# Copyright 2015 Google Inc. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================
from __future__ import absolute_import
from __future__ import print_function
import collections
import math
import os
import random
import zipfile
import numpy as np
from six.moves import xrange # pylint: disable=redefined-builtin
import matplotlib.pyplot as plt
import tensorflow as tf
# Read the data into a string.
[docs]def read_data(filename):
"""
:param filename: A zip file to open and read from
:return: A list of the space delimited tokens from the textfile.
"""
f = zipfile.ZipFile(filename)
for name in f.namelist():
return f.read(name).split()
f.close()
[docs]def build_dataset(words, vocabulary_size):
"""
:param words: A list of word tokens from a text file
:param vocabulary_size: How many word tokens to keep.
:return: data (text transformed into list of word ids 'UNK'=0), count (list of pairs (word:word_count) indexed by word id), dictionary (word:id hashmap), reverse_dictionary (id:word hashmap)
"""
count = [['UNK', -1]]
count.extend(collections.Counter(words).most_common(vocabulary_size - 1))
dictionary = dict()
for word, _ in count:
dictionary[word] = len(dictionary)
data = list()
unk_count = 0
for word in words:
if word in dictionary:
index = dictionary[word]
else:
index = 0 # dictionary['UNK']
unk_count += 1
data.append(index)
count[0][1] = unk_count
reverse_dictionary = dict(zip(dictionary.values(), dictionary.keys()))
return data, count, dictionary, reverse_dictionary
data_index = 0
# Step 3: Function to generate a training batch for the skip-gram model.
[docs]def generate_batch(data, batch_size, num_skips, skip_window):
"""
:param data: list of word ids corresponding to text
:param batch_size: Size of batch to retrieve
:param num_skips: How many times to reuse an input to generate a label.
:param skip_window: How many words to consider left and right.
:return:
"""
global data_index
assert batch_size % num_skips == 0
assert num_skips <= 2 * skip_window
batch = np.ndarray(shape=(batch_size), dtype=np.int32)
labels = np.ndarray(shape=(batch_size, 1), dtype=np.int32)
span = 2 * skip_window + 1 # [ skip_window target skip_window ]
buffer = collections.deque(maxlen=span)
for _ in range(span):
buffer.append(data[data_index])
data_index = (data_index + 1) % len(data)
for i in range(batch_size // num_skips):
target = skip_window # target label at the center of the buffer
targets_to_avoid = [skip_window]
for j in range(num_skips):
while target in targets_to_avoid:
target = random.randint(0, span - 1)
targets_to_avoid.append(target)
batch[i * num_skips + j] = buffer[skip_window]
labels[i * num_skips + j, 0] = buffer[target]
buffer.append(data[data_index])
data_index = (data_index + 1) % len(data)
return batch, labels
# Step 4: Build and train a skip-gram model.
[docs]class SkipGramVecs(object):
"""
Trains a skip gram model from `Distributed Representations of Words and Phrases and their Compositionality`_
:param textfile: Plain text file or zip file with plain text files.
:param vocabulary_size: How many words to use from text
:param batch_size: mini-batch size
:param embedding_size: Dimension of the embedding vector.
:param skip_window: How many words to consider left and right.
:param num_skips: How many times to reuse an input to generate a label.
:param valid_size: Random set of words to evaluate similarity on.
:param valid_window: Only pick dev samples in the head of the distribution.
:param num_sampled: Number of negative examples to sample.
:param num_steps: How many mini-batch steps to take
:param verbose: Whether to calculate and print similarities for a sample of words
"""
def __init__(self, textfile, vocabulary_size=12735,
batch_size=128, embedding_size=128, skip_window=1, num_skips=2,
valid_size=16, valid_window=100, num_sampled=64, num_steps=100000,
verbose=False):
if not textfile.endswith('.zip'):
ziptext = textfile + '.zip'
os.system('zip ' + ziptext + ' ' + textfile)
textfile = ziptext
words = read_data(textfile)
print('Data size', len(words))
self.data, self.count, self.dictionary, self.reverse_dictionary = build_dataset(words, vocabulary_size)
del words # Hint to reduce memory.
