  0.740, Validation Accuracy:  0.750, Loss:  0.411
Epoch   2 Batch  130/137 - Train Accuracy:  0.749, Validation Accuracy:  0.748, Loss:  0.387
Epoch   2 Batch  131/137 - Train Accuracy:  0.729, Validation Accuracy:  0.750, Loss:  0.412
Epoch   2 Batch  132/137 - Train Accuracy:  0.743, Validation Accuracy:  0.745, Loss:  0.399
Epoch   2 Batch  133/137 - Train Accuracy:  0.754, Validation Accuracy:  0.751, Loss:  0.398
Epoch   2 Batch  134/137 - Train Accuracy:  0.733, Validation Accuracy:  0.752, Loss:  0.411
Epoch   2 Batch  135/137 - Train Accuracy:  0.737, Validation Accuracy:  0.749, Loss:  0.414
Epoch   3 Batch    0/137 - Train Accuracy:  0.739, Validation Accuracy:  0.753, Loss:  0.400
Epoch   3 Batch    1/137 - Train Accuracy:  0.727, Validation Accuracy:  0.753, Loss:  0.411
Epoch   3 Batch    2/137 - Train Accuracy:  0.733, Validation Accuracy:  0.750, Loss:  0.393
Epoch   3 Batch    3/137 - Train Accuracy:  0.752, Validation Accuracy:  0.748, Loss:  0.398
Epoch   3 Batch    4/137 - Train Accuracy:  0.747, Validation Accuracy:  0.746, Loss:  0.413
Epoch   3 Batch    5/137 - Train Accuracy:  0.744, Validation Accuracy:  0.752, Loss:  0.402
Epoch   3 Batch    6/137 - Train Accuracy:  0.753, Validation Accuracy:  0.763, Loss:  0.372
Epoch   3 Batch    7/137 - Train Accuracy:  0.758, Validation Accuracy:  0.763, Loss:  0.389
Epoch   3 Batch    8/137 - Train Accuracy:  0.761, Validation Accuracy:  0.766, Loss:  0.390
Epoch   3 Batch    9/137 - Train Accuracy:  0.767, Validation Accuracy:  0.762, Loss:  0.367
Epoch   3 Batch   10/137 - Train Accuracy:  0.760, Validation Accuracy:  0.760, Loss:  0.394
Epoch   3 Batch   11/137 - Train Accuracy:  0.765, Validation Accuracy:  0.760, Loss:  0.384
Epoch   3 Batch   12/137 - Train Accuracy:  0.740, Validation Accuracy:  0.756, Loss:  0.386
Epoch   3 Batch   13/137 - Train Accuracy:  0.756, Validation Accuracy:  0.754, Loss:  0.367
Epoch   3 Batch   14/137 - Train Accuracy:  0.734, Validation Accuracy:  0.753, Loss:  0.406
Epoch   3 Batch   15/137 - Train Accuracy:  0.745, Validation Accuracy:  0.755, Loss:  0.385
Epoch   3 Batch   16/137 - Train Accuracy:  0.758, Validation Accuracy:  0.758, Loss:  0.370
Epoch   3 Batch   17/137 - Train Accuracy:  0.747, Validation Accuracy:  0.759, Loss:  0.385
Epoch   3 Batch   18/137 - Train Accuracy:  0.757, Validation Accuracy:  0.761, Loss:  0.383
Epoch   3 Batch   19/137 - Train Accuracy:  0.753, Validation Accuracy:  0.769, Loss:  0.373
Epoch   3 Batch   20/137 - Train Accuracy:  0.749, Validation Accuracy:  0.760, Loss:  0.384
Epoch   3 Batch   21/137 - Train Accuracy:  0.770, Validation Accuracy:  0.758, Loss:  0.350
Epoch   3 Batch   22/137 - Train Accuracy:  0.757, Validation Accuracy:  0.760, Loss:  0.372
Epoch   3 Batch   23/137 - Train Accuracy:  0.773, Validation Accuracy:  0.759, Loss:  0.350
Epoch   3 Batch   24/137 - Train Accuracy:  0.757, Validation Accuracy:  0.760, Loss:  0.362
Epoch   3 Batch   25/137 - Train Accuracy:  0.755, Validation Accuracy:  0.763, Loss:  0.364
Epoch   3 Batch   26/137 - Train Accuracy:  0.754, Validation Accuracy:  0.754, Loss:  0.360
Epoch   3 Batch   27/137 - Train Accuracy:  0.774, Validation Accuracy:  0.754, Loss:  0.355
Epoch   3 Batch   28/137 - Train Accuracy:  0.762, Validation Accuracy:  0.765, Loss:  0.361
Epoch   3 Batch   29/137 - Train Accuracy:  0.755, Validation Accuracy:  0.759, Loss:  0.366
Epoch   3 Batch   30/137 - Train Accuracy:  0.767, Validation Accuracy:  0.765, Loss:  0.349
Epoch   3 Batch   31/137 - Train Accuracy:  0.777, Validation Accuracy:  0.772, Loss:  0.354
Epoch   3 Batch   32/137 - Train Accuracy:  0.770, Validation Accuracy:  0.769, Loss:  0.355
Epoch   3 Batch   33/137 - Train Accuracy:  0.766, Validation Accuracy:  0.778, Loss:  0.349
Epoch   3 Batch   34/137 - Train Accuracy:  0.759, Validation Accuracy:  0.781, Loss:  0.369
Epoch   3 Batch   35/137 - Train Accuracy:  0.763, Validation Accuracy:  0.768, Loss:  0.367
Epoch   3 Batch   36/137 - Train Accuracy:  0.770, Validation Accuracy:  0.765, Loss:  0.352
Epoch   3 Batch   37/137 - Train Accuracy:  0.767, Validation Accuracy:  0.770, Loss:  0.357
Epoch   3 Batch   38/137 - Train Accuracy:  0.757, Validation Accuracy:  0.766, Loss:  0.352
Epoch   3 Batch   39/137 - Train Accuracy:  0.775, Validation Accuracy:  0.770, Loss:  0.359
Epoch   3 Batch   40/137 - Train Accuracy:  0.763, Validation Accuracy:  0.768, Loss:  0.347
Epoch   3 Batch   41/137 - Train Accuracy:  0.775, Validation Accuracy:  0.762, Loss:  0.356
Epoch   3 Batch   42/137 - Train Accuracy:  0.773, Validation Accuracy:  0.766, Loss:  0.358
Epoch   3 Batch   43/137 - Train Accuracy:  0.775, Validation Accuracy:  0.782, Loss:  0.350
Epoch   3 Batch   44/137 - Train Accuracy:  0.770, Validation Accuracy:  0.776, Loss:  0.353
Epoch   3 Batch   45/137 - Train Accuracy:  0.775, Validation Accuracy:  0.777, Loss:  0.347
Epoch   3 Batch   46/137 - Train Accuracy:  0.774, Validation Accuracy:  0.769, Loss:  0.337
Epoch   3 Batch   47/137 - Train Accuracy:  0.775, Validation Accuracy:  0.769, Loss:  0.328
Epoch   3 Batch   48/137 - Train Accuracy:  0.770, Validation Accuracy:  0.775, Loss:  0.349
Epoch   3 Batch   49/137 - Train Accuracy:  0.769, Validation Accuracy:  0.776, Loss:  0.338
Epoch   3 Batch   50/137 - Train Accuracy:  0.776, Validation Accuracy:  0.774, Loss:  0.339
Epoch   3 Batch   51/137 - Train Accuracy:  0.782, Validation Accuracy:  0.781, Loss:  0.335
Epoch   3 Batch   52/137 - Train Accuracy:  0.762, Validation Accuracy:  0.778, Loss:  0.337
Epoch   3 Batch   53/137 - Train Accuracy:  0.753, Validation Accuracy:  0.774, Loss:  0.339
Epoch   3 Batch   54/137 - Train Accuracy:  0.775, Validation Accuracy:  0.781, Loss:  0.331
Epoch   3 Batch   55/137 - Train Accuracy:  0.774, Validation Accuracy:  0.797, Loss:  0.332
Epoch   3 Batch   56/137 - Train Accuracy:  0.782, Validation Accuracy:  0.795, Loss:  0.331
Epoch   3 Batch   57/137 - Train Accuracy:  0.788, Validation Accuracy:  0.786, Loss:  0.314
Epoch   3 Batch   58/137 - Train Accuracy:  0.789, Validation Accuracy:  0.795, Loss:  0.327
Epoch   3 Batch   59/137 - Train Accuracy:  0.787, Validation Accuracy:  0.791, Loss:  0.327
Epoch   3 Batch   60/137 - Train Accuracy:  0.782, Validation Accuracy:  0.786, Loss:  0.323
Epoch   3 Batch   61/137 - Train Accuracy:  0.790, Validation Accuracy:  0.788, Loss:  0.319
Epoch   3 Batch   62/137 - Train Accuracy:  0.783, Validation Accuracy:  0.793, Loss:  0.315
Epoch   3 Batch   63/137 - Train Accuracy:  0.783, Validation Accuracy:  0.793, Loss:  0.317
Epoch   3 Batch   64/137 - Train Accuracy:  0.796, Validation Accuracy:  0.805, Loss:  0.323
Epoch   3 Batch   65/137 - Train Accuracy:  0.798, Validation Accuracy:  0.809, Loss:  0.325
Epoch   3 Batch   66/137 - Train Accuracy:  0.789, Validation Accuracy:  0.812, Loss:  0.331
Epoch   3 Batch   67/137 - Train Accuracy:  0.779, Validation Accuracy:  0.790, Loss:  0.326
Epoch   3 Batch   68/137 - Train Accuracy:  0.799, Validation Accuracy:  0.804, Loss:  0.323
Epoch   3 Batch   69/137 - Train Accuracy:  0.785, Validation Accuracy:  0.798, Loss:  0.340
Epoch   3 Batch   70/137 - Train Accuracy:  0.792, Validation Accuracy:  0.787, Loss:  0.308
Epoch   3 Batch   71/137 - Train Accuracy:  0.793, Validation Accuracy:  0.788, Loss:  0.323
Epoch   3 Batch   72/137 - Train Accuracy:  0.799, Validation Accuracy:  0.791, Loss:  0.308
Epoch   3 Batch   73/137 - Train Accuracy:  0.789, Validation Accuracy:  0.786, Loss:  0.309
Epoch   3 Batch   74/137 - Train Accuracy:  0.807, Validation Accuracy:  0.804, Loss:  0.301
Epoch   3 Batch   75/137 - Train Accuracy:  0.798, Validation Accuracy:  0.808, Loss:  0.322
Epoch   3 Batch   76/137 - Train Accuracy:  0.808, Validation Accuracy:  0.806, Loss:  0.296
Epoch   3 Batch   77/137 - Train Accuracy:  0.784, Validation Accuracy:  0.793, Loss:  0.309
Epoch   3 Batch   78/137 - Train Accuracy:  0.801, Validation Accuracy:  0.800, Loss:  0.307
Epoch   3 Batch   79/137 - Train Accuracy:  0.807, Validation Accuracy:  0.797, Loss:  0.289
Epoch   3 Batch   80/137 - Train Accuracy:  0.794, Validation Accuracy:  0.798, Loss:  0.308
Epoch   3 Batch   81/137 - Train Accuracy:  0.790, Validation Accuracy:  0.801, Loss:  0.300
Epoch   3 Batch   82/137 - Train Accuracy:  0.811, Validation Accuracy:  0.806, Loss:  0.303
Epoch   3 Batch   83/137 - Train Accuracy:  0.821, Validation Accuracy:  0.817, Loss:  0.305
Epoch   3 Batch   84/137 - Train Accuracy:  0.828, Validation Accuracy:  0.818, Loss:  0.285
Epoch   3 Batch   85/137 - Train Accuracy:  0.822, Validation Accuracy:  0.815, Loss:  0.302
Epoch   3 Batch   86/137 - Train Accuracy:  0.797, Validation Accuracy:  0.808, Loss:  0.301
Epoch   3 Batch   87/137 - Train Accuracy:  0.809, Validation Accuracy:  0.806, Loss:  0.303
Epoch   3 Batch   88/137 - Train Accuracy:  0.823, Validation Accuracy:  0.817, Loss:  0.287
Epoch   3 Batch   89/137 - Train Accuracy:  0.796, Validation Accuracy:  0.805, Loss:  0.319
Epoch   3 Batch   90/137 - Train Accuracy:  0.805, Validation Accuracy:  0.796, Loss:  0.280
Epoch   3 Batch   91/137 - Train Accuracy:  0.789, Validation Accuracy:  0.802, Loss:  0.304
Epoch   3 Batch   92/137 - Train Accuracy:  0.806, Validation Accuracy:  0.822, Loss:  0.295
Epoch   3 Batch   93/137 - Train Accuracy:  0.838, Validation Accuracy:  0.816, Loss:  0.256
Epoch   3 Batch   94/137 - Train Accuracy:  0.806, Validation Accuracy:  0.809, Loss:  0.286
Epoch   3 Batch   95/137 - Train Accuracy:  0.810, Validation Accuracy:  0.813, Loss:  0.290
Epoch   3 Batch   96/137 - Train Accuracy:  0.808, Validation Accuracy:  0.814, Loss:  0.283
Epoch   3 Batch   97/137 - Train Accuracy:  0.811, Validation Accuracy:  0.806, Loss:  0.282
Epoch   3 Batch   98/137 - Train Accuracy:  0.822, Validation Accuracy:  0.818, Loss:  0.291
Epoch   3 Batch   99/137 - Train Accuracy:  0.824, Validation Accuracy:  0.823, Loss:  0.285
Epoch   3 Batch  100/137 - Train Accuracy:  0.797, Validation Accuracy:  0.812, Loss:  0.308
Epoch   3 Batch  101/137 - Train Accuracy:  0.793, Validation Accuracy:  0.805, Loss:  0.289
Epoch   3 Batch  102/137 - Train Accuracy:  0.814, Validation Accuracy:  0.814, Loss:  0.277
Epoch   3 Batch  103/137 - Train Accuracy:  0.818, Validation Accuracy:  0.829, Loss:  0.291
Epoch   3 Batch  104/137 - Train Accuracy:  0.818, Validation Accuracy:  0.821, Loss:  0.281
Epoch   3 Batch  105/137 - Train Accuracy:  0.815, Validation Accuracy:  0.817, Loss:  0.275
Epoch   3 Batch  106/137 - Train Accuracy:  0.818, Validation Accuracy:  0.826, Loss:  0.289
Epoch   3 Batch  107/137 - Train Accuracy:  0.829, Validation Accuracy:  0.838, Loss:  0.288
Epoch   3 Batch  108/137 - Train Accuracy:  0.818, Validation Accuracy:  0.839, Loss:  0.294
Epoch   3 Batch  109/137 - Train Accuracy:  0.838, Validation Accuracy:  0.830, Loss:  0.268
Epoch   3 Batch  110/137 - Train Accuracy:  0.821, Validation Accuracy:  0.830, Loss:  0.291
Epoch   3 Batch  111/137 - Train Accuracy:  0.825, Validation Accuracy:  0.840, Loss:  0.291
Epoch   3 Batch  112/137 - Train Accuracy:  0.839, Validation Accuracy:  0.842, Loss:  0.268
Epoch   3 Batch  113/137 - Train Accuracy:  0.834, Validation Accuracy:  0.840, Loss:  0.272
Epoch   3 Batch  114/137 - Train Accuracy:  0.826, Validation Accuracy:  0.838, Loss:  0.280
Epoch   3 Batch  115/137 - Train Accuracy:  0.841, Validation Accuracy:  0.845, Loss:  0.266
Epoch   3 Batch  116/137 - Train Accuracy:  0.853, Validation Accuracy:  0.843, Loss:  0.244
Epoch   3 Batch  117/137 - Train Accuracy:  0.834, Validation Accuracy:  0.844, Loss:  0.278
Epoch   3 Batch  118/137 - Train Accuracy:  0.816, Validation Accuracy:  0.841, Loss:  0.290
Epoch   3 Batch  119/137 - Train Accuracy:  0.834, Validation Accuracy:  0.837, Loss:  0.271
Epoch   3 Batch  120/137 - Train Accuracy:  0.836, Validation Accuracy:  0.837, Loss:  0.263
Epoch   3 Batch  121/137 - Train Accuracy:  0.848, Validation Accuracy:  0.838, Loss:  0.263
Epoch   3 Batch  122/137 - Train Accuracy:  0.852, Validation Accuracy:  0.840, Loss:  0.246
Epoch   3 Batch  123/137 - Train Accuracy:  0.834, Validation Accuracy:  0.842, Loss:  0.273
Epoch   3 Batch  124/137 - Train Accuracy:  0.850, Validation Accuracy:  0.847, Loss:  0.260
Epoch   3 Batch  125/137 - Train Accuracy:  0.836, Validation Accuracy:  0.845, Loss:  0.266
Epoch   3 Batch  126/137 - Train Accuracy:  0.828, Validation Accuracy:  0.843, Loss:  0.261
Epoch   3 Batch  127/137 - Train Accuracy:  0.862, Validation Accuracy:  0.852, Loss:  0.241
Epoch   3 Batch  128/137 - Train Accuracy:  0.860, Validation Accuracy:  0.846, Loss:  0.257
Epoch   3 Batch  129/137 - Train Accuracy:  0.823, Validation Accuracy:  0.851, Loss:  0.268
