f = open(fasttext_file, 'r', encoding='utf8')
lines = f.readlines()[1:]
f.close()
for line in lines:
values = line.split()
word = values[0]
value = np.asarray(values[1:], dtype='float32')
embeddings[word] = value
print('Loaded %s word vectors.' % len(embeddings))
print(embeddings['man'])
print(embeddings['woman'])
print(embeddings['guy'])
print(embeddings['boy'])
word_vectors = list(embeddings.values())[100:150]
labels = list(embeddings.keys())[100:150]
kutils.viz_vectors(word_vectors, labels)
kutils.viz_vectors_corr(word_vectors, labels)
kutils.viz_vectors_lower_dim(word_vectors, labels)
word_vectors = [
embeddings['man'], embeddings['woman'], embeddings['boy'],
embeddings['guy'], embeddings['cat'], embeddings['dog']
]
labels = ['man', 'woman', 'boy', 'guy', 'cat', 'dog']
kutils.viz_vectors(word_vectors, labels)
kutils.viz_vectors_corr(word_vectors, labels)
kutils.viz_vectors_lower_dim(word_vectors, labels)
#build model
input = Input(shape=(X_train.shape[1], ))
inner = Embedding(input_dim=vocab_size,
output_dim=word_embed_size,
input_length=seq_maxlen,
weights=[emb_layer_weights],
trainable=False)(input)
lstm1 = LSTM(200, return_sequences=True)(inner)
features = LSTM(100, return_sequences=False)(lstm1)
model = Model(inputs=input, outputs=features)
print(model.summary())
feature_vectors = model.predict(X_train)
print(feature_vectors)
#visualize sentence vectors
kutils.viz_vectors(feature_vectors, data)
kutils.viz_vectors_corr(feature_vectors, data)
kutils.viz_vectors_lower_dim(feature_vectors, data)
#visualize layer activations
act = kutils.get_activations(model, X_train[0:1]) # with just one sample.
kutils.display_activations(act,
directory=os.path.join("G:/", 'digit_activations'),
save=True)
kutils.display_heatmaps(act,
tmp[0:1],
directory=os.path.join("G:/", 'digit_heatmaps'),
save=True)