def load_data():
''' '''
print('Loading data...')
(x_train, y_train), (x_test,
y_test) = imdb.load_data(num_words=max_features)
print(len(x_train), 'train sequences')
print(len(x_test), 'test sequences')
print('Pad sequences (samples x time)')
x_train = sequence.pad_sequences(x_train, maxlen=maxlen)
x_test = sequence.pad_sequences(x_test, maxlen=maxlen)
print('x_train shape:', x_train.shape)
print('x_test shape:', x_test.shape)
return [x_train, y_train, x_test, y_test]
def main(unused_argv):
# Loading the data
# data from: https://keras.io/datasets/
# Dataset of 25,000 movies reviews from IMDB, labeled by sentiment
# (positive/negative).
# Reviews have been preprocessed, and each review is encoded as a sequence
# of word indexes (integers).
# For convenience, words are indexed by overall frequency in the dataset.
print('Loading data...')
(x_train, y_train), (x_test, y_test) = imdb.load_data(
num_words=FLAGS.num_words)
print('size of the train dataset:', x_train.shape[0])
print('size of the test dataset:', x_test.shape[0])
# run experiment
run_config = tf.contrib.learn.RunConfig(model_dir=FLAGS.model_dir)
learn_runner.run(generate_experiment_fn(x_train, y_train, x_test, y_test),
run_config=run_config)
def main(unused_argv):
# Loading the data
# data from: https://keras.io/datasets/
# Dataset of 25,000 movies reviews from IMDB, labeled by sentiment
# (positive/negative).
# Reviews have been preprocessed, and each review is encoded as a sequence
# of word indexes (integers).
# For convenience, words are indexed by overall frequency in the dataset.
print('Loading data...')
(x_train, y_train), (x_test,
y_test) = imdb.load_data(num_words=FLAGS.num_words)
print('size of the train dataset:', x_train.shape[0])
print('size of the test dataset:', x_test.shape[0])
# run config
run_config = tf.estimator.RunConfig()
# creates estimator
if FLAGS.use_canned_estimator:
xc = tf.contrib.layers.sparse_column_with_integerized_feature(
'x', FLAGS.num_words)
xc = tf.contrib.layers.embedding_column(xc, FLAGS.embed_dim)
# creates estimator
estimator = tf.contrib.learn.DynamicRnnEstimator(
config=run_config,
model_dir=FLAGS.model_dir,
problem_type=constants.ProblemType.CLASSIFICATION,
prediction_type=PredictionType.SINGLE_VALUE,
sequence_feature_columns=[xc],
context_feature_columns=None,
num_units=FLAGS.num_rnn_units,
cell_type=FLAGS.cell_type,
optimizer=FLAGS.optimizer,
learning_rate=FLAGS.learning_rate,
num_classes=FLAGS.num_classes,
dropout_keep_probabilities=FLAGS.dropout_keep_probabilities)
else:
model_fn = CustomRNNEstimator(rnn_cell_sizes=FLAGS.num_rnn_units,
label_dimension=FLAGS.num_classes,
num_words=FLAGS.num_words,
dnn_layer_sizes=FLAGS.num_dnn_units,
optimizer=FLAGS.optimizer,
learning_rate=FLAGS.learning_rate,
embed_dim=FLAGS.embed_dim)
estimator = tf.estimator.Estimator(model_fn=model_fn,
model_dir=FLAGS.model_dir,
config=run_config)
# input functions
train_input = get_input_fn(x_train,
y_train,
FLAGS.train_batch_size,
epochs=FLAGS.num_epochs,
max_length=FLAGS.max_len,
batch_by_seq_len=FLAGS.batch_by_seq_len)
test_input = get_input_fn(x_test,
y_test,
FLAGS.eval_batch_size,
epochs=1,
max_length=FLAGS.max_len)
# training
# estimator.train(input_fn=train_input)
# evalutaion
# estimator.evaluate(input_fn=test_input)
# predict
predictions = list(estimator.predict(input_fn=test_input))
# loading map from word to index and index to word
word_to_index, index_to_word = _load_map_dicts()
for i in range(5):
index = random.randint(
