.. _sec_sentiment_rnn:
Phân tích tình cảm: Sử dụng mạng nơ-ron tái phát
================================================
Giống như sự tương đồng từ và các nhiệm vụ tương tự, chúng ta cũng có
thể áp dụng các vectơ từ được đào tạo trước vào phân tích tình cảm. Vì
tập dữ liệu đánh giá IMDb trong :numref:`sec_sentiment` không lớn lắm,
sử dụng các đại diện văn bản đã được đào tạo trước trên corpora quy mô
lớn có thể làm giảm quá mức của mô hình. Như một ví dụ cụ thể được minh
họa trong :numref:`fig_nlp-map-sa-rnn`, chúng tôi sẽ đại diện cho mỗi
mã thông báo bằng cách sử dụng mô hình Glove được đào tạo trước và đưa
các biểu diễn token này thành RNN hai chiều nhiều lớp để có được biểu
diễn chuỗi văn bản, sẽ được chuyển thành đầu ra phân tích tình cảm
:cite:`Maas.Daly.Pham.ea.2011`. Đối với cùng một ứng dụng hạ lưu,
chúng tôi sẽ xem xét một sự lựa chọn kiến trúc khác nhau sau này.
.. _fig_nlp-map-sa-rnn:
.. figure:: ../img/nlp-map-sa-rnn.svg
This section feeds pretrained GloVe to an RNN-based architecture for
sentiment analysis.
.. raw:: html
.. raw:: html
.. code:: python
from mxnet import gluon, init, np, npx
from mxnet.gluon import nn, rnn
from d2l import mxnet as d2l
npx.set_np()
batch_size = 64
train_iter, test_iter, vocab = d2l.load_data_imdb(batch_size)
.. raw:: html
.. raw:: html
.. code:: python
import torch
from torch import nn
from d2l import torch as d2l
batch_size = 64
train_iter, test_iter, vocab = d2l.load_data_imdb(batch_size)
.. parsed-literal::
:class: output
Downloading ../data/aclImdb_v1.tar.gz from http://ai.stanford.edu/~amaas/data/sentiment/aclImdb_v1.tar.gz...
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
class BiRNN(nn.Block):
def __init__(self, vocab_size, embed_size, num_hiddens,
num_layers, **kwargs):
super(BiRNN, self).__init__(**kwargs)
self.embedding = nn.Embedding(vocab_size, embed_size)
# Set `bidirectional` to True to get a bidirectional RNN
self.encoder = rnn.LSTM(num_hiddens, num_layers=num_layers,
bidirectional=True, input_size=embed_size)
self.decoder = nn.Dense(2)
def forward(self, inputs):
# The shape of `inputs` is (batch size, no. of time steps). Because
# LSTM requires its input's first dimension to be the temporal
# dimension, the input is transposed before obtaining token
# representations. The output shape is (no. of time steps, batch size,
# word vector dimension)
embeddings = self.embedding(inputs.T)
# Returns hidden states of the last hidden layer at different time
# steps. The shape of `outputs` is (no. of time steps, batch size,
# 2 * no. of hidden units)
outputs = self.encoder(embeddings)
# Concatenate the hidden states at the initial and final time steps as
# the input of the fully-connected layer. Its shape is (batch size,
# 4 * no. of hidden units)
encoding = np.concatenate((outputs[0], outputs[-1]), axis=1)
outs = self.decoder(encoding)
return outs
.. raw:: html
.. raw:: html
.. code:: python
class BiRNN(nn.Module):
def __init__(self, vocab_size, embed_size, num_hiddens,
num_layers, **kwargs):
super(BiRNN, self).__init__(**kwargs)
self.embedding = nn.Embedding(vocab_size, embed_size)
# Set `bidirectional` to True to get a bidirectional RNN
self.encoder = nn.LSTM(embed_size, num_hiddens, num_layers=num_layers,
bidirectional=True)
self.decoder = nn.Linear(4 * num_hiddens, 2)
def forward(self, inputs):
# The shape of `inputs` is (batch size, no. of time steps). Because
# LSTM requires its input's first dimension to be the temporal
# dimension, the input is transposed before obtaining token
# representations. The output shape is (no. of time steps, batch size,
# word vector dimension)
embeddings = self.embedding(inputs.T)
self.encoder.flatten_parameters()
# Returns hidden states of the last hidden layer at different time
# steps. The shape of `outputs` is (no. of time steps, batch size,
# 2 * no. of hidden units)
outputs, _ = self.encoder(embeddings)
# Concatenate the hidden states of the initial time step and final
# time step to use as the input of the fully connected layer. Its
# shape is (batch size, 4 * no. of hidden units)
encoding = torch.cat((outputs[0], outputs[-1]), dim=1)