batch, labels = generate_batch(self.data, batch_size=8, num_skips=2, skip_window=1)
for i in range(8):
print(batch[i], '->', labels[i, 0])
print(self.reverse_dictionary[batch[i]], '->', self.reverse_dictionary[labels[i, 0]])
# We pick a random validation set to sample nearest neighbors. Here we limit the
# validation samples to the words that have a low numeric ID, which by
# construction are also the most frequent.
valid_examples = np.array(random.sample(np.arange(valid_window), valid_size))
graph = tf.Graph()
with graph.as_default():
# Input data.
train_inputs = tf.placeholder(tf.int32, shape=[batch_size])
train_labels = tf.placeholder(tf.int32, shape=[batch_size, 1])
valid_dataset = tf.constant(valid_examples, dtype=tf.int32)
# Ops and variables pinned to the CPU because of missing GPU implementation
with tf.device('/cpu:0'):
# Look up embeddings for inputs.
embeddings = tf.Variable(
tf.random_uniform([vocabulary_size, embedding_size], -1.0, 1.0))
embed = tf.nn.embedding_lookup(embeddings, train_inputs)
# Construct the variables for the NCE loss
nce_weights = tf.Variable(
tf.truncated_normal([vocabulary_size, embedding_size],
stddev=1.0 / math.sqrt(embedding_size)))
nce_biases = tf.Variable(tf.zeros([vocabulary_size]))
# Compute the average NCE loss for the batch.
# tf.nce_loss automatically draws a new sample of the negative labels each
# time we evaluate the loss.
loss = tf.reduce_mean(
tf.nn.nce_loss(nce_weights, nce_biases, embed, train_labels,
num_sampled, vocabulary_size))
optimizer = tf.train.GradientDescentOptimizer(1.0).minimize(loss)
# Compute the cosine similarity between minibatch examples and all embeddings.
norm = tf.sqrt(tf.reduce_sum(tf.square(embeddings), 1, keep_dims=True))
normalized_embeddings = embeddings / norm
valid_embeddings = tf.nn.embedding_lookup(
normalized_embeddings, valid_dataset)
similarity = tf.matmul(
valid_embeddings, normalized_embeddings, transpose_b=True)
with tf.Session(graph=graph) as session:
# We must initialize all variables before we use them.
tf.initialize_all_variables().run()
average_loss = 0
for step in xrange(num_steps):
batch_inputs, batch_labels = generate_batch(
self.data, batch_size, num_skips, skip_window)
feed_dict = {train_inputs: batch_inputs, train_labels: batch_labels}
_, loss_val = session.run([optimizer, loss], feed_dict=feed_dict)
average_loss += loss_val
if step % 2000 == 0:
if step > 0:
average_loss /= 2000
# The average loss is an estimate of the loss over the last 2000 batches.
print("Average loss at step ", step, ": ", average_loss)
average_loss = 0
# Note that this is expensive (~20% slowdown if computed every 500 steps)
if verbose and step % 10000 == 0:
sim = similarity.eval()
for i in xrange(valid_size):
valid_word = self.reverse_dictionary[valid_examples[i]]
top_k = 8 # number of nearest neighbors
nearest = (-sim[i, :]).argsort()[1:top_k + 1]
log_str = "Nearest to %s:" % valid_word
for k in xrange(top_k):
close_word = self.reverse_dictionary[nearest[k]]
log_str = "%s %s," % (log_str, close_word)
print(log_str)
self.final_embeddings = normalized_embeddings.eval()
[docs] def plot_embeddings(self, filename='tsne.png', num_terms=500):
"""
Plot tsne reduction of learned word embeddings in 2-space.
:param filename: File to save plot to.
:param num_terms: How many words to plot.
"""
plot_tsne(self.final_embeddings,
[self.reverse_dictionary[i] for i in xrange(num_terms)],
filename, num_terms)
[docs]def plot_tsne(embeddings, labels, filename='tsne.png', num_terms=500):
"""
Makes tsne plot to visualize word embeddings. Need sklearn, matplotlib for this to work.
:param filename: Location to save labeled tsne plots
:param num_terms: Num of words to plot
"""
try:
from sklearn.manifold import TSNE
tsne = TSNE(perplexity=30, n_components=2, init='pca', n_iter=5000)
low_dim_embs = tsne.fit_transform(embeddings[:num_terms, :])
assert low_dim_embs.shape[0] >= len(labels), "More labels than embeddings"
plt.figure(figsize=(18, 18)) # in inches
for i, label in enumerate(labels):
x, y = low_dim_embs[i, :]
plt.scatter(x, y)
plt.annotate(label,
xy=(x, y),
xytext=(5, 2),
textcoords='offset points',
ha='right',
va='bottom')
plt.savefig(filename)
except ImportError:
print("Please install sklearn and matplotlib to visualize embeddings.")