def test_training(self):
model = FT_gensim(size=10, min_count=1, hs=1, negative=0, seed=42, workers=1)
model.build_vocab(sentences)
self.model_sanity(model)
model.train(sentences, total_examples=model.corpus_count, epochs=model.iter)
sims = model.most_similar('graph', topn=10)
self.assertEqual(model.wv.syn0.shape, (12, 10))
self.assertEqual(len(model.wv.vocab), 12)
self.assertEqual(model.wv.syn0_vocab.shape[1], 10)
self.assertEqual(model.wv.syn0_ngrams.shape[1], 10)
self.model_sanity(model)
# test querying for "most similar" by vector
graph_vector = model.wv.syn0norm[model.wv.vocab['graph'].index]
sims2 = model.most_similar(positive=[graph_vector], topn=11)
sims2 = [(w, sim) for w, sim in sims2 if w != 'graph'] # ignore 'graph' itself
self.assertEqual(sims, sims2)
# build vocab and train in one step; must be the same as above
model2 = FT_gensim(sentences, size=10, min_count=1, hs=1, negative=0, seed=42, workers=1)
self.models_equal(model, model2)
# verify oov-word vector retrieval
invocab_vec = model['minors'] # invocab word
self.assertEqual(len(invocab_vec), 10)
oov_vec = model['minor'] # oov word
self.assertEqual(len(oov_vec), 10)
def test_training_fromfile(self):
with temporary_file(get_tmpfile('gensim_fasttext.tst')) as corpus_file:
utils.save_as_line_sentence(sentences, corpus_file)
model = FT_gensim(size=10, min_count=1, hs=1, negative=0, seed=42, workers=1)
model.build_vocab(corpus_file=corpus_file)
self.model_sanity(model)
model.train(corpus_file=corpus_file, total_words=model.corpus_total_words, epochs=model.iter)
sims = model.most_similar('graph', topn=10)
self.assertEqual(model.wv.syn0.shape, (12, 10))
self.assertEqual(len(model.wv.vocab), 12)
self.assertEqual(model.wv.syn0_vocab.shape[1], 10)
self.assertEqual(model.wv.syn0_ngrams.shape[1], 10)
self.model_sanity(model)
# test querying for "most similar" by vector
graph_vector = model.wv.syn0norm[model.wv.vocab['graph'].index]
sims2 = model.most_similar(positive=[graph_vector], topn=11)
sims2 = [(w, sim) for w, sim in sims2 if w != 'graph'] # ignore 'graph' itself
self.assertEqual(sims, sims2)
# verify oov-word vector retrieval
invocab_vec = model['minors'] # invocab word
self.assertEqual(len(invocab_vec), 10)
oov_vec = model['minor'] # oov word
self.assertEqual(len(oov_vec), 10)
def test_training(self):
model = FT_gensim(size=10, min_count=1, hs=1, negative=0, seed=42, workers=1)
model.build_vocab(sentences)
self.model_sanity(model)
model.train(sentences, total_examples=model.corpus_count, epochs=model.iter)
sims = model.most_similar('graph', topn=10)
self.assertEqual(model.wv.syn0.shape, (12, 10))
self.assertEqual(len(model.wv.vocab), 12)
self.assertEqual(model.wv.syn0_vocab.shape[1], 10)
self.assertEqual(model.wv.syn0_ngrams.shape[1], 10)
self.model_sanity(model)
# test querying for "most similar" by vector
graph_vector = model.wv.syn0norm[model.wv.vocab['graph'].index]
sims2 = model.most_similar(positive=[graph_vector], topn=11)
sims2 = [(w, sim) for w, sim in sims2 if w != 'graph'] # ignore 'graph' itself
self.assertEqual(sims, sims2)
# build vocab and train in one step; must be the same as above
model2 = FT_gensim(sentences, size=10, min_count=1, hs=1, negative=0, seed=42, workers=1)
self.models_equal(model, model2)
# verify oov-word vector retrieval
invocab_vec = model['minors'] # invocab word
self.assertEqual(len(invocab_vec), 10)
oov_vec = model['minor'] # oov word
self.assertEqual(len(oov_vec), 10)
#
print(model.similarity("night", "nights"))
###############################################################################
#
# Syntactically similar words generally have high similarity in fastText models, since a large number of the component char-ngrams will be the same. As a result, fastText generally does better at syntactic tasks than Word2Vec. A detailed comparison is provided `here <Word2Vec_FastText_Comparison.ipynb>`_.