Epoch   3 Batch  130/137 - Train Accuracy:  0.834, Validation Accuracy:  0.852, Loss:  0.254
Epoch   3 Batch  131/137 - Train Accuracy:  0.834, Validation Accuracy:  0.839, Loss:  0.267
Epoch   3 Batch  132/137 - Train Accuracy:  0.852, Validation Accuracy:  0.845, Loss:  0.262
Epoch   3 Batch  133/137 - Train Accuracy:  0.833, Validation Accuracy:  0.837, Loss:  0.258
Epoch   3 Batch  134/137 - Train Accuracy:  0.842, Validation Accuracy:  0.854, Loss:  0.271
Epoch   3 Batch  135/137 - Train Accuracy:  0.842, Validation Accuracy:  0.845, Loss:  0.270
Epoch   4 Batch    0/137 - Train Accuracy:  0.835, Validation Accuracy:  0.836, Loss:  0.253
Epoch   4 Batch    1/137 - Train Accuracy:  0.818, Validation Accuracy:  0.838, Loss:  0.264
Epoch   4 Batch    2/137 - Train Accuracy:  0.838, Validation Accuracy:  0.840, Loss:  0.252
Epoch   4 Batch    3/137 - Train Accuracy:  0.848, Validation Accuracy:  0.846, Loss:  0.251
Epoch   4 Batch    4/137 - Train Accuracy:  0.854, Validation Accuracy:  0.857, Loss:  0.263
Epoch   4 Batch    5/137 - Train Accuracy:  0.846, Validation Accuracy:  0.849, Loss:  0.259
Epoch   4 Batch    6/137 - Train Accuracy:  0.839, Validation Accuracy:  0.849, Loss:  0.251
Epoch   4 Batch    7/137 - Train Accuracy:  0.846, Validation Accuracy:  0.851, Loss:  0.248
Epoch   4 Batch    8/137 - Train Accuracy:  0.839, Validation Accuracy:  0.843, Loss:  0.252
Epoch   4 Batch    9/137 - Train Accuracy:  0.851, Validation Accuracy:  0.847, Loss:  0.235
Epoch   4 Batch   10/137 - Train Accuracy:  0.842, Validation Accuracy:  0.851, Loss:  0.251
Epoch   4 Batch   11/137 - Train Accuracy:  0.857, Validation Accuracy:  0.861, Loss:  0.248
Epoch   4 Batch   12/137 - Train Accuracy:  0.841, Validation Accuracy:  0.854, Loss:  0.249
Epoch   4 Batch   13/137 - Train Accuracy:  0.859, Validation Accuracy:  0.864, Loss:  0.232
Epoch   4 Batch   14/137 - Train Accuracy:  0.840, Validation Accuracy:  0.861, Loss:  0.260
Epoch   4 Batch   15/137 - Train Accuracy:  0.854, Validation Accuracy:  0.859, Loss:  0.243
Epoch   4 Batch   16/137 - Train Accuracy:  0.858, Validation Accuracy:  0.862, Loss:  0.233
Epoch   4 Batch   17/137 - Train Accuracy:  0.852, Validation Accuracy:  0.865, Loss:  0.251
Epoch   4 Batch   18/137 - Train Accuracy:  0.857, Validation Accuracy:  0.859, Loss:  0.245
Epoch   4 Batch   19/137 - Train Accuracy:  0.850, Validation Accuracy:  0.864, Loss:  0.242
Epoch   4 Batch   20/137 - Train Accuracy:  0.846, Validation Accuracy:  0.866, Loss:  0.252
Epoch   4 Batch   21/137 - Train Accuracy:  0.866, Validation Accuracy:  0.865, Loss:  0.225
Epoch   4 Batch   22/137 - Train Accuracy:  0.859, Validation Accuracy:  0.861, Loss:  0.244
Epoch   4 Batch   23/137 - Train Accuracy:  0.859, Validation Accuracy:  0.858, Loss:  0.225
Epoch   4 Batch   24/137 - Train Accuracy:  0.864, Validation Accuracy:  0.865, Loss:  0.233
Epoch   4 Batch   25/137 - Train Accuracy:  0.846, Validation Accuracy:  0.855, Loss:  0.237
Epoch   4 Batch   26/137 - Train Accuracy:  0.853, Validation Accuracy:  0.863, Loss:  0.231
Epoch   4 Batch   27/137 - Train Accuracy:  0.870, Validation Accuracy:  0.871, Loss:  0.229
Epoch   4 Batch   28/137 - Train Accuracy:  0.854, Validation Accuracy:  0.867, Loss:  0.241
Epoch   4 Batch   29/137 - Train Accuracy:  0.863, Validation Accuracy:  0.868, Loss:  0.234
Epoch   4 Batch   30/137 - Train Accuracy:  0.875, Validation Accuracy:  0.872, Loss:  0.217
Epoch   4 Batch   31/137 - Train Accuracy:  0.867, Validation Accuracy:  0.873, Loss:  0.230
Epoch   4 Batch   32/137 - Train Accuracy:  0.856, Validation Accuracy:  0.870, Loss:  0.231
Epoch   4 Batch   33/137 - Train Accuracy:  0.852, Validation Accuracy:  0.866, Loss:  0.223
Epoch   4 Batch   34/137 - Train Accuracy:  0.834, Validation Accuracy:  0.865, Loss:  0.240
Epoch   4 Batch   35/137 - Train Accuracy:  0.856, Validation Accuracy:  0.871, Loss:  0.241
Epoch   4 Batch   36/137 - Train Accuracy:  0.875, Validation Accuracy:  0.874, Loss:  0.227
Epoch   4 Batch   37/137 - Train Accuracy:  0.870, Validation Accuracy:  0.874, Loss:  0.226
Epoch   4 Batch   38/137 - Train Accuracy:  0.861, Validation Accuracy:  0.875, Loss:  0.231
Epoch   4 Batch   39/137 - Train Accuracy:  0.873, Validation Accuracy:  0.875, Loss:  0.227
Epoch   4 Batch   40/137 - Train Accuracy:  0.866, Validation Accuracy:  0.879, Loss:  0.228
Epoch   4 Batch   41/137 - Train Accuracy:  0.878, Validation Accuracy:  0.876, Loss:  0.223
Epoch   4 Batch   42/137 - Train Accuracy:  0.862, Validation Accuracy:  0.875, Loss:  0.231
Epoch   4 Batch   43/137 - Train Accuracy:  0.858, Validation Accuracy:  0.873, Loss:  0.221
Epoch   4 Batch   44/137 - Train Accuracy:  0.863, Validation Accuracy:  0.874, Loss:  0.230
Epoch   4 Batch   45/137 - Train Accuracy:  0.865, Validation Accuracy:  0.873, Loss:  0.221
Epoch   4 Batch   46/137 - Train Accuracy:  0.876, Validation Accuracy:  0.873, Loss:  0.218
Epoch   4 Batch   47/137 - Train Accuracy:  0.871, Validation Accuracy:  0.868, Loss:  0.216
Epoch   4 Batch   48/137 - Train Accuracy:  0.857, Validation Accuracy:  0.875, Loss:  0.236
Epoch   4 Batch   49/137 - Train Accuracy:  0.863, Validation Accuracy:  0.878, Loss:  0.226
Epoch   4 Batch   50/137 - Train Accuracy:  0.868, Validation Accuracy:  0.875, Loss:  0.218
Epoch   4 Batch   51/137 - Train Accuracy:  0.862, Validation Accuracy:  0.868, Loss:  0.225
Epoch   4 Batch   52/137 - Train Accuracy:  0.873, Validation Accuracy:  0.881, Loss:  0.223
Epoch   4 Batch   53/137 - Train Accuracy:  0.856, Validation Accuracy:  0.875, Loss:  0.219
Epoch   4 Batch   54/137 - Train Accuracy:  0.871, Validation Accuracy:  0.877, Loss:  0.216
Epoch   4 Batch   55/137 - Train Accuracy:  0.862, Validation Accuracy:  0.877, Loss:  0.214
Epoch   4 Batch   56/137 - Train Accuracy:  0.868, Validation Accuracy:  0.873, Loss:  0.211
Epoch   4 Batch   57/137 - Train Accuracy:  0.877, Validation Accuracy:  0.877, Loss:  0.210
Epoch   4 Batch   58/137 - Train Accuracy:  0.869, Validation Accuracy:  0.873, Loss:  0.211
Epoch   4 Batch   59/137 - Train Accuracy:  0.873, Validation Accuracy:  0.882, Loss:  0.212
Epoch   4 Batch   60/137 - Train Accuracy:  0.864, Validation Accuracy:  0.878, Loss:  0.205
Epoch   4 Batch   61/137 - Train Accuracy:  0.872, Validation Accuracy:  0.884, Loss:  0.209
Epoch   4 Batch   62/137 - Train Accuracy:  0.869, Validation Accuracy:  0.874, Loss:  0.215
Epoch   4 Batch   63/137 - Train Accuracy:  0.875, Validation Accuracy:  0.874, Loss:  0.206
Epoch   4 Batch   64/137 - Train Accuracy:  0.866, Validation Accuracy:  0.875, Loss:  0.204
Epoch   4 Batch   65/137 - Train Accuracy:  0.864, Validation Accuracy:  0.874, Loss:  0.214
Epoch   4 Batch   66/137 - Train Accuracy:  0.851, Validation Accuracy:  0.869, Loss:  0.226
Epoch   4 Batch   67/137 - Train Accuracy:  0.866, Validation Accuracy:  0.881, Loss:  0.215
Epoch   4 Batch   68/137 - Train Accuracy:  0.880, Validation Accuracy:  0.889, Loss:  0.209
Epoch   4 Batch   69/137 - Train Accuracy:  0.859, Validation Accuracy:  0.887, Loss:  0.223
Epoch   4 Batch   70/137 - Train Accuracy:  0.874, Validation Accuracy:  0.884, Loss:  0.204
Epoch   4 Batch   71/137 - Train Accuracy:  0.875, Validation Accuracy:  0.885, Loss:  0.214
Epoch   4 Batch   72/137 - Train Accuracy:  0.878, Validation Accuracy:  0.879, Loss:  0.200
Epoch   4 Batch   73/137 - Train Accuracy:  0.885, Validation Accuracy:  0.884, Loss:  0.203
Epoch   4 Batch   74/137 - Train Accuracy:  0.885, Validation Accuracy:  0.887, Loss:  0.192
Epoch   4 Batch   75/137 - Train Accuracy:  0.863, Validation Accuracy:  0.878, Loss:  0.218
Epoch   4 Batch   76/137 - Train Accuracy:  0.874, Validation Accuracy:  0.877, Loss:  0.193
Epoch   4 Batch   77/137 - Train Accuracy:  0.868, Validation Accuracy:  0.881, Loss:  0.205
Epoch   4 Batch   78/137 - Train Accuracy:  0.886, Validation Accuracy:  0.885, Loss:  0.199
Epoch   4 Batch   79/137 - Train Accuracy:  0.878, Validation Accuracy:  0.891, Loss:  0.192
Epoch   4 Batch   80/137 - Train Accuracy:  0.871, Validation Accuracy:  0.886, Loss:  0.202
Epoch   4 Batch   81/137 - Train Accuracy:  0.867, Validation Accuracy:  0.881, Loss:  0.197
Epoch   4 Batch   82/137 - Train Accuracy:  0.887, Validation Accuracy:  0.886, Loss:  0.200
Epoch   4 Batch   83/137 - Train Accuracy:  0.885, Validation Accuracy:  0.888, Loss:  0.200
Epoch   4 Batch   84/137 - Train Accuracy:  0.887, Validation Accuracy:  0.887, Loss:  0.188
Epoch   4 Batch   85/137 - Train Accuracy:  0.874, Validation Accuracy:  0.884, Loss:  0.196
Epoch   4 Batch   86/137 - Train Accuracy:  0.860, Validation Accuracy:  0.885, Loss:  0.202
Epoch   4 Batch   87/137 - Train Accuracy:  0.869, Validation Accuracy:  0.884, Loss:  0.209
Epoch   4 Batch   88/137 - Train Accuracy:  0.884, Validation Accuracy:  0.885, Loss:  0.193
Epoch   4 Batch   89/137 - Train Accuracy:  0.861, Validation Accuracy:  0.883, Loss:  0.210
Epoch   4 Batch   90/137 - Train Accuracy:  0.878, Validation Accuracy:  0.884, Loss:  0.183
Epoch   4 Batch   91/137 - Train Accuracy:  0.861, Validation Accuracy:  0.881, Loss:  0.196
Epoch   4 Batch   92/137 - Train Accuracy:  0.866, Validation Accuracy:  0.889, Loss:  0.200
Epoch   4 Batch   93/137 - Train Accuracy:  0.896, Validation Accuracy:  0.889, Loss:  0.167
Epoch   4 Batch   94/137 - Train Accuracy:  0.881, Validation Accuracy:  0.882, Loss:  0.190
Epoch   4 Batch   95/137 - Train Accuracy:  0.879, Validation Accuracy:  0.886, Loss:  0.191
Epoch   4 Batch   96/137 - Train Accuracy:  0.881, Validation Accuracy:  0.887, Loss:  0.182
Epoch   4 Batch   97/137 - Train Accuracy:  0.871, Validation Accuracy:  0.892, Loss:  0.192
Epoch   4 Batch   98/137 - Train Accuracy:  0.868, Validation Accuracy:  0.884, Loss:  0.188
Epoch   4 Batch   99/137 - Train Accuracy:  0.875, Validation Accuracy:  0.894, Loss:  0.198
Epoch   4 Batch  100/137 - Train Accuracy:  0.872, Validation Accuracy:  0.894, Loss:  0.200
Epoch   4 Batch  101/137 - Train Accuracy:  0.879, Validation Accuracy:  0.892, Loss:  0.193
Epoch   4 Batch  102/137 - Train Accuracy:  0.880, Validation Accuracy:  0.890, Loss:  0.186
Epoch   4 Batch  103/137 - Train Accuracy:  0.880, Validation Accuracy:  0.892, Loss:  0.193
Epoch   4 Batch  104/137 - Train Accuracy:  0.882, Validation Accuracy:  0.888, Loss:  0.185
Epoch   4 Batch  105/137 - Train Accuracy:  0.876, Validation Accuracy:  0.891, Loss:  0.185
Epoch   4 Batch  106/137 - Train Accuracy:  0.882, Validation Accuracy:  0.889, Loss:  0.187
Epoch   4 Batch  107/137 - Train Accuracy:  0.879, Validation Accuracy:  0.888, Loss:  0.181
Epoch   4 Batch  108/137 - Train Accuracy:  0.876, Validation Accuracy:  0.892, Loss:  0.202
Epoch   4 Batch  109/137 - Train Accuracy:  0.887, Validation Accuracy:  0.894, Loss:  0.175
Epoch   4 Batch  110/137 - Train Accuracy:  0.870, Validation Accuracy:  0.895, Loss:  0.193
Epoch   4 Batch  111/137 - Train Accuracy:  0.879, Validation Accuracy:  0.893, Loss:  0.191
Epoch   4 Batch  112/137 - Train Accuracy:  0.881, Validation Accuracy:  0.889, Loss:  0.176
Epoch   4 Batch  113/137 - Train Accuracy:  0.882, Validation Accuracy:  0.888, Loss:  0.182
Epoch   4 Batch  114/137 - Train Accuracy:  0.879, Validation Accuracy:  0.892, Loss:  0.188
Epoch   4 Batch  115/137 - Train Accuracy:  0.893, Validation Accuracy:  0.893, Loss:  0.176
Epoch   4 Batch  116/137 - Train Accuracy:  0.899, Validation Accuracy:  0.893, Loss:  0.163
Epoch   4 Batch  117/137 - Train Accuracy:  0.878, Validation Accuracy:  0.888, Loss:  0.190
Epoch   4 Batch  118/137 - Train Accuracy:  0.868, Validation Accuracy:  0.890, Loss:  0.196
Epoch   4 Batch  119/137 - Train Accuracy:  0.893, Validation Accuracy:  0.893, Loss:  0.181
Epoch   4 Batch  120/137 - Train Accuracy:  0.892, Validation Accuracy:  0.895, Loss:  0.169
Epoch   4 Batch  121/137 - Train Accuracy:  0.889, Validation Accuracy:  0.894, Loss:  0.172
Epoch   4 Batch  122/137 - Train Accuracy:  0.899, Validation Accuracy:  0.894, Loss:  0.164
Epoch   4 Batch  123/137 - Train Accuracy:  0.884, Validation Accuracy:  0.893, Loss:  0.184
Epoch   4 Batch  124/137 - Train Accuracy:  0.893, Validation Accuracy:  0.897, Loss:  0.181
Epoch   4 Batch  125/137 - Train Accuracy:  0.878, Validation Accuracy:  0.899, Loss:  0.176
Epoch   4 Batch  126/137 - Train Accuracy:  0.888, Validation Accuracy:  0.893, Loss:  0.173
Epoch   4 Batch  127/137 - Train Accuracy:  0.902, Validation Accuracy:  0.899, Loss:  0.160
Epoch   4 Batch  128/137 - Train Accuracy:  0.903, Validation Accuracy:  0.897, Loss:  0.170
Epoch   4 Batch  129/137 - Train Accuracy:  0.873, Validation Accuracy:  0.892, Loss:  0.181
Epoch   4 Batch  130/137 - Train Accuracy:  0.883, Validation Accuracy:  0.897, Loss:  0.170
