mirror of
https://github.com/explosion/spaCy.git
synced 2024-12-26 09:56:28 +03:00
Fix use of dropout in sentiment analysis LSTM example
This commit is contained in:
parent
1ed40682a3
commit
7793e2ad82
|
@ -111,10 +111,9 @@ def compile_lstm(embeddings, shape, settings):
|
||||||
mask_zero=True
|
mask_zero=True
|
||||||
)
|
)
|
||||||
)
|
)
|
||||||
model.add(TimeDistributed(Dense(shape['nr_hidden'] * 2, bias=False)))
|
model.add(TimeDistributed(Dense(shape['nr_hidden'], bias=False)))
|
||||||
model.add(Dropout(settings['dropout']))
|
model.add(Bidirectional(LSTM(shape['nr_hidden'], dropout_U=settings['dropout'],
|
||||||
model.add(Bidirectional(LSTM(shape['nr_hidden'])))
|
dropout_W=settings['dropout'])))
|
||||||
model.add(Dropout(settings['dropout']))
|
|
||||||
model.add(Dense(shape['nr_class'], activation='sigmoid'))
|
model.add(Dense(shape['nr_class'], activation='sigmoid'))
|
||||||
model.compile(optimizer=Adam(lr=settings['lr']), loss='binary_crossentropy',
|
model.compile(optimizer=Adam(lr=settings['lr']), loss='binary_crossentropy',
|
||||||
metrics=['accuracy'])
|
metrics=['accuracy'])
|
||||||
|
@ -195,7 +194,7 @@ def main(model_dir, train_dir, dev_dir,
|
||||||
dev_labels = numpy.asarray(dev_labels, dtype='int32')
|
dev_labels = numpy.asarray(dev_labels, dtype='int32')
|
||||||
lstm = train(train_texts, train_labels, dev_texts, dev_labels,
|
lstm = train(train_texts, train_labels, dev_texts, dev_labels,
|
||||||
{'nr_hidden': nr_hidden, 'max_length': max_length, 'nr_class': 1},
|
{'nr_hidden': nr_hidden, 'max_length': max_length, 'nr_class': 1},
|
||||||
{'dropout': 0.5, 'lr': learn_rate},
|
{'dropout': dropout, 'lr': learn_rate},
|
||||||
{},
|
{},
|
||||||
nb_epoch=nb_epoch, batch_size=batch_size)
|
nb_epoch=nb_epoch, batch_size=batch_size)
|
||||||
weights = lstm.get_weights()
|
weights = lstm.get_weights()
|
||||||
|
|
Loading…
Reference in New Issue
Block a user