#
###############################################################################
#
# Other similarity operations
# ^^^^^^^^^^^^^^^^^^^^^^^^^^^
#
# The example training corpus is a toy corpus, results are not expected to be good, for proof-of-concept only
print(model.most_similar("nights"))
###############################################################################
#
print(model.n_similarity(['sushi', 'shop'], ['japanese', 'restaurant']))
###############################################################################
#
print(model.doesnt_match("breakfast cereal dinner lunch".split()))
###############################################################################
#
print(model.most_similar(positive=['baghdad', 'england'], negative=['london']))
###############################################################################
#
for activation in activations:
if activation == 0:
for negat in negatives:
ftxt= FastText(min_count = min_cnt, size = 300, window = 3, workers = 3, sg = architecture, hs = activation, negative = negat, seed = 0)
ftxt.build_vocab(sentences = sent)
ftxt.train(sentences = sent, total_examples = len(sent), epochs = 30)
keys = ['sval', 'oko', 'srdce']
embedding_clusters = []
word_clusters = []
for word in keys:
embeddings = []
words = []
for similar_word, _ in ftxt.most_similar(word, topn = 30):
words.append(similar_word)
embeddings.append(ftxt[similar_word])
embedding_clusters.append(embeddings)
word_clusters.append(words)
embedding_clusters = np.array(embedding_clusters)
n, m, k = embedding_clusters.shape
tsne_model_en_2d = TSNE(perplexity = 15, n_components = 2, init = 'pca', n_iter = 3500, random_state = 32)
embeddings_en_2d = np.array(tsne_model_en_2d.fit_transform(embedding_clusters.reshape(n * m, k))).reshape(n, m, 2)
tsne_plot_similar_words('Sval oko srdce', keys, embeddings_en_2d, word_clusters, 0.7,
f'Sval oko srdce FastText OPTIMED NZIP min_count = ' + str(min_cnt) + ' , size = 300, workers = 3, window = 3, sg = ' + str(architecture) +', hs = ' + str(activation) + ', negative = '+ str(negat) + ', seed = 0.png')
else:
ftxt= FastText(min_count = min_cnt, size = 300, window = 3, workers = 3, sg = architecture, hs = activation, seed = 0)
## Note: Not necessary. Uses sentences.csv from the word2vec model
# %% Train Fasttext
from gensim.models.fasttext import FastText as FT_gensim
from gensim.models.word2vec import LineSentence
sentences = LineSentence(dataPath + "sentences.csv")
model_gensim = FT_gensim(size=300)
# build the vocabulary
model_gensim.build_vocab(sentences)
# train the model
model_gensim.train(
sentences, total_examples=model_gensim.corpus_count, epochs=model_gensim.epochs
)
print(model_gensim)
# %% save model
model_gensim.save(dataPath + "fasttext.model")
# %% sanity check
model_gensim.most_similar("cattle")
model_gensim.most_similar(["super", "market"])
model_gensim.most_similar(["pharma"])
model_gensim.wv.doesnt_match("pharma medical pharmaceutic cattle".split())
model_gensim.wv.doesnt_match("pharma medical pharmaceutic".split())
from pprint import pprint as print
from gensim.models.fasttext import FastText as FT_gensim
from gensim.test.utils import datapath
import time
# Set file names for train and test data
corpus_file = datapath('lee_background.cor')
model = FT_gensim(size=100)
start = time.time()
# build the vocabulary
model.build_vocab(corpus_file=corpus_file)
# train the model
model.train(
corpus_file=corpus_file, epochs=model.epochs,
total_examples=model.corpus_count, total_words=model.corpus_total_words
)
print(model)
print("Model trained")
print(str(model.most_similar("rest",topn=5)) + "\nTook " + str(time.time()-start))