Epoch   4 Batch  131/137 - Train Accuracy:  0.873, Validation Accuracy:  0.901, Loss:  0.187
Epoch   4 Batch  132/137 - Train Accuracy:  0.894, Validation Accuracy:  0.898, Loss:  0.178
Epoch   4 Batch  133/137 - Train Accuracy:  0.885, Validation Accuracy:  0.896, Loss:  0.175
Epoch   4 Batch  134/137 - Train Accuracy:  0.876, Validation Accuracy:  0.900, Loss:  0.188
Epoch   4 Batch  135/137 - Train Accuracy:  0.890, Validation Accuracy:  0.902, Loss:  0.174
Epoch   5 Batch    0/137 - Train Accuracy:  0.891, Validation Accuracy:  0.898, Loss:  0.176
Epoch   5 Batch    1/137 - Train Accuracy:  0.880, Validation Accuracy:  0.899, Loss:  0.180
Epoch   5 Batch    2/137 - Train Accuracy:  0.899, Validation Accuracy:  0.902, Loss:  0.170
Epoch   5 Batch    3/137 - Train Accuracy:  0.897, Validation Accuracy:  0.905, Loss:  0.168
Epoch   5 Batch    4/137 - Train Accuracy:  0.889, Validation Accuracy:  0.905, Loss:  0.180
Epoch   5 Batch    5/137 - Train Accuracy:  0.897, Validation Accuracy:  0.898, Loss:  0.172
Epoch   5 Batch    6/137 - Train Accuracy:  0.888, Validation Accuracy:  0.899, Loss:  0.165
Epoch   5 Batch    7/137 - Train Accuracy:  0.890, Validation Accuracy:  0.899, Loss:  0.168
Epoch   5 Batch    8/137 - Train Accuracy:  0.891, Validation Accuracy:  0.899, Loss:  0.172
Epoch   5 Batch    9/137 - Train Accuracy:  0.896, Validation Accuracy:  0.902, Loss:  0.158
Epoch   5 Batch   10/137 - Train Accuracy:  0.891, Validation Accuracy:  0.901, Loss:  0.170
Epoch   5 Batch   11/137 - Train Accuracy:  0.894, Validation Accuracy:  0.899, Loss:  0.165
Epoch   5 Batch   12/137 - Train Accuracy:  0.878, Validation Accuracy:  0.902, Loss:  0.171
Epoch   5 Batch   13/137 - Train Accuracy:  0.889, Validation Accuracy:  0.902, Loss:  0.153
Epoch   5 Batch   14/137 - Train Accuracy:  0.892, Validation Accuracy:  0.904, Loss:  0.174
Epoch   5 Batch   15/137 - Train Accuracy:  0.891, Validation Accuracy:  0.903, Loss:  0.166
Epoch   5 Batch   16/137 - Train Accuracy:  0.899, Validation Accuracy:  0.903, Loss:  0.155
Epoch   5 Batch   17/137 - Train Accuracy:  0.893, Validation Accuracy:  0.898, Loss:  0.169
Epoch   5 Batch   18/137 - Train Accuracy:  0.891, Validation Accuracy:  0.903, Loss:  0.166
Epoch   5 Batch   19/137 - Train Accuracy:  0.884, Validation Accuracy:  0.904, Loss:  0.165
Epoch   5 Batch   20/137 - Train Accuracy:  0.886, Validation Accuracy:  0.904, Loss:  0.172
Epoch   5 Batch   21/137 - Train Accuracy:  0.903, Validation Accuracy:  0.897, Loss:  0.157
Epoch   5 Batch   22/137 - Train Accuracy:  0.892, Validation Accuracy:  0.898, Loss:  0.164
Epoch   5 Batch   23/137 - Train Accuracy:  0.897, Validation Accuracy:  0.897, Loss:  0.145
Epoch   5 Batch   24/137 - Train Accuracy:  0.886, Validation Accuracy:  0.897, Loss:  0.160
Epoch   5 Batch   25/137 - Train Accuracy:  0.888, Validation Accuracy:  0.900, Loss:  0.162
Epoch   5 Batch   26/137 - Train Accuracy:  0.882, Validation Accuracy:  0.895, Loss:  0.157
Epoch   5 Batch   27/137 - Train Accuracy:  0.904, Validation Accuracy:  0.904, Loss:  0.162
Epoch   5 Batch   28/137 - Train Accuracy:  0.894, Validation Accuracy:  0.899, Loss:  0.160
Epoch   5 Batch   29/137 - Train Accuracy:  0.889, Validation Accuracy:  0.893, Loss:  0.162
Epoch   5 Batch   30/137 - Train Accuracy:  0.905, Validation Accuracy:  0.901, Loss:  0.155
Epoch   5 Batch   31/137 - Train Accuracy:  0.896, Validation Accuracy:  0.900, Loss:  0.154
Epoch   5 Batch   32/137 - Train Accuracy:  0.899, Validation Accuracy:  0.900, Loss:  0.165
Epoch   5 Batch   33/137 - Train Accuracy:  0.887, Validation Accuracy:  0.904, Loss:  0.155
Epoch   5 Batch   34/137 - Train Accuracy:  0.879, Validation Accuracy:  0.903, Loss:  0.174
Epoch   5 Batch   35/137 - Train Accuracy:  0.891, Validation Accuracy:  0.903, Loss:  0.174
Epoch   5 Batch   36/137 - Train Accuracy:  0.902, Validation Accuracy:  0.900, Loss:  0.162
Epoch   5 Batch   37/137 - Train Accuracy:  0.904, Validation Accuracy:  0.900, Loss:  0.158
Epoch   5 Batch   38/137 - Train Accuracy:  0.890, Validation Accuracy:  0.902, Loss:  0.159
Epoch   5 Batch   39/137 - Train Accuracy:  0.902, Validation Accuracy:  0.904, Loss:  0.156
Epoch   5 Batch   40/137 - Train Accuracy:  0.889, Validation Accuracy:  0.903, Loss:  0.156
Epoch   5 Batch   41/137 - Train Accuracy:  0.904, Validation Accuracy:  0.905, Loss:  0.153
Epoch   5 Batch   42/137 - Train Accuracy:  0.893, Validation Accuracy:  0.904, Loss:  0.166
Epoch   5 Batch   43/137 - Train Accuracy:  0.890, Validation Accuracy:  0.907, Loss:  0.158
Epoch   5 Batch   44/137 - Train Accuracy:  0.895, Validation Accuracy:  0.907, Loss:  0.158
Epoch   5 Batch   45/137 - Train Accuracy:  0.904, Validation Accuracy:  0.904, Loss:  0.154
Epoch   5 Batch   46/137 - Train Accuracy:  0.915, Validation Accuracy:  0.904, Loss:  0.145
Epoch   5 Batch   47/137 - Train Accuracy:  0.913, Validation Accuracy:  0.909, Loss:  0.148
Epoch   5 Batch   48/137 - Train Accuracy:  0.895, Validation Accuracy:  0.910, Loss:  0.158
Epoch   5 Batch   49/137 - Train Accuracy:  0.893, Validation Accuracy:  0.909, Loss:  0.157
Epoch   5 Batch   50/137 - Train Accuracy:  0.899, Validation Accuracy:  0.905, Loss:  0.152
Epoch   5 Batch   51/137 - Train Accuracy:  0.905, Validation Accuracy:  0.908, Loss:  0.154
Epoch   5 Batch   52/137 - Train Accuracy:  0.903, Validation Accuracy:  0.906, Loss:  0.148
Epoch   5 Batch   53/137 - Train Accuracy:  0.894, Validation Accuracy:  0.909, Loss:  0.148
Epoch   5 Batch   54/137 - Train Accuracy:  0.910, Validation Accuracy:  0.911, Loss:  0.144
Epoch   5 Batch   55/137 - Train Accuracy:  0.893, Validation Accuracy:  0.909, Loss:  0.146
Epoch   5 Batch   56/137 - Train Accuracy:  0.900, Validation Accuracy:  0.913, Loss:  0.143
Epoch   5 Batch   57/137 - Train Accuracy:  0.899, Validation Accuracy:  0.912, Loss:  0.142
Epoch   5 Batch   58/137 - Train Accuracy:  0.902, Validation Accuracy:  0.911, Loss:  0.140
Epoch   5 Batch   59/137 - Train Accuracy:  0.905, Validation Accuracy:  0.914, Loss:  0.143
Epoch   5 Batch   60/137 - Train Accuracy:  0.900, Validation Accuracy:  0.911, Loss:  0.139
Epoch   5 Batch   61/137 - Train Accuracy:  0.909, Validation Accuracy:  0.914, Loss:  0.144
Epoch   5 Batch   62/137 - Train Accuracy:  0.905, Validation Accuracy:  0.916, Loss:  0.138
Epoch   5 Batch   63/137 - Train Accuracy:  0.908, Validation Accuracy:  0.918, Loss:  0.135
Epoch   5 Batch   64/137 - Train Accuracy:  0.903, Validation Accuracy:  0.917, Loss:  0.140
Epoch   5 Batch   65/137 - Train Accuracy:  0.902, Validation Accuracy:  0.920, Loss:  0.137
Epoch   5 Batch   66/137 - Train Accuracy:  0.905, Validation Accuracy:  0.918, Loss:  0.146
Epoch   5 Batch   67/137 - Train Accuracy:  0.906, Validation Accuracy:  0.916, Loss:  0.138
Epoch   5 Batch   68/137 - Train Accuracy:  0.911, Validation Accuracy:  0.919, Loss:  0.134
Epoch   5 Batch   69/137 - Train Accuracy:  0.895, Validation Accuracy:  0.918, Loss:  0.154
Epoch   5 Batch   70/137 - Train Accuracy:  0.905, Validation Accuracy:  0.919, Loss:  0.130
Epoch   5 Batch   71/137 - Train Accuracy:  0.903, Validation Accuracy:  0.921, Loss:  0.139
Epoch   5 Batch   72/137 - Train Accuracy:  0.912, Validation Accuracy:  0.920, Loss:  0.132
Epoch   5 Batch   73/137 - Train Accuracy:  0.916, Validation Accuracy:  0.916, Loss:  0.126
Epoch   5 Batch   74/137 - Train Accuracy:  0.914, Validation Accuracy:  0.919, Loss:  0.124
Epoch   5 Batch   75/137 - Train Accuracy:  0.901, Validation Accuracy:  0.918, Loss:  0.142
Epoch   5 Batch   76/137 - Train Accuracy:  0.907, Validation Accuracy:  0.920, Loss:  0.135
Epoch   5 Batch   77/137 - Train Accuracy:  0.904, Validation Accuracy:  0.921, Loss:  0.134
Epoch   5 Batch   78/137 - Train Accuracy:  0.926, Validation Accuracy:  0.923, Loss:  0.130
Epoch   5 Batch   79/137 - Train Accuracy:  0.906, Validation Accuracy:  0.920, Loss:  0.124
Epoch   5 Batch   80/137 - Train Accuracy:  0.912, Validation Accuracy:  0.924, Loss:  0.128
Epoch   5 Batch   81/137 - Train Accuracy:  0.910, Validation Accuracy:  0.925, Loss:  0.131
Epoch   5 Batch   82/137 - Train Accuracy:  0.921, Validation Accuracy:  0.925, Loss:  0.122
Epoch   5 Batch   83/137 - Train Accuracy:  0.921, Validation Accuracy:  0.922, Loss:  0.125
Epoch   5 Batch   84/137 - Train Accuracy:  0.921, Validation Accuracy:  0.922, Loss:  0.115
Epoch   5 Batch   85/137 - Train Accuracy:  0.913, Validation Accuracy:  0.926, Loss:  0.133
Epoch   5 Batch   86/137 - Train Accuracy:  0.903, Validation Accuracy:  0.928, Loss:  0.125
Epoch   5 Batch   87/137 - Train Accuracy:  0.916, Validation Accuracy:  0.926, Loss:  0.127
Epoch   5 Batch   88/137 - Train Accuracy:  0.923, Validation Accuracy:  0.924, Loss:  0.118
Epoch   5 Batch   89/137 - Train Accuracy:  0.906, Validation Accuracy:  0.923, Loss:  0.139
Epoch   5 Batch   90/137 - Train Accuracy:  0.924, Validation Accuracy:  0.921, Loss:  0.112
Epoch   5 Batch   91/137 - Train Accuracy:  0.904, Validation Accuracy:  0.924, Loss:  0.128
Epoch   5 Batch   92/137 - Train Accuracy:  0.909, Validation Accuracy:  0.924, Loss:  0.128
Epoch   5 Batch   93/137 - Train Accuracy:  0.931, Validation Accuracy:  0.922, Loss:  0.106
Epoch   5 Batch   94/137 - Train Accuracy:  0.922, Validation Accuracy:  0.923, Loss:  0.121
Epoch   5 Batch   95/137 - Train Accuracy:  0.918, Validation Accuracy:  0.921, Loss:  0.122
Epoch   5 Batch   96/137 - Train Accuracy:  0.918, Validation Accuracy:  0.921, Loss:  0.120
Epoch   5 Batch   97/137 - Train Accuracy:  0.909, Validation Accuracy:  0.927, Loss:  0.123
Epoch   5 Batch   98/137 - Train Accuracy:  0.919, Validation Accuracy:  0.924, Loss:  0.121
Epoch   5 Batch   99/137 - Train Accuracy:  0.909, Validation Accuracy:  0.924, Loss:  0.126
Epoch   5 Batch  100/137 - Train Accuracy:  0.904, Validation Accuracy:  0.922, Loss:  0.122
Epoch   5 Batch  101/137 - Train Accuracy:  0.918, Validation Accuracy:  0.921, Loss:  0.125
Epoch   5 Batch  102/137 - Train Accuracy:  0.914, Validation Accuracy:  0.919, Loss:  0.117
Epoch   5 Batch  103/137 - Train Accuracy:  0.911, Validation Accuracy:  0.922, Loss:  0.127
Epoch   5 Batch  104/137 - Train Accuracy:  0.919, Validation Accuracy:  0.921, Loss:  0.113
Epoch   5 Batch  105/137 - Train Accuracy:  0.918, Validation Accuracy:  0.920, Loss:  0.123
Epoch   5 Batch  106/137 - Train Accuracy:  0.920, Validation Accuracy:  0.919, Loss:  0.119
Epoch   5 Batch  107/137 - Train Accuracy:  0.918, Validation Accuracy:  0.923, Loss:  0.117
Epoch   5 Batch  108/137 - Train Accuracy:  0.911, Validation Accuracy:  0.923, Loss:  0.125
Epoch   5 Batch  109/137 - Train Accuracy:  0.927, Validation Accuracy:  0.925, Loss:  0.115
Epoch   5 Batch  110/137 - Train Accuracy:  0.909, Validation Accuracy:  0.921, Loss:  0.127
Epoch   5 Batch  111/137 - Train Accuracy:  0.921, Validation Accuracy:  0.923, Loss:  0.126
Epoch   5 Batch  112/137 - Train Accuracy:  0.916, Validation Accuracy:  0.925, Loss:  0.121
Epoch   5 Batch  113/137 - Train Accuracy:  0.925, Validation Accuracy:  0.927, Loss:  0.111
Epoch   5 Batch  114/137 - Train Accuracy:  0.916, Validation Accuracy:  0.927, Loss:  0.122
Epoch   5 Batch  115/137 - Train Accuracy:  0.923, Validation Accuracy:  0.928, Loss:  0.113
Epoch   5 Batch  116/137 - Train Accuracy:  0.930, Validation Accuracy:  0.924, Loss:  0.104
Epoch   5 Batch  117/137 - Train Accuracy:  0.909, Validation Accuracy:  0.921, Loss:  0.126
Epoch   5 Batch  118/137 - Train Accuracy:  0.912, Validation Accuracy:  0.928, Loss:  0.121
Epoch   5 Batch  119/137 - Train Accuracy:  0.923, Validation Accuracy:  0.925, Loss:  0.117
Epoch   5 Batch  120/137 - Train Accuracy:  0.932, Validation Accuracy:  0.929, Loss:  0.111
Epoch   5 Batch  121/137 - Train Accuracy:  0.923, Validation Accuracy:  0.929, Loss:  0.109
Epoch   5 Batch  122/137 - Train Accuracy:  0.926, Validation Accuracy:  0.929, Loss:  0.102
Epoch   5 Batch  123/137 - Train Accuracy:  0.919, Validation Accuracy:  0.928, Loss:  0.124
Epoch   5 Batch  124/137 - Train Accuracy:  0.919, Validation Accuracy:  0.928, Loss:  0.114
Epoch   5 Batch  125/137 - Train Accuracy:  0.918, Validation Accuracy:  0.927, Loss:  0.114
Epoch   5 Batch  126/137 - Train Accuracy:  0.925, Validation Accuracy:  0.928, Loss:  0.110
Epoch   5 Batch  127/137 - Train Accuracy:  0.934, Validation Accuracy:  0.927, Loss:  0.104
Epoch   5 Batch  128/137 - Train Accuracy:  0.930, Validation Accuracy:  0.929, Loss:  0.110
Epoch   5 Batch  129/137 - Train Accuracy:  0.912, Validation Accuracy:  0.930, Loss:  0.114
Epoch   5 Batch  130/137 - Train Accuracy:  0.917, Validation Accuracy:  0.929, Loss:  0.111
Epoch   5 Batch  131/137 - Train Accuracy:  0.913, Validation Accuracy:  0.929, Loss:  0.118
