電子發(fā)燒友網(wǎng)>電子資料下載>電子資料>PyTorch教程15.7之詞的相似性和類比

PyTorch教程15.7之詞的相似性和類比

2074048 2023-06-05 | pdf | 0.10 MB | 次下載 | 免費

資料介紹

在15.4 節(jié)中，我們在一個小數(shù)據(jù)集上訓(xùn)練了一個 word2vec 模型，并將其應(yīng)用于為輸入詞尋找語義相似的詞。在實踐中，在大型語料庫上預(yù)訓(xùn)練的詞向量可以應(yīng)用于下游的自然語言處理任務(wù)，這將在第 16 節(jié)后面介紹。為了以直接的方式展示來自大型語料庫的預(yù)訓(xùn)練詞向量的語義，讓我們將它們應(yīng)用到詞相似度和類比任務(wù)中。

						import os
import torch
from torch import nn
from d2l import torch as d2l

						 

						import os
from mxnet import np, npx
from d2l import mxnet as d2l

npx.set_np()

15.7.1。加載預(yù)訓(xùn)練詞向量

下面列出了維度為 50、100 和 300 的預(yù)訓(xùn)練 GloVe 嵌入，可以從GloVe 網(wǎng)站下載。預(yù)訓(xùn)練的 fastText 嵌入有多種語言版本。這里我們考慮一個可以從fastText 網(wǎng)站下載的英文版本（300 維“wiki.en”）。

							#@save
d2l.DATA_HUB['glove.6b.50d'] = (d2l.DATA_URL + 'glove.6B.50d.zip',
                '0b8703943ccdb6eb788e6f091b8946e82231bc4d')

#@save
d2l.DATA_HUB['glove.6b.100d'] = (d2l.DATA_URL + 'glove.6B.100d.zip',
                 'cd43bfb07e44e6f27cbcc7bc9ae3d80284fdaf5a')

#@save
d2l.DATA_HUB['glove.42b.300d'] = (d2l.DATA_URL + 'glove.42B.300d.zip',
                 'b5116e234e9eb9076672cfeabf5469f3eec904fa')

#@save
d2l.DATA_HUB['wiki.en'] = (d2l.DATA_URL + 'wiki.en.zip',
              'c1816da3821ae9f43899be655002f6c723e91b88')

							 

							#@save
d2l.DATA_HUB['glove.6b.50d'] = (d2l.DATA_URL + 'glove.6B.50d.zip',
                '0b8703943ccdb6eb788e6f091b8946e82231bc4d')

#@save
d2l.DATA_HUB['glove.6b.100d'] = (d2l.DATA_URL + 'glove.6B.100d.zip',
                 'cd43bfb07e44e6f27cbcc7bc9ae3d80284fdaf5a')

#@save
d2l.DATA_HUB['glove.42b.300d'] = (d2l.DATA_URL + 'glove.42B.300d.zip',
                 'b5116e234e9eb9076672cfeabf5469f3eec904fa')

#@save
d2l.DATA_HUB['wiki.en'] = (d2l.DATA_URL + 'wiki.en.zip',
              'c1816da3821ae9f43899be655002f6c723e91b88')

							 

為了加載這些預(yù)訓(xùn)練的 GloVe 和 fastText 嵌入，我們定義了以下TokenEmbedding類。

							#@save
class TokenEmbedding:
  """Token Embedding."""
  def __init__(self, embedding_name):
    self.idx_to_token, self.idx_to_vec = self._load_embedding(
      embedding_name)
    self.unknown_idx = 0
    self.token_to_idx = {token: idx for idx, token in
               enumerate(self.idx_to_token)}

  def _load_embedding(self, embedding_name):
    idx_to_token, idx_to_vec = [''], []
    data_dir = d2l.download_extract(embedding_name)
    # GloVe website: https://nlp.stanford.edu/projects/glove/
    # fastText website: https://fasttext.cc/
    with open(os.path.join(data_dir, 'vec.txt'), 'r') as f:
      for line in f:
        elems = line.rstrip().split(' ')
        token, elems = elems[0], [float(elem) for elem in elems[1:]]
        # Skip header information, such as the top row in fastText
        if len(elems) > 1:
          idx_to_token.append(token)
          idx_to_vec.append(elems)
    idx_to_vec = [[0] * len(idx_to_vec[0])] + idx_to_vec
    return idx_to_token, torch.tensor(idx_to_vec)

  def __getitem__(self, tokens):
    indices = [self.token_to_idx.get(token, self.unknown_idx)
          for token in tokens]
    vecs = self.idx_to_vec[torch.tensor(indices)]
    return vecs

  def __len__(self):
    return len(self.idx_to_token)

