import collections
import math
import torch
from torch import nn
from d2l import torch as d2l
from encoder_decoder import Encoder
from encoder_decoder import Decoder
from encoder_decoder import EncoderDecoder
from machine_translation import load_data_nmt
from RNN.rnn_self import grad_clipping
class Seq2SeqEncoder(Encoder):
"""用于序列到序列学习的循环神经网络编码器"""
def __init__(self, vocab_size, embed_size, num_hidden, num_layers, dropout=0, **kwargs):
super(Seq2SeqEncoder, self).__init__(**kwargs)
self.embedding = nn.Embedding(vocab_size, embed_size)
self.rnn = nn.GRU(embed_size, num_hidden, num_layers, dropout=dropout)
def forward(self, X, *args):
X = self.embedding(X)
X = X.permute(1, 0, 2)
output, state = self.rnn(X)
return output, state
class Seq2SeqDecoder(Decoder):
"""用于序列到序列学习的循环神经网络解码器"""
def __init__(self, vocab_size, embed_size, num_hidden, num_layer, dropout=0, **kwargs):
super(Seq2SeqDecoder, self).__init__(**kwargs)
self.embedding = nn.Embedding(vocab_size, embed_size)
self.rnn = nn.GRU(embed_size + num_hidden, num_hidden, num_layer, dropout=dropout)
self.dense = nn.Linear(num_hidden, vocab_size)
def init_state(self, enc_outputs, *args):
"""
:param enc_outputs: 编码器的输出
:param args: 其余参数
:return:
"""
return enc_outputs[1]
def forward(self, X, state):
X = self.embedding(X).permute(1, 0, 2)
context = state[-1].repeat(X.shape[0], 1, 1)
X_and_context = torch.cat((X, context), 2)
output, state = self.rnn(X_and_context, state)
output = self.dense(output).permute(1, 0, 2)
return output, state
def sequence_mask(X, valid_len, value=0):
"""屏蔽序列中不相关的项"""
max_len = X.size(1)
mask = torch.arange((max_len), dtype=torch.float32, device=X.device)[None, :] < valid_len[:, None]
X[~mask] = value
return X
class MaskedSoftmaxCELoss(nn.CrossEntropyLoss):
"""带遮蔽的softmax交叉熵损失函数"""
def forward(self, pred, label, valid_len):
weights = torch.ones_like(label)
weights = sequence_mask(weights, valid_len)
self.reduction = 'none'
unweighted_loss = super(MaskedSoftmaxCELoss, self).forward(pred.permute(0, 2, 1), label)
weighted_loss = (unweighted_loss * weights).mean(dim=1)
return weighted_loss
def train_seq2seq(net, train_iter, lr, num_epochs, target_vocab, device):
"""训练序列到序列模型"""
def xavier_init_weights(m):
if type(m) == nn.Linear:
nn.init.xavier_uniform_(m.weight)
if type(m) == nn.GRU:
for param in m._flat_weights_names:
if "weight" in param:
nn.init.xavier_uniform_(m._parameters[param])
net.apply(xavier_init_weights)
net.to(device)
optimizer = torch.optim.Adam(net.parameters(), lr=lr)
loas = MaskedSoftmaxCELoss()
net.train()
animator = d2l.Animator(xlabel='epoch', ylabel='loss', xlim=[10, num_epochs])
for epoch in range(num_epochs):
timer = d2l.Timer()
metric = d2l.Accumulator(2)
for batch in train_iter:
optimizer.zero_grad()
X, X_valid_len, Y, Y_valid_len = [x.to(device) for x in batch]
bos = torch.tensor([target_vocab['<bos>']] * Y.shape[0], device=device).reshape(-1, 1)
dec_input = torch.cat([bos, Y[:, :-1]], 1)
Y_hat, _ = net(X, dec_input, X_valid_len)
l = loss(Y_hat, Y, Y_valid_len)
l.sum().backward()
grad_clipping(net, l.sum())
num_tokens = Y_valid_len.sum()
optimizer.step()
with torch.no_grad():
metric.add(l.sum(), num_tokens)
if (epoch + 1) % 10 == 0:
animator.add(epoch + 1, (metric[0] / metric[1], ))
print(f'loss {metric[0] / metric[1]}, {metric[1] / timer.stop()} tokens / sec on {device}')