Epoch   5 Batch  132/137 - Train Accuracy:  0.922, Validation Accuracy:  0.929, Loss:  0.115
Epoch   5 Batch  133/137 - Train Accuracy:  0.926, Validation Accuracy:  0.928, Loss:  0.109
Epoch   5 Batch  134/137 - Train Accuracy:  0.909, Validation Accuracy:  0.926, Loss:  0.113
Epoch   5 Batch  135/137 - Train Accuracy:  0.918, Validation Accuracy:  0.930, Loss:  0.112
Epoch   6 Batch    0/137 - Train Accuracy:  0.925, Validation Accuracy:  0.931, Loss:  0.107
Epoch   6 Batch    1/137 - Train Accuracy:  0.919, Validation Accuracy:  0.932, Loss:  0.112
Epoch   6 Batch    2/137 - Train Accuracy:  0.930, Validation Accuracy:  0.932, Loss:  0.107
Epoch   6 Batch    3/137 - Train Accuracy:  0.930, Validation Accuracy:  0.931, Loss:  0.106
Epoch   6 Batch    4/137 - Train Accuracy:  0.924, Validation Accuracy:  0.930, Loss:  0.109
Epoch   6 Batch    5/137 - Train Accuracy:  0.922, Validation Accuracy:  0.931, Loss:  0.110
Epoch   6 Batch    6/137 - Train Accuracy:  0.917, Validation Accuracy:  0.932, Loss:  0.102
Epoch   6 Batch    7/137 - Train Accuracy:  0.923, Validation Accuracy:  0.932, Loss:  0.097
Epoch   6 Batch    8/137 - Train Accuracy:  0.921, Validation Accuracy:  0.932, Loss:  0.106
Epoch   6 Batch    9/137 - Train Accuracy:  0.931, Validation Accuracy:  0.935, Loss:  0.096
Epoch   6 Batch   10/137 - Train Accuracy:  0.918, Validation Accuracy:  0.934, Loss:  0.106
Epoch   6 Batch   11/137 - Train Accuracy:  0.930, Validation Accuracy:  0.931, Loss:  0.110
Epoch   6 Batch   12/137 - Train Accuracy:  0.910, Validation Accuracy:  0.932, Loss:  0.109
Epoch   6 Batch   13/137 - Train Accuracy:  0.927, Validation Accuracy:  0.929, Loss:  0.099
Epoch   6 Batch   14/137 - Train Accuracy:  0.919, Validation Accuracy:  0.927, Loss:  0.107
Epoch   6 Batch   15/137 - Train Accuracy:  0.920, Validation Accuracy:  0.930, Loss:  0.106
Epoch   6 Batch   16/137 - Train Accuracy:  0.931, Validation Accuracy:  0.933, Loss:  0.092
Epoch   6 Batch   17/137 - Train Accuracy:  0.919, Validation Accuracy:  0.932, Loss:  0.113
Epoch   6 Batch   18/137 - Train Accuracy:  0.929, Validation Accuracy:  0.930, Loss:  0.105
Epoch   6 Batch   19/137 - Train Accuracy:  0.919, Validation Accuracy:  0.933, Loss:  0.102
Epoch   6 Batch   20/137 - Train Accuracy:  0.913, Validation Accuracy:  0.932, Loss:  0.104
Epoch   6 Batch   21/137 - Train Accuracy:  0.930, Validation Accuracy:  0.928, Loss:  0.097
Epoch   6 Batch   22/137 - Train Accuracy:  0.926, Validation Accuracy:  0.931, Loss:  0.106
Epoch   6 Batch   23/137 - Train Accuracy:  0.934, Validation Accuracy:  0.928, Loss:  0.091
Epoch   6 Batch   24/137 - Train Accuracy:  0.930, Validation Accuracy:  0.933, Loss:  0.098
Epoch   6 Batch   25/137 - Train Accuracy:  0.920, Validation Accuracy:  0.933, Loss:  0.107
Epoch   6 Batch   26/137 - Train Accuracy:  0.924, Validation Accuracy:  0.934, Loss:  0.095
Epoch   6 Batch   27/137 - Train Accuracy:  0.931, Validation Accuracy:  0.936, Loss:  0.096
Epoch   6 Batch   28/137 - Train Accuracy:  0.918, Validation Accuracy:  0.934, Loss:  0.104
Epoch   6 Batch   29/137 - Train Accuracy:  0.937, Validation Accuracy:  0.936, Loss:  0.102
Epoch   6 Batch   30/137 - Train Accuracy:  0.937, Validation Accuracy:  0.935, Loss:  0.091
Epoch   6 Batch   31/137 - Train Accuracy:  0.923, Validation Accuracy:  0.936, Loss:  0.102
Epoch   6 Batch   32/137 - Train Accuracy:  0.933, Validation Accuracy:  0.936, Loss:  0.102
Epoch   6 Batch   33/137 - Train Accuracy:  0.926, Validation Accuracy:  0.935, Loss:  0.098
Epoch   6 Batch   34/137 - Train Accuracy:  0.916, Validation Accuracy:  0.935, Loss:  0.113
Epoch   6 Batch   35/137 - Train Accuracy:  0.926, Validation Accuracy:  0.937, Loss:  0.105
Epoch   6 Batch   36/137 - Train Accuracy:  0.929, Validation Accuracy:  0.933, Loss:  0.104
Epoch   6 Batch   37/137 - Train Accuracy:  0.930, Validation Accuracy:  0.935, Loss:  0.100
Epoch   6 Batch   38/137 - Train Accuracy:  0.928, Validation Accuracy:  0.935, Loss:  0.099
Epoch   6 Batch   39/137 - Train Accuracy:  0.931, Validation Accuracy:  0.935, Loss:  0.098
Epoch   6 Batch   40/137 - Train Accuracy:  0.923, Validation Accuracy:  0.934, Loss:  0.102
Epoch   6 Batch   41/137 - Train Accuracy:  0.931, Validation Accuracy:  0.934, Loss:  0.102
Epoch   6 Batch   42/137 - Train Accuracy:  0.927, Validation Accuracy:  0.932, Loss:  0.105
Epoch   6 Batch   43/137 - Train Accuracy:  0.922, Validation Accuracy:  0.931, Loss:  0.093
Epoch   6 Batch   44/137 - Train Accuracy:  0.924, Validation Accuracy:  0.931, Loss:  0.100
Epoch   6 Batch   45/137 - Train Accuracy:  0.928, Validation Accuracy:  0.932, Loss:  0.098
Epoch   6 Batch   46/137 - Train Accuracy:  0.941, Validation Accuracy:  0.933, Loss:  0.096
Epoch   6 Batch   47/137 - Train Accuracy:  0.930, Validation Accuracy:  0.935, Loss:  0.095
Epoch   6 Batch   48/137 - Train Accuracy:  0.927, Validation Accuracy:  0.935, Loss:  0.108
Epoch   6 Batch   49/137 - Train Accuracy:  0.920, Validation Accuracy:  0.937, Loss:  0.103
Epoch   6 Batch   50/137 - Train Accuracy:  0.930, Validation Accuracy:  0.937, Loss:  0.102
Epoch   6 Batch   51/137 - Train Accuracy:  0.937, Validation Accuracy:  0.936, Loss:  0.100
Epoch   6 Batch   52/137 - Train Accuracy:  0.925, Validation Accuracy:  0.938, Loss:  0.096
Epoch   6 Batch   53/137 - Train Accuracy:  0.922, Validation Accuracy:  0.934, Loss:  0.096
Epoch   6 Batch   54/137 - Train Accuracy:  0.940, Validation Accuracy:  0.934, Loss:  0.094
Epoch   6 Batch   55/137 - Train Accuracy:  0.927, Validation Accuracy:  0.938, Loss:  0.095
Epoch   6 Batch   56/137 - Train Accuracy:  0.928, Validation Accuracy:  0.935, Loss:  0.095
Epoch   6 Batch   57/137 - Train Accuracy:  0.927, Validation Accuracy:  0.932, Loss:  0.100
Epoch   6 Batch   58/137 - Train Accuracy:  0.932, Validation Accuracy:  0.937, Loss:  0.096
Epoch   6 Batch   59/137 - Train Accuracy:  0.934, Validation Accuracy:  0.940, Loss:  0.093
Epoch   6 Batch   60/137 - Train Accuracy:  0.924, Validation Accuracy:  0.940, Loss:  0.098
Epoch   6 Batch   61/137 - Train Accuracy:  0.933, Validation Accuracy:  0.938, Loss:  0.094
Epoch   6 Batch   62/137 - Train Accuracy:  0.928, Validation Accuracy:  0.937, Loss:  0.097
Epoch   6 Batch   63/137 - Train Accuracy:  0.927, Validation Accuracy:  0.936, Loss:  0.090
Epoch   6 Batch   64/137 - Train Accuracy:  0.923, Validation Accuracy:  0.935, Loss:  0.094
Epoch   6 Batch   65/137 - Train Accuracy:  0.923, Validation Accuracy:  0.936, Loss:  0.096
Epoch   6 Batch   66/137 - Train Accuracy:  0.923, Validation Accuracy:  0.942, Loss:  0.105
Epoch   6 Batch   67/137 - Train Accuracy:  0.923, Validation Accuracy:  0.939, Loss:  0.093
Epoch   6 Batch   68/137 - Train Accuracy:  0.933, Validation Accuracy:  0.938, Loss:  0.089
Epoch   6 Batch   69/137 - Train Accuracy:  0.924, Validation Accuracy:  0.938, Loss:  0.105
Epoch   6 Batch   70/137 - Train Accuracy:  0.921, Validation Accuracy:  0.939, Loss:  0.090
Epoch   6 Batch   71/137 - Train Accuracy:  0.926, Validation Accuracy:  0.944, Loss:  0.098
Epoch   6 Batch   72/137 - Train Accuracy:  0.931, Validation Accuracy:  0.941, Loss:  0.092
Epoch   6 Batch   73/137 - Train Accuracy:  0.937, Validation Accuracy:  0.939, Loss:  0.088
Epoch   6 Batch   74/137 - Train Accuracy:  0.934, Validation Accuracy:  0.937, Loss:  0.086
Epoch   6 Batch   75/137 - Train Accuracy:  0.925, Validation Accuracy:  0.938, Loss:  0.106
Epoch   6 Batch   76/137 - Train Accuracy:  0.926, Validation Accuracy:  0.938, Loss:  0.094
Epoch   6 Batch   77/137 - Train Accuracy:  0.931, Validation Accuracy:  0.938, Loss:  0.096
Epoch   6 Batch   78/137 - Train Accuracy:  0.946, Validation Accuracy:  0.939, Loss:  0.087
Epoch   6 Batch   79/137 - Train Accuracy:  0.927, Validation Accuracy:  0.939, Loss:  0.087
Epoch   6 Batch   80/137 - Train Accuracy:  0.924, Validation Accuracy:  0.935, Loss:  0.091
Epoch   6 Batch   81/137 - Train Accuracy:  0.935, Validation Accuracy:  0.942, Loss:  0.096
Epoch   6 Batch   82/137 - Train Accuracy:  0.943, Validation Accuracy:  0.940, Loss:  0.092
Epoch   6 Batch   83/137 - Train Accuracy:  0.937, Validation Accuracy:  0.938, Loss:  0.090
Epoch   6 Batch   84/137 - Train Accuracy:  0.935, Validation Accuracy:  0.941, Loss:  0.088
Epoch   6 Batch   85/137 - Train Accuracy:  0.924, Validation Accuracy:  0.939, Loss:  0.092
Epoch   6 Batch   86/137 - Train Accuracy:  0.925, Validation Accuracy:  0.940, Loss:  0.089
Epoch   6 Batch   87/137 - Train Accuracy:  0.937, Validation Accuracy:  0.942, Loss:  0.093
Epoch   6 Batch   88/137 - Train Accuracy:  0.941, Validation Accuracy:  0.942, Loss:  0.085
Epoch   6 Batch   89/137 - Train Accuracy:  0.926, Validation Accuracy:  0.945, Loss:  0.098
Epoch   6 Batch   90/137 - Train Accuracy:  0.947, Validation Accuracy:  0.941, Loss:  0.082
Epoch   6 Batch   91/137 - Train Accuracy:  0.930, Validation Accuracy:  0.941, Loss:  0.087
Epoch   6 Batch   92/137 - Train Accuracy:  0.932, Validation Accuracy:  0.942, Loss:  0.092
Epoch   6 Batch   93/137 - Train Accuracy:  0.948, Validation Accuracy:  0.944, Loss:  0.074
Epoch   6 Batch   94/137 - Train Accuracy:  0.945, Validation Accuracy:  0.942, Loss:  0.085
Epoch   6 Batch   95/137 - Train Accuracy:  0.936, Validation Accuracy:  0.941, Loss:  0.087
Epoch   6 Batch   96/137 - Train Accuracy:  0.938, Validation Accuracy:  0.944, Loss:  0.081
Epoch   6 Batch   97/137 - Train Accuracy:  0.926, Validation Accuracy:  0.945, Loss:  0.087
Epoch   6 Batch   98/137 - Train Accuracy:  0.938, Validation Accuracy:  0.941, Loss:  0.088
Epoch   6 Batch   99/137 - Train Accuracy:  0.930, Validation Accuracy:  0.942, Loss:  0.089
Epoch   6 Batch  100/137 - Train Accuracy:  0.931, Validation Accuracy:  0.944, Loss:  0.089
Epoch   6 Batch  101/137 - Train Accuracy:  0.933, Validation Accuracy:  0.942, Loss:  0.090
Epoch   6 Batch  102/137 - Train Accuracy:  0.934, Validation Accuracy:  0.944, Loss:  0.085
Epoch   6 Batch  103/137 - Train Accuracy:  0.933, Validation Accuracy:  0.947, Loss:  0.092
Epoch   6 Batch  104/137 - Train Accuracy:  0.936, Validation Accuracy:  0.949, Loss:  0.082
Epoch   6 Batch  105/137 - Train Accuracy:  0.928, Validation Accuracy:  0.946, Loss:  0.086
Epoch   6 Batch  106/137 - Train Accuracy:  0.937, Validation Accuracy:  0.945, Loss:  0.087
Epoch   6 Batch  107/137 - Train Accuracy:  0.936, Validation Accuracy:  0.942, Loss:  0.085
Epoch   6 Batch  108/137 - Train Accuracy:  0.931, Validation Accuracy:  0.941, Loss:  0.096
Epoch   6 Batch  109/137 - Train Accuracy:  0.937, Validation Accuracy:  0.939, Loss:  0.084
Epoch   6 Batch  110/137 - Train Accuracy:  0.931, Validation Accuracy:  0.944, Loss:  0.088
Epoch   6 Batch  111/137 - Train Accuracy:  0.940, Validation Accuracy:  0.943, Loss:  0.090
Epoch   6 Batch  112/137 - Train Accuracy:  0.933, Validation Accuracy:  0.941, Loss:  0.083
Epoch   6 Batch  113/137 - Train Accuracy:  0.939, Validation Accuracy:  0.944, Loss:  0.084
Epoch   6 Batch  114/137 - Train Accuracy:  0.928, Validation Accuracy:  0.944, Loss:  0.092
Epoch   6 Batch  115/137 - Train Accuracy:  0.942, Validation Accuracy:  0.944, Loss:  0.084
Epoch   6 Batch  116/137 - Train Accuracy:  0.944, Validation Accuracy:  0.944, Loss:  0.079
Epoch   6 Batch  117/137 - Train Accuracy:  0.932, Validation Accuracy:  0.943, Loss:  0.087
Epoch   6 Batch  118/137 - Train Accuracy:  0.927, Validation Accuracy:  0.942, Loss:  0.090
Epoch   6 Batch  119/137 - Train Accuracy:  0.936, Validation Accuracy:  0.945, Loss:  0.089
Epoch   6 Batch  120/137 - Train Accuracy:  0.948, Validation Accuracy:  0.946, Loss:  0.076
Epoch   6 Batch  121/137 - Train Accuracy:  0.939, Validation Accuracy:  0.942, Loss:  0.082
Epoch   6 Batch  122/137 - Train Accuracy:  0.941, Validation Accuracy:  0.938, Loss:  0.075
Epoch   6 Batch  123/137 - Train Accuracy:  0.934, Validation Accuracy:  0.942, Loss:  0.092
Epoch   6 Batch  124/137 - Train Accuracy:  0.933, Validation Accuracy:  0.946, Loss:  0.080
Epoch   6 Batch  125/137 - Train Accuracy:  0.937, Validation Accuracy:  0.944, Loss:  0.088
Epoch   6 Batch  126/137 - Train Accuracy:  0.944, Validation Accuracy:  0.946, Loss:  0.077
Epoch   6 Batch  127/137 - Train Accuracy:  0.942, Validation Accuracy:  0.946, Loss:  0.077
Epoch   6 Batch  128/137 - Train Accuracy:  0.947, Validation Accuracy:  0.945, Loss:  0.076
Epoch   6 Batch  129/137 - Train Accuracy:  0.931, Validation Accuracy:  0.945, Loss:  0.084
Epoch   6 Batch  130/137 - Train Accuracy:  0.931, Validation Accuracy:  0.944, Loss:  0.080
Epoch   6 Batch  131/137 - Train Accuracy:  0.924, Validation Accuracy:  0.944, Loss:  0.086
Epoch   6 Batch  132/137 - Train Accuracy:  0.933, Validation Accuracy:  0.944, Loss:  0.088