# Concatenate the hidden states at the initial and final time steps as
# the input of the fully-connected layer. Its shape is (batch size,
# 4 * no. of hidden units)
outs = self.decoder(encoding)
return outs
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
embed_size, num_hiddens, num_layers, devices = 100, 100, 2, d2l.try_all_gpus()
net = BiRNN(len(vocab), embed_size, num_hiddens, num_layers)
net.initialize(init.Xavier(), ctx=devices)
.. raw:: html
.. raw:: html
.. code:: python
embed_size, num_hiddens, num_layers, devices = 100, 100, 2, d2l.try_all_gpus()
net = BiRNN(len(vocab), embed_size, num_hiddens, num_layers)
def init_weights(m):
if type(m) == nn.Linear:
nn.init.xavier_uniform_(m.weight)
if type(m) == nn.LSTM:
for param in m._flat_weights_names:
if "weight" in param:
nn.init.xavier_uniform_(m._parameters[param])
net.apply(init_weights);
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
glove_embedding = d2l.TokenEmbedding('glove.6b.100d')
.. raw:: html
.. raw:: html
.. code:: python
glove_embedding = d2l.TokenEmbedding('glove.6b.100d')
.. parsed-literal::
:class: output
Downloading ../data/glove.6B.100d.zip from http://d2l-data.s3-accelerate.amazonaws.com/glove.6B.100d.zip...
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
embeds = glove_embedding[vocab.idx_to_token]
embeds.shape
.. parsed-literal::
:class: output
(49346, 100)
.. raw:: html
.. raw:: html
.. code:: python
embeds = glove_embedding[vocab.idx_to_token]
embeds.shape
.. parsed-literal::
:class: output
torch.Size([49346, 100])
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
net.embedding.weight.set_data(embeds)
net.embedding.collect_params().setattr('grad_req', 'null')
.. raw:: html
.. raw:: html
.. code:: python
net.embedding.weight.data.copy_(embeds)
net.embedding.weight.requires_grad = False
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
lr, num_epochs = 0.01, 5
trainer = gluon.Trainer(net.collect_params(), 'adam', {'learning_rate': lr})
loss = gluon.loss.SoftmaxCrossEntropyLoss()
d2l.train_ch13(net, train_iter, test_iter, loss, trainer, num_epochs, devices)
.. parsed-literal::
:class: output
loss 0.297, train acc 0.872, test acc 0.846
707.2 examples/sec on [gpu(0), gpu(1)]
.. figure:: output_sentiment-analysis-rnn_6199ad_57_1.svg
.. raw:: html
.. raw:: html
.. code:: python
lr, num_epochs = 0.01, 5
trainer = torch.optim.Adam(net.parameters(), lr=lr)
loss = nn.CrossEntropyLoss(reduction="none")
d2l.train_ch13(net, train_iter, test_iter, loss, trainer, num_epochs, devices)
.. parsed-literal::
:class: output
loss 0.304, train acc 0.875, test acc 0.849
709.9 examples/sec on [device(type='cuda', index=0), device(type='cuda', index=1)]
.. figure:: output_sentiment-analysis-rnn_6199ad_60_1.svg
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
#@save
def predict_sentiment(net, vocab, sequence):
"""Predict the sentiment of a text sequence."""
sequence = np.array(vocab[sequence.split()], ctx=d2l.try_gpu())
label = np.argmax(net(sequence.reshape(1, -1)), axis=1)
return 'positive' if label == 1 else 'negative'
.. raw:: html
.. raw:: html
.. code:: python
#@save
def predict_sentiment(net, vocab, sequence):
"""Predict the sentiment of a text sequence."""
sequence = torch.tensor(vocab[sequence.split()], device=d2l.try_gpu())
label = torch.argmax(net(sequence.reshape(1, -1)), dim=1)
return 'positive' if label == 1 else 'negative'
.. raw:: html
.. raw:: html
.. raw:: html
.. code:: python
predict_sentiment(net, vocab, 'this movie is so great')
.. parsed-literal::
:class: output
'positive'
.. code:: python
predict_sentiment(net, vocab, 'this movie is so bad')
.. parsed-literal::
:class: output
'negative'
.. raw:: html
.. raw:: html
.. code:: python
predict_sentiment(net, vocab, 'this movie is so great')
.. parsed-literal::
:class: output
'positive'
.. code:: python
predict_sentiment(net, vocab, 'this movie is so bad')
.. parsed-literal::
:class: output
'negative'
.. raw:: html
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html
`Discussions `__
.. raw:: html
.. raw:: html