d2l.plt.show()
def predict_seq2seq(net, src_sentence, src_vocab, target_vocab, num_steps, device, save_attention_weights=False):
"""序列到序列模型的预测"""
net.eval()
src_tokens = src_vocab[src_sentence.lower().split(' ')] + [src_vocab['<eos>']]
end_valid_len = torch.tensor([len(src_tokens)], device=device)
src_tokens = d2l.truncate_pad(src_tokens, num_steps, src_vocab['<pad>'])
enc_X = torch.unsqueeze(
torch.tensor(src_tokens, dtype=torch.long, device=device), dim=0
)
enc_outputs = net.encoder(enc_X, end_valid_len)
dec_state = net.decoder.init_state(enc_outputs, end_valid_len)
dec_X = torch.unsqueeze(
torch.tensor([target_vocab['<bos>']], dtype=torch.long, device=device), dim=0
)
output_seq, attention_weight_seq = [], []
for _ in range(num_steps):
Y, dec_state = net.decoder(dec_X, dec_state)
dec_X = Y.argmax(dim=2)
pred = dec_X.squeeze(dim=0).type(torch.int32).item()
if save_attention_weights:
attention_weight_seq.append(net.decoder.attention_weights)
if pred == target_vocab['<eos>']:
break
output_seq.append(pred)
return ' '.join(target_vocab.to_tokens(output_seq)), attention_weight_seq
def bleu(pred_seq, label_seq, k):
"""计算BLEU"""
pred_tokens, label_tokens = pred_seq.split(' '), label_seq.split(' ')
len_pred, len_label = len(pred_tokens), len(label_tokens)
score = math.exp(min(0, 1 - len_label / len_pred))
for n in range(1, k + 1):
num_matches, label_subs = 0, collections.defaultdict(int)
for i in range(len_label - n + 1):
label_subs[' '.join(label_tokens[i: i + n])] += 1
for i in range(len_pred - n + 1):
if label_subs[' '.join(pred_tokens[i: i + n])] > 0:
num_matches += 1
label_subs[' '.join(pred_tokens[i: i + n])] += 1
score *= math.pow(num_matches / (len_pred - n + 1), math.pow(0.5, n))
return score
if __name__ == "__main__":
encoder = Seq2SeqEncoder(vocab_size=10, embed_size=8, num_hidden=16, num_layers=2)
encoder.eval()
X = torch.zeros((4, 7), dtype=torch.long)
output, state = encoder(X)
print("encoder:")
print('output shape:', output.shape, 'state shape:', state.shape)
decoder = Seq2SeqDecoder(vocab_size=10, embed_size=8, num_hidden=16, num_layer=2)
decoder.eval()
state = decoder.init_state(encoder(X))
output, state = decoder(X, state)
print("decoder:")
print('output shape:', output.shape, 'state shape:', state.shape)
loss = MaskedSoftmaxCELoss()
print("loss demo:")
print(loss(torch.ones(3, 4, 10), torch.ones((3, 4), dtype=torch.long), torch.tensor([4, 2, 0])))
embed_size , num_hidden, num_layers, dropout = 32, 32, 2, 0.1
batch_size, num_steps = 64, 10
lr, num_epochs, device = 0.005, 300, d2l.try_gpu()
train_iter, src_vocab, target_vocab = load_data_nmt(batch_size, num_steps)
encoder = Seq2SeqEncoder(len(src_vocab), embed_size, num_hidden, num_layers, dropout)
decoder = Seq2SeqDecoder(len(target_vocab), embed_size, num_hidden, num_layers, dropout)
net = EncoderDecoder(encoder, decoder)
train_seq2seq(net, train_iter, lr, num_epochs, target_vocab, device)
print("start translating:")
engs = ['go .', 'i lost .', 'he\'s calm .', 'i\'m home .']
fras = ['va !', 'j\'ai perdu .', 'il est calme .', 'je suis chez moi .']
for eng, fra in zip(engs, fras):
translation, attention_weight_seq = predict_seq2seq(net, eng, src_vocab, target_vocab, num_steps, device)
print(f'{eng} => {translation}, bleu {bleu(translation, fra, k=2):.3f}')
|
微信支付