Epoch   6 Batch  133/137 - Train Accuracy:  0.943, Validation Accuracy:  0.941, Loss:  0.088
Epoch   6 Batch  134/137 - Train Accuracy:  0.928, Validation Accuracy:  0.942, Loss:  0.090
Epoch   6 Batch  135/137 - Train Accuracy:  0.933, Validation Accuracy:  0.943, Loss:  0.087
Epoch   7 Batch    0/137 - Train Accuracy:  0.938, Validation Accuracy:  0.948, Loss:  0.084
Epoch   7 Batch    1/137 - Train Accuracy:  0.931, Validation Accuracy:  0.946, Loss:  0.088
Epoch   7 Batch    2/137 - Train Accuracy:  0.948, Validation Accuracy:  0.950, Loss:  0.084
Epoch   7 Batch    3/137 - Train Accuracy:  0.945, Validation Accuracy:  0.945, Loss:  0.078
Epoch   7 Batch    4/137 - Train Accuracy:  0.941, Validation Accuracy:  0.950, Loss:  0.091
Epoch   7 Batch    5/137 - Train Accuracy:  0.936, Validation Accuracy:  0.941, Loss:  0.084
Epoch   7 Batch    6/137 - Train Accuracy:  0.934, Validation Accuracy:  0.942, Loss:  0.084
Epoch   7 Batch    7/137 - Train Accuracy:  0.936, Validation Accuracy:  0.938, Loss:  0.087
Epoch   7 Batch    8/137 - Train Accuracy:  0.934, Validation Accuracy:  0.946, Loss:  0.084
Epoch   7 Batch    9/137 - Train Accuracy:  0.946, Validation Accuracy:  0.943, Loss:  0.072
Epoch   7 Batch   10/137 - Train Accuracy:  0.930, Validation Accuracy:  0.944, Loss:  0.090
Epoch   7 Batch   11/137 - Train Accuracy:  0.946, Validation Accuracy:  0.949, Loss:  0.081
Epoch   7 Batch   12/137 - Train Accuracy:  0.933, Validation Accuracy:  0.947, Loss:  0.085
Epoch   7 Batch   13/137 - Train Accuracy:  0.939, Validation Accuracy:  0.948, Loss:  0.077
Epoch   7 Batch   14/137 - Train Accuracy:  0.936, Validation Accuracy:  0.946, Loss:  0.087
Epoch   7 Batch   15/137 - Train Accuracy:  0.936, Validation Accuracy:  0.947, Loss:  0.082
Epoch   7 Batch   16/137 - Train Accuracy:  0.940, Validation Accuracy:  0.945, Loss:  0.078
Epoch   7 Batch   17/137 - Train Accuracy:  0.933, Validation Accuracy:  0.945, Loss:  0.086
Epoch   7 Batch   18/137 - Train Accuracy:  0.938, Validation Accuracy:  0.946, Loss:  0.080
Epoch   7 Batch   19/137 - Train Accuracy:  0.935, Validation Accuracy:  0.951, Loss:  0.080
Epoch   7 Batch   20/137 - Train Accuracy:  0.930, Validation Accuracy:  0.943, Loss:  0.081
Epoch   7 Batch   21/137 - Train Accuracy:  0.944, Validation Accuracy:  0.943, Loss:  0.075
Epoch   7 Batch   22/137 - Train Accuracy:  0.936, Validation Accuracy:  0.943, Loss:  0.081
Epoch   7 Batch   23/137 - Train Accuracy:  0.941, Validation Accuracy:  0.942, Loss:  0.070
Epoch   7 Batch   24/137 - Train Accuracy:  0.945, Validation Accuracy:  0.947, Loss:  0.076
Epoch   7 Batch   25/137 - Train Accuracy:  0.933, Validation Accuracy:  0.949, Loss:  0.081
Epoch   7 Batch   26/137 - Train Accuracy:  0.933, Validation Accuracy:  0.945, Loss:  0.076
Epoch   7 Batch   27/137 - Train Accuracy:  0.948, Validation Accuracy:  0.946, Loss:  0.074
Epoch   7 Batch   28/137 - Train Accuracy:  0.933, Validation Accuracy:  0.942, Loss:  0.077
Epoch   7 Batch   29/137 - Train Accuracy:  0.942, Validation Accuracy:  0.942, Loss:  0.079
Epoch   7 Batch   30/137 - Train Accuracy:  0.946, Validation Accuracy:  0.948, Loss:  0.072
Epoch   7 Batch   31/137 - Train Accuracy:  0.935, Validation Accuracy:  0.950, Loss:  0.079
Epoch   7 Batch   32/137 - Train Accuracy:  0.945, Validation Accuracy:  0.949, Loss:  0.073
Epoch   7 Batch   33/137 - Train Accuracy:  0.939, Validation Accuracy:  0.951, Loss:  0.078
Epoch   7 Batch   34/137 - Train Accuracy:  0.929, Validation Accuracy:  0.942, Loss:  0.089
Epoch   7 Batch   35/137 - Train Accuracy:  0.930, Validation Accuracy:  0.939, Loss:  0.084
Epoch   7 Batch   36/137 - Train Accuracy:  0.938, Validation Accuracy:  0.942, Loss:  0.079
Epoch   7 Batch   37/137 - Train Accuracy:  0.944, Validation Accuracy:  0.947, Loss:  0.077
Epoch   7 Batch   38/137 - Train Accuracy:  0.941, Validation Accuracy:  0.944, Loss:  0.080
Epoch   7 Batch   39/137 - Train Accuracy:  0.939, Validation Accuracy:  0.943, Loss:  0.078
Epoch   7 Batch   40/137 - Train Accuracy:  0.938, Validation Accuracy:  0.946, Loss:  0.075
Epoch   7 Batch   41/137 - Train Accuracy:  0.941, Validation Accuracy:  0.946, Loss:  0.077
Epoch   7 Batch   42/137 - Train Accuracy:  0.931, Validation Accuracy:  0.945, Loss:  0.083
Epoch   7 Batch   43/137 - Train Accuracy:  0.942, Validation Accuracy:  0.943, Loss:  0.069
Epoch   7 Batch   44/137 - Train Accuracy:  0.938, Validation Accuracy:  0.948, Loss:  0.079
Epoch   7 Batch   45/137 - Train Accuracy:  0.943, Validation Accuracy:  0.948, Loss:  0.076
Epoch   7 Batch   46/137 - Train Accuracy:  0.953, Validation Accuracy:  0.944, Loss:  0.069
Epoch   7 Batch   47/137 - Train Accuracy:  0.947, Validation Accuracy:  0.952, Loss:  0.074
Epoch   7 Batch   48/137 - Train Accuracy:  0.940, Validation Accuracy:  0.949, Loss:  0.080
Epoch   7 Batch   49/137 - Train Accuracy:  0.939, Validation Accuracy:  0.951, Loss:  0.081
Epoch   7 Batch   50/137 - Train Accuracy:  0.939, Validation Accuracy:  0.953, Loss:  0.076
Epoch   7 Batch   51/137 - Train Accuracy:  0.947, Validation Accuracy:  0.949, Loss:  0.079
Epoch   7 Batch   52/137 - Train Accuracy:  0.942, Validation Accuracy:  0.950, Loss:  0.076
Epoch   7 Batch   53/137 - Train Accuracy:  0.933, Validation Accuracy:  0.950, Loss:  0.078
Epoch   7 Batch   54/137 - Train Accuracy:  0.953, Validation Accuracy:  0.942, Loss:  0.072
Epoch   7 Batch   55/137 - Train Accuracy:  0.937, Validation Accuracy:  0.942, Loss:  0.079
Epoch   7 Batch   56/137 - Train Accuracy:  0.935, Validation Accuracy:  0.943, Loss:  0.076
Epoch   7 Batch   57/137 - Train Accuracy:  0.938, Validation Accuracy:  0.944, Loss:  0.072
Epoch   7 Batch   58/137 - Train Accuracy:  0.937, Validation Accuracy:  0.944, Loss:  0.071
Epoch   7 Batch   59/137 - Train Accuracy:  0.942, Validation Accuracy:  0.946, Loss:  0.068
Epoch   7 Batch   60/137 - Train Accuracy:  0.944, Validation Accuracy:  0.944, Loss:  0.074
Epoch   7 Batch   61/137 - Train Accuracy:  0.935, Validation Accuracy:  0.946, Loss:  0.072
Epoch   7 Batch   62/137 - Train Accuracy:  0.939, Validation Accuracy:  0.945, Loss:  0.071
Epoch   7 Batch   63/137 - Train Accuracy:  0.936, Validation Accuracy:  0.946, Loss:  0.068
Epoch   7 Batch   64/137 - Train Accuracy:  0.932, Validation Accuracy:  0.946, Loss:  0.077
Epoch   7 Batch   65/137 - Train Accuracy:  0.933, Validation Accuracy:  0.946, Loss:  0.075
Epoch   7 Batch   66/137 - Train Accuracy:  0.935, Validation Accuracy:  0.948, Loss:  0.084
Epoch   7 Batch   67/137 - Train Accuracy:  0.934, Validation Accuracy:  0.950, Loss:  0.073
Epoch   7 Batch   68/137 - Train Accuracy:  0.943, Validation Accuracy:  0.949, Loss:  0.068
Epoch   7 Batch   69/137 - Train Accuracy:  0.935, Validation Accuracy:  0.949, Loss:  0.085
Epoch   7 Batch   70/137 - Train Accuracy:  0.940, Validation Accuracy:  0.944, Loss:  0.070
Epoch   7 Batch   71/137 - Train Accuracy:  0.932, Validation Accuracy:  0.944, Loss:  0.075
Epoch   7 Batch   72/137 - Train Accuracy:  0.944, Validation Accuracy:  0.949, Loss:  0.070
Epoch   7 Batch   73/137 - Train Accuracy:  0.951, Validation Accuracy:  0.951, Loss:  0.069
Epoch   7 Batch   74/137 - Train Accuracy:  0.947, Validation Accuracy:  0.951, Loss:  0.069
Epoch   7 Batch   75/137 - Train Accuracy:  0.941, Validation Accuracy:  0.944, Loss:  0.082
Epoch   7 Batch   76/137 - Train Accuracy:  0.939, Validation Accuracy:  0.943, Loss:  0.073
Epoch   7 Batch   77/137 - Train Accuracy:  0.937, Validation Accuracy:  0.947, Loss:  0.077
Epoch   7 Batch   78/137 - Train Accuracy:  0.949, Validation Accuracy:  0.950, Loss:  0.069
Epoch   7 Batch   79/137 - Train Accuracy:  0.946, Validation Accuracy:  0.951, Loss:  0.070
Epoch   7 Batch   80/137 - Train Accuracy:  0.937, Validation Accuracy:  0.949, Loss:  0.069
Epoch   7 Batch   81/137 - Train Accuracy:  0.940, Validation Accuracy:  0.949, Loss:  0.074
Epoch   7 Batch   82/137 - Train Accuracy:  0.946, Validation Accuracy:  0.946, Loss:  0.071
Epoch   7 Batch   83/137 - Train Accuracy:  0.948, Validation Accuracy:  0.944, Loss:  0.072
Epoch   7 Batch   84/137 - Train Accuracy:  0.952, Validation Accuracy:  0.947, Loss:  0.068
Epoch   7 Batch   85/137 - Train Accuracy:  0.941, Validation Accuracy:  0.951, Loss:  0.076
Epoch   7 Batch   86/137 - Train Accuracy:  0.939, Validation Accuracy:  0.952, Loss:  0.066
Epoch   7 Batch   87/137 - Train Accuracy:  0.944, Validation Accuracy:  0.951, Loss:  0.075
Epoch   7 Batch   88/137 - Train Accuracy:  0.949, Validation Accuracy:  0.951, Loss:  0.066
Epoch   7 Batch   89/137 - Train Accuracy:  0.929, Validation Accuracy:  0.952, Loss:  0.087
Epoch   7 Batch   90/137 - Train Accuracy:  0.951, Validation Accuracy:  0.951, Loss:  0.066
Epoch   7 Batch   91/137 - Train Accuracy:  0.942, Validation Accuracy:  0.955, Loss:  0.071
Epoch   7 Batch   92/137 - Train Accuracy:  0.941, Validation Accuracy:  0.955, Loss:  0.077
Epoch   7 Batch   93/137 - Train Accuracy:  0.954, Validation Accuracy:  0.950, Loss:  0.059
Epoch   7 Batch   94/137 - Train Accuracy:  0.951, Validation Accuracy:  0.950, Loss:  0.066
Epoch   7 Batch   95/137 - Train Accuracy:  0.943, Validation Accuracy:  0.953, Loss:  0.069
Epoch   7 Batch   96/137 - Train Accuracy:  0.950, Validation Accuracy:  0.952, Loss:  0.066
Epoch   7 Batch   97/137 - Train Accuracy:  0.938, Validation Accuracy:  0.954, Loss:  0.069
Epoch   7 Batch   98/137 - Train Accuracy:  0.948, Validation Accuracy:  0.952, Loss:  0.068
Epoch   7 Batch   99/137 - Train Accuracy:  0.936, Validation Accuracy:  0.953, Loss:  0.073
Epoch   7 Batch  100/137 - Train Accuracy:  0.939, Validation Accuracy:  0.953, Loss:  0.072
Epoch   7 Batch  101/137 - Train Accuracy:  0.948, Validation Accuracy:  0.953, Loss:  0.071
Epoch   7 Batch  102/137 - Train Accuracy:  0.948, Validation Accuracy:  0.955, Loss:  0.066
Epoch   7 Batch  103/137 - Train Accuracy:  0.940, Validation Accuracy:  0.957, Loss:  0.070
Epoch   7 Batch  104/137 - Train Accuracy:  0.945, Validation Accuracy:  0.955, Loss:  0.071
Epoch   7 Batch  105/137 - Train Accuracy:  0.936, Validation Accuracy:  0.955, Loss:  0.068
Epoch   7 Batch  106/137 - Train Accuracy:  0.950, Validation Accuracy:  0.950, Loss:  0.073
Epoch   7 Batch  107/137 - Train Accuracy:  0.947, Validation Accuracy:  0.951, Loss:  0.068
Epoch   7 Batch  108/137 - Train Accuracy:  0.946, Validation Accuracy:  0.951, Loss:  0.077
Epoch   7 Batch  109/137 - Train Accuracy:  0.945, Validation Accuracy:  0.954, Loss:  0.067
Epoch   7 Batch  110/137 - Train Accuracy:  0.937, Validation Accuracy:  0.954, Loss:  0.076
Epoch   7 Batch  111/137 - Train Accuracy:  0.949, Validation Accuracy:  0.953, Loss:  0.071
Epoch   7 Batch  112/137 - Train Accuracy:  0.943, Validation Accuracy:  0.947, Loss:  0.066
Epoch   7 Batch  113/137 - Train Accuracy:  0.948, Validation Accuracy:  0.945, Loss:  0.065
Epoch   7 Batch  114/137 - Train Accuracy:  0.939, Validation Accuracy:  0.949, Loss:  0.071
Epoch   7 Batch  115/137 - Train Accuracy:  0.952, Validation Accuracy:  0.953, Loss:  0.069
Epoch   7 Batch  116/137 - Train Accuracy:  0.956, Validation Accuracy:  0.954, Loss:  0.064
Epoch   7 Batch  117/137 - Train Accuracy:  0.942, Validation Accuracy:  0.951, Loss:  0.072
Epoch   7 Batch  118/137 - Train Accuracy:  0.940, Validation Accuracy:  0.952, Loss:  0.072
Epoch   7 Batch  119/137 - Train Accuracy:  0.951, Validation Accuracy:  0.952, Loss:  0.073
Epoch   7 Batch  120/137 - Train Accuracy:  0.961, Validation Accuracy:  0.950, Loss:  0.064
Epoch   7 Batch  121/137 - Train Accuracy:  0.949, Validation Accuracy:  0.949, Loss:  0.067
Epoch   7 Batch  122/137 - Train Accuracy:  0.949, Validation Accuracy:  0.951, Loss:  0.063
Epoch   7 Batch  123/137 - Train Accuracy:  0.942, Validation Accuracy:  0.950, Loss:  0.076
Epoch   7 Batch  124/137 - Train Accuracy:  0.947, Validation Accuracy:  0.950, Loss:  0.063
Epoch   7 Batch  125/137 - Train Accuracy:  0.941, Validation Accuracy:  0.949, Loss:  0.072
Epoch   7 Batch  126/137 - Train Accuracy:  0.952, Validation Accuracy:  0.954, Loss:  0.064
Epoch   7 Batch  127/137 - Train Accuracy:  0.950, Validation Accuracy:  0.951, Loss:  0.064
Epoch   7 Batch  128/137 - Train Accuracy:  0.953, Validation Accuracy:  0.952, Loss:  0.065
Epoch   7 Batch  129/137 - Train Accuracy:  0.944, Validation Accuracy:  0.954, Loss:  0.069
Epoch   7 Batch  130/137 - Train Accuracy:  0.942, Validation Accuracy:  0.951, Loss:  0.068
Epoch   7 Batch  131/137 - Train Accuracy:  0.941, Validation Accuracy:  0.945, Loss:  0.069
Epoch   7 Batch  132/137 - Train Accuracy:  0.943, Validation Accuracy:  0.950, Loss:  0.071
Epoch   7 Batch  133/137 - Train Accuracy:  0.944, Validation Accuracy:  0.956, Loss:  0.067