0, x_test.shape[0]) # choose random index in the test dataset
print(_ids_to_sentence(x_test[index], index_to_word))
print('Prediction:', predictions[index])
print('Label:', y_test[index])
from tensorflow.contrib.keras.python.keras.datasets import imdb
from tensorflow.contrib.keras.python.keras.layers import Embedding, SimpleRNN, Dropout, Dense, Activation, LSTM, GRU
from tensorflow.contrib.keras.python.keras.models import Sequential
from tensorflow.contrib.keras.python.keras.preprocessing import sequence
max_features = 20000
maxlen = 100
batch_size = 32
(X_train, y_train), (X_test, y_test) = imdb.load_data(num_words=max_features)
X_train = sequence.pad_sequences(X_train, maxlen=maxlen)
X_test = sequence.pad_sequences(X_test, maxlen=maxlen)
print('X_train shape:', X_train.shape)
print('X_test shape:', X_test.shape)
model = Sequential()
model.add(Embedding(max_features, 128, input_length=maxlen))
#model.add(SimpleRNN(128))
#model.add(GRU(128))
model.add(LSTM(128))
model.add(Dropout(0.5))
model.add(Dense(1))
model.add(Activation('sigmoid'))
model.compile(loss='binary_crossentropy', optimizer='adam')
# Global params:
NB_WORDS = 20000
SKIP_TOP = 0
TEST_SPLIT = 0.2
INIT_SEED = 2017
GLOBAL_SEED = 2018
MAXLEN = 80
BATCH_SIZE = 128
TEST_BATCH_SIZE = 512
# In[2]:
(X_train, Y_train), (X_test, Y_test) = imdb.load_data(num_words=NB_WORDS)
print(len(X_train), 'train sequences')
print(len(X_test), 'test sequences')
print('Pad sequences (samples x time)')
X_train = sequence.pad_sequences(X_train, maxlen=MAXLEN)
X_test = sequence.pad_sequences(X_test, maxlen=MAXLEN)
print('x_train shape:', X_train.shape)
print('x_test shape:', X_test.shape)
# In[3]:
class Model(nn.Module):
def __init__(self, nb_words, hidden_size=128, embedding_size=128, n_layers=1,
wdrop=0.25, odrop=0.25, edrop=0.1, idrop=0.25, variational=False,
def main(unused_argv):
# loading data
(x_train, y_train), (x_test,
y_test) = imdb.load_data(maxlen=FLAGS.max_len,
num_words=FLAGS.num_words,
index_from=FLAGS.index_from)
# loading map from word to index and index to word
word_to_index, index_to_word = _load_map_dicts()
# run config
run_config = tf.estimator.RunConfig()
# loading estimators
if FLAGS.use_canned_estimator:
xc = tf.contrib.layers.sparse_column_with_integerized_feature(
'x', FLAGS.num_words)
xc = tf.contrib.layers.embedding_column(xc, FLAGS.embed_dim)
# creates estimator
estimator = tf.contrib.learn.DynamicRnnEstimator(
config=run_config,
model_dir=FLAGS.model_dir,
problem_type=constants.ProblemType.CLASSIFICATION,
prediction_type=PredictionType.SINGLE_VALUE,
sequence_feature_columns=[xc],
context_feature_columns=None,
num_units=FLAGS.num_rnn_units,
cell_type=FLAGS.cell_type,
optimizer=FLAGS.optimizer,
learning_rate=FLAGS.learning_rate,
num_classes=FLAGS.num_classes,
dropout_keep_probabilities=FLAGS.dropout_keep_probabilities)
else:
model_fn = CustomRNNEstimator(rnn_cell_sizes=FLAGS.num_rnn_units,
label_dimension=FLAGS.num_classes,
num_words=FLAGS.num_words,
dnn_layer_sizes=FLAGS.num_dnn_units,
optimizer=FLAGS.optimizer,
learning_rate=FLAGS.learning_rate,
embed_dim=FLAGS.embed_dim)
estimator = tf.estimator.Estimator(model_fn=model_fn,
model_dir=FLAGS.model_dir,
config=run_config)
# getting test input_function
test_input = get_input_fn(x_test,
y_test,
FLAGS.eval_batch_size,
epochs=1,
max_length=FLAGS.max_len,
shuffle=False)
predictions = list(estimator.predict(input_fn=test_input))
for i in range(5):
print(_ids_to_sentence(x_test[i], index_to_word))
print('Prediction:', predictions[i])
print('Label:', y_test[i])