							 

							#@save
class TokenEmbedding:
  """Token Embedding."""
  def __init__(self, embedding_name):
    self.idx_to_token, self.idx_to_vec = self._load_embedding(
      embedding_name)
    self.unknown_idx = 0
    self.token_to_idx = {token: idx for idx, token in
               enumerate(self.idx_to_token)}

  def _load_embedding(self, embedding_name):
    idx_to_token, idx_to_vec = [''], []
    data_dir = d2l.download_extract(embedding_name)
    # GloVe website: https://nlp.stanford.edu/projects/glove/
    # fastText website: https://fasttext.cc/
    with open(os.path.join(data_dir, 'vec.txt'), 'r') as f:
      for line in f:
        elems = line.rstrip().split(' ')
        token, elems = elems[0], [float(elem) for elem in elems[1:]]
        # Skip header information, such as the top row in fastText
        if len(elems) > 1:
          idx_to_token.append(token)
          idx_to_vec.append(elems)
    idx_to_vec = [[0] * len(idx_to_vec[0])] + idx_to_vec
    return idx_to_token, np.array(idx_to_vec)

  def __getitem__(self, tokens):
    indices = [self.token_to_idx.get(token, self.unknown_idx)
          for token in tokens]
    vecs = self.idx_to_vec[np.array(indices)]
    return vecs

  def __len__(self):
    return len(self.idx_to_token)

							 

下面我們加載 50 維 GloVe 嵌入（在維基百科子集上預(yù)訓(xùn)練）。創(chuàng)建TokenEmbedding實例時，如果尚未下載指定的嵌入文件，則必須下載。

							glove_6b50d = TokenEmbedding('glove.6b.50d')

							 

							Downloading ../data/glove.6B.50d.zip from http://d2l-data.s3-accelerate.amazonaws.com/glove.6B.50d.zip...

						

							glove_6b50d = TokenEmbedding('glove.6b.50d')

							 

							Downloading ../data/glove.6B.50d.zip from http://d2l-data.s3-accelerate.amazonaws.com/glove.6B.50d.zip...

						

輸出詞匯量。詞匯表包含 400000 個單詞（標記）和一個特殊的未知標記。

							len(glove_6b50d)

							 

							len(glove_6b50d)

							 

我們可以獲得一個詞在詞匯表中的索引，反之亦然。

							glove_6b50d.token_to_idx['beautiful'], glove_6b50d.idx_to_token[3367]

							 

							(3367, 'beautiful')

						

							glove_6b50d.token_to_idx['beautiful'], glove_6b50d.idx_to_token[3367]

							 

							(3367, 'beautiful')

						

15.7.2。應(yīng)用預(yù)訓(xùn)練詞向量

使用加載的 GloVe 向量，我們將通過將它們應(yīng)用于以下單詞相似性和類比任務(wù)來演示它們的語義。

15.7.2.1。詞相似度

與第 15.4.3 節(jié)類似，為了根據(jù)詞向量之間的余弦相似度為輸入詞找到語義相似的詞，我們實現(xiàn)以下knn （k-最近的鄰居）功能。

								def knn(W, x, k):
  # Add 1e-9 for numerical stability
  cos = torch.mv(W, x.reshape(-1,)) / (
    torch.sqrt(torch.sum(W * W, axis=1) + 1e-9) *
    torch.sqrt((x * x).sum()))
  _, topk = torch.topk(cos, k=k)
  return topk, [cos[int(i)] for i in topk]

								 

								def knn(W, x, k):
  # Add 1e-9 for numerical stability
  cos = np.dot(W, x.reshape(-1,)) / (
    np.sqrt(np.sum(W * W, axis=1) + 1e-9) * np
							