Epoch   7 Batch  134/137 - Train Accuracy:  0.940, Validation Accuracy:  0.957, Loss:  0.075
Epoch   7 Batch  135/137 - Train Accuracy:  0.936, Validation Accuracy:  0.955, Loss:  0.069
Epoch   8 Batch    0/137 - Train Accuracy:  0.950, Validation Accuracy:  0.951, Loss:  0.067
Epoch   8 Batch    1/137 - Train Accuracy:  0.938, Validation Accuracy:  0.952, Loss:  0.067
Epoch   8 Batch    2/137 - Train Accuracy:  0.952, Validation Accuracy:  0.951, Loss:  0.064
Epoch   8 Batch    3/137 - Train Accuracy:  0.948, Validation Accuracy:  0.952, Loss:  0.067
Epoch   8 Batch    4/137 - Train Accuracy:  0.955, Validation Accuracy:  0.950, Loss:  0.067
Epoch   8 Batch    5/137 - Train Accuracy:  0.945, Validation Accuracy:  0.951, Loss:  0.070
Epoch   8 Batch    6/137 - Train Accuracy:  0.947, Validation Accuracy:  0.951, Loss:  0.059
Epoch   8 Batch    7/137 - Train Accuracy:  0.948, Validation Accuracy:  0.956, Loss:  0.062
Epoch   8 Batch    8/137 - Train Accuracy:  0.939, Validation Accuracy:  0.955, Loss:  0.067
Epoch   8 Batch    9/137 - Train Accuracy:  0.955, Validation Accuracy:  0.956, Loss:  0.061
Epoch   8 Batch   10/137 - Train Accuracy:  0.942, Validation Accuracy:  0.956, Loss:  0.067
Epoch   8 Batch   11/137 - Train Accuracy:  0.952, Validation Accuracy:  0.954, Loss:  0.065
Epoch   8 Batch   12/137 - Train Accuracy:  0.944, Validation Accuracy:  0.954, Loss:  0.069
Epoch   8 Batch   13/137 - Train Accuracy:  0.941, Validation Accuracy:  0.952, Loss:  0.063
Epoch   8 Batch   14/137 - Train Accuracy:  0.944, Validation Accuracy:  0.958, Loss:  0.072
Epoch   8 Batch   15/137 - Train Accuracy:  0.940, Validation Accuracy:  0.959, Loss:  0.063
Epoch   8 Batch   16/137 - Train Accuracy:  0.951, Validation Accuracy:  0.960, Loss:  0.059
Epoch   8 Batch   17/137 - Train Accuracy:  0.946, Validation Accuracy:  0.958, Loss:  0.073
Epoch   8 Batch   18/137 - Train Accuracy:  0.948, Validation Accuracy:  0.949, Loss:  0.064
Epoch   8 Batch   19/137 - Train Accuracy:  0.940, Validation Accuracy:  0.952, Loss:  0.062
Epoch   8 Batch   20/137 - Train Accuracy:  0.938, Validation Accuracy:  0.952, Loss:  0.069
Epoch   8 Batch   21/137 - Train Accuracy:  0.950, Validation Accuracy:  0.952, Loss:  0.063
Epoch   8 Batch   22/137 - Train Accuracy:  0.948, Validation Accuracy:  0.949, Loss:  0.067
Epoch   8 Batch   23/137 - Train Accuracy:  0.946, Validation Accuracy:  0.948, Loss:  0.057
Epoch   8 Batch   24/137 - Train Accuracy:  0.951, Validation Accuracy:  0.950, Loss:  0.057
Epoch   8 Batch   25/137 - Train Accuracy:  0.946, Validation Accuracy:  0.950, Loss:  0.069
Epoch   8 Batch   26/137 - Train Accuracy:  0.948, Validation Accuracy:  0.954, Loss:  0.058
Epoch   8 Batch   27/137 - Train Accuracy:  0.957, Validation Accuracy:  0.955, Loss:  0.061
Epoch   8 Batch   28/137 - Train Accuracy:  0.946, Validation Accuracy:  0.954, Loss:  0.065
Epoch   8 Batch   29/137 - Train Accuracy:  0.948, Validation Accuracy:  0.952, Loss:  0.066
Epoch   8 Batch   30/137 - Train Accuracy:  0.955, Validation Accuracy:  0.952, Loss:  0.057
Epoch   8 Batch   31/137 - Train Accuracy:  0.946, Validation Accuracy:  0.954, Loss:  0.066
Epoch   8 Batch   32/137 - Train Accuracy:  0.949, Validation Accuracy:  0.954, Loss:  0.058
Epoch   8 Batch   33/137 - Train Accuracy:  0.941, Validation Accuracy:  0.958, Loss:  0.067
Epoch   8 Batch   34/137 - Train Accuracy:  0.941, Validation Accuracy:  0.957, Loss:  0.071
Epoch   8 Batch   35/137 - Train Accuracy:  0.946, Validation Accuracy:  0.953, Loss:  0.072
Epoch   8 Batch   36/137 - Train Accuracy:  0.951, Validation Accuracy:  0.952, Loss:  0.071
Epoch   8 Batch   37/137 - Train Accuracy:  0.957, Validation Accuracy:  0.952, Loss:  0.064
Epoch   8 Batch   38/137 - Train Accuracy:  0.948, Validation Accuracy:  0.955, Loss:  0.065
Epoch   8 Batch   39/137 - Train Accuracy:  0.955, Validation Accuracy:  0.953, Loss:  0.067
Epoch   8 Batch   40/137 - Train Accuracy:  0.942, Validation Accuracy:  0.953, Loss:  0.062
Epoch   8 Batch   41/137 - Train Accuracy:  0.950, Validation Accuracy:  0.951, Loss:  0.067
Epoch   8 Batch   42/137 - Train Accuracy:  0.938, Validation Accuracy:  0.950, Loss:  0.069
Epoch   8 Batch   43/137 - Train Accuracy:  0.952, Validation Accuracy:  0.954, Loss:  0.060
Epoch   8 Batch   44/137 - Train Accuracy:  0.949, Validation Accuracy:  0.955, Loss:  0.065
Epoch   8 Batch   45/137 - Train Accuracy:  0.953, Validation Accuracy:  0.953, Loss:  0.059
Epoch   8 Batch   46/137 - Train Accuracy:  0.964, Validation Accuracy:  0.958, Loss:  0.058
Epoch   8 Batch   47/137 - Train Accuracy:  0.954, Validation Accuracy:  0.954, Loss:  0.063
Epoch   8 Batch   48/137 - Train Accuracy:  0.950, Validation Accuracy:  0.948, Loss:  0.067
Epoch   8 Batch   49/137 - Train Accuracy:  0.942, Validation Accuracy:  0.955, Loss:  0.064
Epoch   8 Batch   50/137 - Train Accuracy:  0.947, Validation Accuracy:  0.954, Loss:  0.067
Epoch   8 Batch   51/137 - Train Accuracy:  0.951, Validation Accuracy:  0.956, Loss:  0.065
Epoch   8 Batch   52/137 - Train Accuracy:  0.957, Validation Accuracy:  0.952, Loss:  0.064
Epoch   8 Batch   53/137 - Train Accuracy:  0.948, Validation Accuracy:  0.948, Loss:  0.062
Epoch   8 Batch   54/137 - Train Accuracy:  0.955, Validation Accuracy:  0.947, Loss:  0.062
Epoch   8 Batch   55/137 - Train Accuracy:  0.954, Validation Accuracy:  0.950, Loss:  0.066
Epoch   8 Batch   56/137 - Train Accuracy:  0.948, Validation Accuracy:  0.955, Loss:  0.064
Epoch   8 Batch   57/137 - Train Accuracy:  0.950, Validation Accuracy:  0.954, Loss:  0.065
Epoch   8 Batch   58/137 - Train Accuracy:  0.950, Validation Accuracy:  0.951, Loss:  0.061
Epoch   8 Batch   59/137 - Train Accuracy:  0.948, Validation Accuracy:  0.951, Loss:  0.059
Epoch   8 Batch   60/137 - Train Accuracy:  0.952, Validation Accuracy:  0.949, Loss:  0.061
Epoch   8 Batch   61/137 - Train Accuracy:  0.943, Validation Accuracy:  0.954, Loss:  0.064
Epoch   8 Batch   62/137 - Train Accuracy:  0.949, Validation Accuracy:  0.954, Loss:  0.059
Epoch   8 Batch   63/137 - Train Accuracy:  0.950, Validation Accuracy:  0.953, Loss:  0.055
Epoch   8 Batch   64/137 - Train Accuracy:  0.940, Validation Accuracy:  0.955, Loss:  0.061
Epoch   8 Batch   65/137 - Train Accuracy:  0.945, Validation Accuracy:  0.958, Loss:  0.062
Epoch   8 Batch   66/137 - Train Accuracy:  0.948, Validation Accuracy:  0.957, Loss:  0.068
Epoch   8 Batch   67/137 - Train Accuracy:  0.943, Validation Accuracy:  0.958, Loss:  0.066
Epoch   8 Batch   68/137 - Train Accuracy:  0.957, Validation Accuracy:  0.960, Loss:  0.054
Epoch   8 Batch   69/137 - Train Accuracy:  0.945, Validation Accuracy:  0.957, Loss:  0.071
Epoch   8 Batch   70/137 - Train Accuracy:  0.949, Validation Accuracy:  0.955, Loss:  0.060
Epoch   8 Batch   71/137 - Train Accuracy:  0.944, Validation Accuracy:  0.955, Loss:  0.059
Epoch   8 Batch   72/137 - Train Accuracy:  0.947, Validation Accuracy:  0.957, Loss:  0.057
Epoch   8 Batch   73/137 - Train Accuracy:  0.960, Validation Accuracy:  0.957, Loss:  0.054
Epoch   8 Batch   74/137 - Train Accuracy:  0.951, Validation Accuracy:  0.957, Loss:  0.056
Epoch   8 Batch   75/137 - Train Accuracy:  0.951, Validation Accuracy:  0.956, Loss:  0.070
Epoch   8 Batch   76/137 - Train Accuracy:  0.955, Validation Accuracy:  0.951, Loss:  0.062
Epoch   8 Batch   77/137 - Train Accuracy:  0.954, Validation Accuracy:  0.951, Loss:  0.063
Epoch   8 Batch   78/137 - Train Accuracy:  0.953, Validation Accuracy:  0.951, Loss:  0.057
Epoch   8 Batch   79/137 - Train Accuracy:  0.953, Validation Accuracy:  0.952, Loss:  0.060
Epoch   8 Batch   80/137 - Train Accuracy:  0.945, Validation Accuracy:  0.957, Loss:  0.060
Epoch   8 Batch   81/137 - Train Accuracy:  0.948, Validation Accuracy:  0.954, Loss:  0.060
Epoch   8 Batch   82/137 - Train Accuracy:  0.952, Validation Accuracy:  0.952, Loss:  0.064
Epoch   8 Batch   83/137 - Train Accuracy:  0.957, Validation Accuracy:  0.954, Loss:  0.057
Epoch   8 Batch   84/137 - Train Accuracy:  0.954, Validation Accuracy:  0.953, Loss:  0.058
Epoch   8 Batch   85/137 - Train Accuracy:  0.949, Validation Accuracy:  0.951, Loss:  0.065
Epoch   8 Batch   86/137 - Train Accuracy:  0.947, Validation Accuracy:  0.950, Loss:  0.059
Epoch   8 Batch   87/137 - Train Accuracy:  0.955, Validation Accuracy:  0.954, Loss:  0.062
Epoch   8 Batch   88/137 - Train Accuracy:  0.954, Validation Accuracy:  0.953, Loss:  0.055
Epoch   8 Batch   89/137 - Train Accuracy:  0.948, Validation Accuracy:  0.952, Loss:  0.071
Epoch   8 Batch   90/137 - Train Accuracy:  0.959, Validation Accuracy:  0.953, Loss:  0.054
Epoch   8 Batch   91/137 - Train Accuracy:  0.955, Validation Accuracy:  0.953, Loss:  0.057
Epoch   8 Batch   92/137 - Train Accuracy:  0.952, Validation Accuracy:  0.955, Loss:  0.062
Epoch   8 Batch   93/137 - Train Accuracy:  0.955, Validation Accuracy:  0.956, Loss:  0.046
Epoch   8 Batch   94/137 - Train Accuracy:  0.951, Validation Accuracy:  0.955, Loss:  0.059
Epoch   8 Batch   95/137 - Train Accuracy:  0.951, Validation Accuracy:  0.957, Loss:  0.060
Epoch   8 Batch   96/137 - Train Accuracy:  0.955, Validation Accuracy:  0.961, Loss:  0.054
Epoch   8 Batch   97/137 - Train Accuracy:  0.944, Validation Accuracy:  0.957, Loss:  0.057
Epoch   8 Batch   98/137 - Train Accuracy:  0.954, Validation Accuracy:  0.957, Loss:  0.057
Epoch   8 Batch   99/137 - Train Accuracy:  0.946, Validation Accuracy:  0.959, Loss:  0.058
Epoch   8 Batch  100/137 - Train Accuracy:  0.948, Validation Accuracy:  0.950, Loss:  0.059
Epoch   8 Batch  101/137 - Train Accuracy:  0.953, Validation Accuracy:  0.954, Loss:  0.061
Epoch   8 Batch  102/137 - Train Accuracy:  0.958, Validation Accuracy:  0.961, Loss:  0.053
Epoch   8 Batch  103/137 - Train Accuracy:  0.947, Validation Accuracy:  0.955, Loss:  0.063
Epoch   8 Batch  104/137 - Train Accuracy:  0.956, Validation Accuracy:  0.955, Loss:  0.057
Epoch   8 Batch  105/137 - Train Accuracy:  0.948, Validation Accuracy:  0.957, Loss:  0.059
Epoch   8 Batch  106/137 - Train Accuracy:  0.958, Validation Accuracy:  0.954, Loss:  0.060
Epoch   8 Batch  107/137 - Train Accuracy:  0.950, Validation Accuracy:  0.954, Loss:  0.057
Epoch   8 Batch  108/137 - Train Accuracy:  0.949, Validation Accuracy:  0.957, Loss:  0.067
Epoch   8 Batch  109/137 - Train Accuracy:  0.951, Validation Accuracy:  0.959, Loss:  0.058
Epoch   8 Batch  110/137 - Train Accuracy:  0.946, Validation Accuracy:  0.961, Loss:  0.064
Epoch   8 Batch  111/137 - Train Accuracy:  0.951, Validation Accuracy:  0.959, Loss:  0.060
Epoch   8 Batch  112/137 - Train Accuracy:  0.951, Validation Accuracy:  0.958, Loss:  0.057
Epoch   8 Batch  113/137 - Train Accuracy:  0.953, Validation Accuracy:  0.957, Loss:  0.055
Epoch   8 Batch  114/137 - Train Accuracy:  0.949, Validation Accuracy:  0.960, Loss:  0.059
Epoch   8 Batch  115/137 - Train Accuracy:  0.958, Validation Accuracy:  0.960, Loss:  0.057
Epoch   8 Batch  116/137 - Train Accuracy:  0.964, Validation Accuracy:  0.956, Loss:  0.056
Epoch   8 Batch  117/137 - Train Accuracy:  0.949, Validation Accuracy:  0.957, Loss:  0.061
Epoch   8 Batch  118/137 - Train Accuracy:  0.952, Validation Accuracy:  0.958, Loss:  0.057
Epoch   8 Batch  119/137 - Train Accuracy:  0.957, Validation Accuracy:  0.957, Loss:  0.060
Epoch   8 Batch  120/137 - Train Accuracy:  0.960, Validation Accuracy:  0.958, Loss:  0.052
Epoch   8 Batch  121/137 - Train Accuracy:  0.961, Validation Accuracy:  0.957, Loss:  0.054
Epoch   8 Batch  122/137 - Train Accuracy:  0.954, Validation Accuracy:  0.956, Loss:  0.055
Epoch   8 Batch  123/137 - Train Accuracy:  0.944, Validation Accuracy:  0.957, Loss:  0.067
Epoch   8 Batch  124/137 - Train Accuracy:  0.955, Validation Accuracy:  0.957, Loss:  0.056
Epoch   8 Batch  125/137 - Train Accuracy:  0.951, Validation Accuracy:  0.959, Loss:  0.060
Epoch   8 Batch  126/137 - Train Accuracy:  0.958, Validation Accuracy:  0.957, Loss:  0.055
Epoch   8 Batch  127/137 - Train Accuracy:  0.955, Validation Accuracy:  0.956, Loss:  0.056
Epoch   8 Batch  128/137 - Train Accuracy:  0.957, Validation Accuracy:  0.954, Loss:  0.051
Epoch   8 Batch  129/137 - Train Accuracy:  0.949, Validation Accuracy:  0.951, Loss:  0.060
Epoch   8 Batch  130/137 - Train Accuracy:  0.951, Validation Accuracy:  0.949, Loss:  0.060
Epoch   8 Batch  131/137 - Train Accuracy:  0.952, Validation Accuracy:  0.950, Loss:  0.063
Epoch   8 Batch  132/137 - Train Accuracy:  0.948, Validation Accuracy:  0.951, Loss:  0.064
Epoch   8 Batch  133/137 - Train Accuracy:  0.958, Validation Accuracy:  0.956, Loss:  0.056
Epoch   8 Batch  134/137 - Train Accuracy:  0.945, Validation Accuracy:  0.958, Loss:  0.058
Epoch   8 Batch  135/137 - Train Accuracy:  0.949, Validation Accuracy:  0.958, Loss:  0.059
Epoch   9 Batch    0/137 - Train Accuracy:  0.955, Validation Accuracy:  0.960, Loss:  0.059
Epoch   9 Batch    1/137 - Train Accuracy:  0.947, Validation Accuracy:  0.963, Loss:  0.059
Epoch   9 Batch    2/137 - Train Accuracy:  0.953, Validation Accuracy:  0.960, Loss:  0.055
Epoch   9 Batch    3/137 - Train Accuracy:  0.958, Validation Accuracy:  0.963, Loss:  0.055
Epoch   9 Batch    4/137 - Train Accuracy:  0.959, Validation Accuracy:  0.962, Loss:  0.057
Epoch   9 Batch    5/137 - Train Accuracy:  0.957, Validation Accuracy:  0.959, Loss:  0.061
Epoch   9 Batch    6/137 - Train Accuracy:  0.955, Validation Accuracy:  0.958, Loss:  0.054
Epoch   9 Batch    7/137 - Train Accuracy:  0.958, Validation Accuracy:  0.961, Loss:  0.050
Epoch   9 Batch    8/137 - Train Accuracy:  0.950, Validation Accuracy:  0.959, Loss:  0.058
Epoch   9 Batch    9/137 - Train Accuracy:  0.963, Validation Accuracy:  0.961, Loss:  0.051
Epoch   9 Batch   10/137 - Train Accuracy:  0.953, Validation Accuracy:  0.964, Loss:  0.058
Epoch   9 Batch   11/137 - Train Accuracy:  0.953, Validation Accuracy:  0.965, Loss:  0.056
Epoch   9 Batch   12/137 - Train Accuracy:  0.946, Validation Accuracy:  0.963, Loss:  0.061
Epoch   9 Batch   13/137 - Train Accuracy:  0.954, Validation Accuracy:  0.964, Loss:  0.050
Epoch   9 Batch   14/137 - Train Accuracy:  0.953, Validation Accuracy:  0.964, Loss:  0.056
Epoch   9 Batch   15/137 - Train Accuracy:  0.955, Validation Accuracy:  0.965, Loss:  0.056
Epoch   9 Batch   16/137 - Train Accuracy:  0.959, Validation Accuracy:  0.964, Loss:  0.050
Epoch   9 Batch   17/137 - Train Accuracy:  0.953, Validation Accuracy:  0.960, Loss:  0.064
Epoch   9 Batch   18/137 - Train Accuracy:  0.957, Validation Accuracy:  0.960, Loss:  0.058
Epoch   9 Batch   19/137 - Train Accuracy:  0.952, Validation Accuracy:  0.958, Loss:  0.056
Epoch   9 Batch   20/137 - Train Accuracy:  0.946, Validation Accuracy:  0.959, Loss:  0.060
Epoch   9 Batch   21/137 - Train Accuracy:  0.961, Validation Accuracy:  0.956, Loss:  0.051
Epoch   9 Batch   22/137 - Train Accuracy:  0.951, Validation Accuracy:  0.956, Loss:  0.056
Epoch   9 Batch   23/137 - Train Accuracy:  0.956, Validation Accuracy:  0.959, Loss:  0.049
Epoch   9 Batch   24/137 - Train Accuracy:  0.952, Validation Accuracy:  0.959, Loss:  0.051
Epoch   9 Batch   25/137 - Train Accuracy:  0.951, Validation Accuracy:  0.958, Loss:  0.057
Epoch   9 Batch   26/137 - Train Accuracy:  0.956, Validation Accuracy:  0.958, Loss:  0.049
Epoch   9 Batch   27/137 - Train Accuracy:  0.960, Validation Accuracy:  0.959, Loss:  0.053
Epoch   9 Batch   28/137 - Train Accuracy:  0.951, Validation Accuracy:  0.960, Loss:  0.054
Epoch   9 Batch   29/137 - Train Accuracy:  0.959, Validation Accuracy:  0.959, Loss:  0.053
Epoch   9 Batch   30/137 - Train Accuracy:  0.960, Validation Accuracy:  0.959, Loss:  0.050
Epoch   9 Batch   31/137 - Train Accuracy:  0.954, Validation Accuracy:  0.960, Loss:  0.056
Epoch   9 Batch   32/137 - Train Accuracy:  0.960, Validation Accuracy:  0.960, Loss:  0.052
Epoch   9 Batch   33/137 - Train Accuracy:  0.952, Validation Accuracy:  0.958, Loss:  0.057
Epoch   9 Batch   34/137 - Train Accuracy:  0.945, Validation Accuracy:  0.961, Loss:  0.061
Epoch   9 Batch   35/137 - Train Accuracy:  0.950, Validation Accuracy:  0.961, Loss:  0.063
Epoch   9 Batch   36/137 - Train Accuracy:  0.954, Validation Accuracy:  0.960, Loss:  0.059
Epoch   9 Batch   37/137 - Train Accuracy:  0.957, Validation Accuracy:  0.960, Loss:  0.057
Epoch   9 Batch   38/137 - Train Accuracy:  0.955, Validation Accuracy:  0.959, Loss:  0.055
Epoch   9 Batch   39/137 - Train Accuracy:  0.958, Validation Accuracy:  0.959, Loss:  0.054
Epoch   9 Batch   40/137 - Train Accuracy:  0.955, Validation Accuracy:  0.962, Loss:  0.054
Epoch   9 Batch   41/137 - Train Accuracy:  0.956, Validation Accuracy:  0.962, Loss:  0.057
Epoch   9 Batch   42/137 - Train Accuracy:  0.957, Validation Accuracy:  0.964, Loss:  0.061
Epoch   9 Batch   43/137 - Train Accuracy:  0.957, Validation Accuracy:  0.966, Loss:  0.050
Epoch   9 Batch   44/137 - Train Accuracy:  0.951, Validation Accuracy:  0.960, Loss:  0.057
Epoch   9 Batch   45/137 - Train Accuracy:  0.952, Validation Accuracy:  0.960, Loss:  0.054
Epoch   9 Batch   46/137 - Train Accuracy:  0.965, Validation Accuracy:  0.961, Loss:  0.048
Epoch   9 Batch   47/137 - Train Accuracy:  0.962, Validation Accuracy:  0.963, Loss:  0.054
Epoch   9 Batch   48/137 - Train Accuracy:  0.957, Validation Accuracy:  0.961, Loss:  0.057
Epoch   9 Batch   49/137 - Train Accuracy:  0.949, Validation Accuracy:  0.960, Loss:  0.060
Epoch   9 Batch   50/137 - Train Accuracy:  0.954, Validation Accuracy:  0.954, Loss:  0.053
Epoch   9 Batch   51/137 - Train Accuracy:  0.955, Validation Accuracy:  0.960, Loss:  0.057
Epoch   9 Batch   52/137 - Train Accuracy:  0.962, Validation Accuracy:  0.959, Loss:  0.056
Epoch   9 Batch   53/137 - Train Accuracy:  0.953, Validation Accuracy:  0.962, Loss:  0.056
Epoch   9 Batch   54/137 - Train Accuracy:  0.963, Validation Accuracy:  0.964, Loss:  0.052
Epoch   9 Batch   55/137 - Train Accuracy:  0.952, Validation Accuracy:  0.961, Loss:  0.052
Epoch   9 Batch   56/137 - Train Accuracy:  0.955, Validation Accuracy:  0.959, Loss:  0.052
Epoch   9 Batch   57/137 - Train Accuracy:  0.949, Validation Accuracy:  0.957, Loss:  0.054
Epoch   9 Batch   58/137 - Train Accuracy:  0.955, Validation Accuracy:  0.959, Loss:  0.056
Epoch   9 Batch   59/137 - Train Accuracy:  0.959, Validation Accuracy:  0.960, Loss:  0.051
Epoch   9 Batch   60/137 - Train Accuracy:  0.955, Validation Accuracy:  0.960, Loss:  0.054
Epoch   9 Batch   61/137 - Train Accuracy:  0.954, Validation Accuracy:  0.959, Loss:  0.055
Epoch   9 Batch   62/137 - Train Accuracy:  0.955, Validation Accuracy:  0.956, Loss:  0.050
Epoch   9 Batch   63/137 - Train Accuracy:  0.955, Validation Accuracy:  0.956, Loss:  0.046
Epoch   9 Batch   64/137 - Train Accuracy:  0.951, Validation Accuracy:  0.959, Loss:  0.053
Epoch   9 Batch   65/137 - Train Accuracy:  0.951, Validation Accuracy:  0.961, Loss:  0.057
Epoch   9 Batch   66/137 - Train Accuracy:  0.955, Validation Accuracy:  0.960, Loss:  0.058
Epoch   9 Batch   67/137 - Train Accuracy:  0.949, Validation Accuracy:  0.961, Loss:  0.055
Epoch   9 Batch   68/137 - Train Accuracy:  0.960, Validation Accuracy:  0.961, Loss:  0.048
Epoch   9 Batch   69/137 - Train Accuracy:  0.948, Validation Accuracy:  0.962, Loss:  0.061
Epoch   9 Batch   70/137 - Train Accuracy:  0.950, Validation Accuracy:  0.958, Loss:  0.047
Epoch   9 Batch   71/137 - Train Accuracy:  0.954, Validation Accuracy:  0.958, Loss:  0.055
Epoch   9 Batch   72/137 - Train Accuracy:  0.953, Validation Accuracy:  0.959, Loss:  0.052
Epoch   9 Batch   73/137 - Train Accuracy:  0.959, Validation Accuracy:  0.962, Loss:  0.050
Epoch   9 Batch   74/137 - Train Accuracy:  0.959, Validation Accuracy:  0.959, Loss:  0.047
Epoch   9 Batch   75/137 - Train Accuracy:  0.958, Validation Accuracy:  0.961, Loss:  0.059
Epoch   9 Batch   76/137 - Train Accuracy:  0.957, Validation Accuracy:  0.957, Loss:  0.054
Epoch   9 Batch   77/137 - Train Accuracy:  0.957, Validation Accuracy:  0.958, Loss:  0.055
Epoch   9 Batch   78/137 - Train Accuracy:  0.961, Validation Accuracy:  0.961, Loss:  0.052
Epoch   9 Batch   79/137 - Train Accuracy:  0.960, Validation Accuracy:  0.962, Loss:  0.052
Epoch   9 Batch   80/137 - Train Accuracy:  0.947, Validation Accuracy:  0.960, Loss:  0.051
Epoch   9 Batch   81/137 - Train Accuracy:  0.955, Validation Accuracy:  0.961, Loss:  0.057
Epoch   9 Batch   82/137 - Train Accuracy:  0.963, Validation Accuracy:  0.960, Loss:  0.049
Epoch   9 Batch   83/137 - Train Accuracy:  0.966, Validation Accuracy:  0.960, Loss:  0.050
Epoch   9 Batch   84/137 - Train Accuracy:  0.955, Validation Accuracy:  0.960, Loss:  0.051
Epoch   9 Batch   85/137 - Train Accuracy:  0.958, Validation Accuracy:  0.961, Loss:  0.057
Epoch   9 Batch   86/137 - Train Accuracy:  0.954, Validation Accuracy:  0.960, Loss:  0.051
Epoch   9 Batch   87/137 - Train Accuracy:  0.956, Validation Accuracy:  0.958, Loss:  0.055
Epoch   9 Batch   88/137 - Train Accuracy:  0.958, Validation Accuracy:  0.958, Loss:  0.048
Epoch   9 Batch   89/137 - Train Accuracy:  0.947, Validation Accuracy:  0.958, Loss:  0.064
Epoch   9 Batch   90/137 - Train Accuracy:  0.962, Validation Accuracy:  0.960, Loss:  0.047
Epoch   9 Batch   91/137 - Train Accuracy:  0.957, Validation Accuracy:  0.966, Loss:  0.052
Epoch   9 Batch   92/137 - Train Accuracy:  0.956, Validation Accuracy:  0.966, Loss:  0.053
Epoch   9 Batch   93/137 - Train Accuracy:  0.962, Validation Accuracy:  0.964, Loss:  0.046
Epoch   9 Batch   94/137 - Train Accuracy:  0.959, Validation Accuracy:  0.962, Loss:  0.052
Epoch   9 Batch   95/137 - Train Accuracy:  0.957, Validation Accuracy:  0.966, Loss:  0.050
Epoch   9 Batch   96/137 - Train Accuracy:  0.964, Validation Accuracy:  0.965, Loss:  0.047
Epoch   9 Batch   97/137 - Train Accuracy:  0.956, Validation Accuracy:  0.964, Loss:  0.053
Epoch   9 Batch   98/137 - Train Accuracy:  0.961, Validation Accuracy:  0.963, Loss:  0.052
Epoch   9 Batch   99/137 - Train Accuracy:  0.953, Validation Accuracy:  0.961, Loss:  0.052
Epoch   9 Batch  100/137 - Train Accuracy:  0.952, Validation Accuracy:  0.964, Loss:  0.051
Epoch   9 Batch  101/137 - Train Accuracy:  0.964, Validation Accuracy:  0.968, Loss:  0.056
Epoch   9 Batch  102/137 - Train Accuracy:  0.965, Validation Accuracy:  0.967, Loss:  0.048
Epoch   9 Batch  103/137 - Train Accuracy:  0.958, Validation Accuracy:  0.967, Loss:  0.056
Epoch   9 Batch  104/137 - Train Accuracy:  0.966, Validation Accuracy:  0.963, Loss:  0.047
Epoch   9 Batch  105/137 - Train Accuracy:  0.953, Validation Accuracy:  0.963, Loss:  0.051
Epoch   9 Batch  106/137 - Train Accuracy:  0.960, Validation Accuracy:  0.965, Loss:  0.052
Epoch   9 Batch  107/137 - Train Accuracy:  0.958, Validation Accuracy:  0.968, Loss:  0.051
Epoch   9 Batch  108/137 - Train Accuracy:  0.956, Validation Accuracy:  0.970, Loss:  0.058
Epoch   9 Batch  109/137 - Train Accuracy:  0.962, Validation Accuracy:  0.970, Loss:  0.047
Epoch   9 Batch  110/137 - Train Accuracy:  0.951, Validation Accuracy:  0.972, Loss:  0.057
Epoch   9 Batch  111/137 - Train Accuracy:  0.960, Validation Accuracy:  0.962, Loss:  0.053
Epoch   9 Batch  112/137 - Train Accuracy:  0.952, Validation Accuracy:  0.958, Loss:  0.048
Epoch   9 Batch  113/137 - Train Accuracy:  0.958, Validation Accuracy:  0.959, Loss:  0.048
Epoch   9 Batch  114/137 - Train Accuracy:  0.955, Validation Accuracy:  0.963, Loss:  0.055
Epoch   9 Batch  115/137 - Train Accuracy:  0.961, Validation Accuracy:  0.965, Loss:  0.053
Epoch   9 Batch  116/137 - Train Accuracy:  0.966, Validation Accuracy:  0.964, Loss:  0.047
Epoch   9 Batch  117/137 - Train Accuracy:  0.955, Validation Accuracy:  0.964, Loss:  0.053
Epoch   9 Batch  118/137 - Train Accuracy:  0.957, Validation Accuracy:  0.968, Loss:  0.052
Epoch   9 Batch  119/137 - Train Accuracy:  0.966, Validation Accuracy:  0.965, Loss:  0.052
Epoch   9 Batch  120/137 - Train Accuracy:  0.969, Validation Accuracy:  0.964, Loss:  0.049
Epoch   9 Batch  121/137 - Train Accuracy:  0.965, Validation Accuracy:  0.966, Loss:  0.049
Epoch   9 Batch  122/137 - Train Accuracy:  0.958, Validation Accuracy:  0.969, Loss:  0.048
Epoch   9 Batch  123/137 - Train Accuracy:  0.954, Validation Accuracy:  0.965, Loss:  0.059
Epoch   9 Batch  124/137 - Train Accuracy:  0.966, Validation Accuracy:  0.964, Loss:  0.045
Epoch   9 Batch  125/137 - Train Accuracy:  0.956, Validation Accuracy:  0.964, Loss:  0.056
Epoch   9 Batch  126/137 - Train Accuracy:  0.964, Validation Accuracy:  0.964, Loss:  0.048
Epoch   9 Batch  127/137 - Train Accuracy:  0.958, Validation Accuracy:  0.963, Loss:  0.049
Epoch   9 Batch  128/137 - Train Accuracy:  0.965, Validation Accuracy:  0.964, Loss:  0.046
Epoch   9 Batch  129/137 - Train Accuracy:  0.956, Validation Accuracy:  0.964, Loss:  0.053
Epoch   9 Batch  130/137 - Train Accuracy:  0.961, Validation Accuracy:  0.965, Loss:  0.053
Epoch   9 Batch  131/137 - Train Accuracy:  0.961, Validation Accuracy:  0.964, Loss:  0.051
Epoch   9 Batch  132/137 - Train Accuracy:  0.955, Validation Accuracy:  0.965, Loss:  0.056
Epoch   9 Batch  133/137 - Train Accuracy:  0.963, Validation Accuracy:  0.964, Loss:  0.049
Epoch   9 Batch  134/137 - Train Accuracy:  0.955, Validation Accuracy:  0.965, Loss:  0.053
Epoch   9 Batch  135/137 - Train Accuracy:  0.960, Validation Accuracy:  0.963, Loss:  0.050
Model Trained and Saved

Save Parameters

Save the batch_size and save_path parameters for inference.





In [80]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
# Save parameters for checkpoint
helper.save_params(save_path)

Checkpoint





In [81]:



    


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import tensorflow as tf
import numpy as np
import helper
import problem_unittests as tests

_, (source_vocab_to_int, target_vocab_to_int), (source_int_to_vocab, target_int_to_vocab) = helper.load_preprocess()
load_path = helper.load_params()

Sentence to Sequence

To feed a sentence into the model for translation, you first need to preprocess it. Implement the function sentence_to_seq() to preprocess new sentences.

  • Convert the sentence to lowercase
  • Convert words into ids using vocab_to_int
  • Convert words not in the vocabulary, to the <UNK> word id.




In [82]:



    


def sentence_to_seq(sentence, vocab_to_int):
    """
    Convert a sentence to a sequence of ids
    :param sentence: String
    :param vocab_to_int: Dictionary to go from the words to an id
    :return: List of word ids
    """
    # TODO: Implement Function
    
    return [vocab_to_int.get(word,vocab_to_int['<UNK>']) for word in sentence.lower().split()]


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_sentence_to_seq(sentence_to_seq)



    


















    






Tests Passed

Translate

This will translate translate_sentence from English to French.





In [83]:



    


translate_sentence = 'he saw a old yellow truck .'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
translate_sentence = sentence_to_seq(translate_sentence, source_vocab_to_int)

loaded_graph = tf.Graph()
with tf.Session(graph=loaded_graph) as sess:
    # Load saved model
    loader = tf.train.import_meta_graph(load_path + '.meta')
    loader.restore(sess, load_path)

    input_data = loaded_graph.get_tensor_by_name('input:0')
    logits = loaded_graph.get_tensor_by_name('logits:0')
    keep_prob = loaded_graph.get_tensor_by_name('keep_prob:0')

    translate_logits = sess.run(logits, {input_data: [translate_sentence], keep_prob: 1.0})[0]

print('Input')
print('  Word Ids:      {}'.format([i for i in translate_sentence]))
print('  English Words: {}'.format([source_int_to_vocab[i] for i in translate_sentence]))

print('\nPrediction')
print('  Word Ids:      {}'.format([i for i in np.argmax(translate_logits, 1)]))
print('  French Words: {}'.format([target_int_to_vocab[i] for i in np.argmax(translate_logits, 1)]))



    


















    






Input
  Word Ids:      [30, 4, 65, 5, 17, 72, 166]
  English Words: ['he', 'saw', 'a', 'old', 'yellow', 'truck', '.']

Prediction
  Word Ids:      [217, 294, 355, 78, 235, 345, 335, 1]
  French Words: ['il', 'a', 'vu', 'un', 'vieux', 'et', '.', '<EOS>']

Imperfect Translation

You might notice that some sentences translate better than others. Since the dataset you're using only has a vocabulary of 227 English words of the thousands that you use, you're only going to see good results using these words. Additionally, the translations in this data set were made by Google translate, so the translations themselves aren't particularly good. (We apologize to the French speakers out there!) Thankfully, for this project, you don't need a perfect translation. However, if you want to create a better translation model, you'll need better data.

You can train on the WMT10 French-English corpus. This dataset has more vocabulary and richer in topics discussed. However, this will take you days to train, so make sure you've a GPU and the neural network is performing well on dataset we provided. Just make sure you play with the WMT10 corpus after you've submitted this project.

Submitting This Project

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_language_translation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.

Content source: zhuanxuhit/deep-learning

Similar notebooks:

notebook.community | gallery | about