文本主要介绍一下如何使用 PyTorch 复现 Transformer,实现简单的机器翻译任务。请先花上 15 分钟阅读我的这篇文章 Transformer详解,再来看文本,方能达到醍醐灌顶,事半功倍的效果
数据预处理
这里我并没有用什么大型的数据集,而是手动输入了两对德语→英语的句子,还有每个字的索引也是我手动硬编码上去的,主要是为了降低代码阅读难度,我希望读者能更关注模型实现的部分
# S: Symbol that shows starting of decoding input
# E: Symbol that shows starting of decoding output
# P: Symbol that will fill in blank sequence if current batch data size is short than time steps
sentences = [
# enc_input dec_input dec_output
['ich mochte ein bier P', 'S i want a beer .', 'i want a beer . E'],
['ich mochte ein cola P', 'S i want a coke .', 'i want a coke . E']
]
# Padding Should be Zero
src_vocab = {'P' : 0, 'ich' : 1, 'mochte' : 2, 'ein' : 3, 'bier' : 4, 'cola' : 5}
src_vocab_size = len(src_vocab)
tgt_vocab = {'P' : 0, 'i' : 1, 'want' : 2, 'a' : 3, 'beer' : 4, 'coke' : 5, 'S' : 6, 'E' : 7, '.' : 8}
idx2word = {i: w for i, w in enumerate(tgt_vocab)}
tgt_vocab_size = len(tgt_vocab)
src_len = 5 # enc_input max sequence length
tgt_len = 6 # dec_input(=dec_output) max sequence length
def make_data(sentences):
enc_inputs, dec_inputs, dec_outputs = [], [], []
for i in range(len(sentences)):
enc_input = [[src_vocab[n] for n in sentences[i][0].split()]] # [[1, 2, 3, 4, 0], [1, 2, 3, 5, 0]]
dec_input = [[tgt_vocab[n] for n in sentences[i][1].split()]] # [[6, 1, 2, 3, 4, 8], [6, 1, 2, 3, 5, 8]]
dec_output = [[tgt_vocab[n] for n in sentences[i][2].split()]] # [[1, 2, 3, 4, 8, 7], [1, 2, 3, 5, 8, 7]]
enc_inputs.extend(enc_input)
dec_inputs.extend(dec_input)
dec_outputs.extend(dec_output)
return torch.LongTensor(enc_inputs), torch.LongTensor(dec_inputs), torch.LongTensor(dec_outputs)
enc_inputs, dec_inputs, dec_outputs = make_data(sentences)
class MyDataSet(Data.Dataset):
def __init__(self, enc_inputs, dec_inputs, dec_outputs):
super(MyDataSet, self).__init__()
self.enc_inputs = enc_inputs
self.dec_inputs = dec_inputs
self.dec_outputs = dec_outputs
def __len__(self):
return self.enc_inputs.shape[0]
def __getitem__(self, idx):
return self.enc_inputs[idx], self.dec_inputs[idx], self.dec_outputs[idx]
loader = Data.DataLoader(MyDataSet(enc_inputs, dec_inputs, dec_outputs), 2, True)
模型参数
下面变量代表的含义依次是
- 字嵌入&位置嵌入的维度,这俩值是相同的,因此用一个变量就行了
- FeedForward层隐藏神经元个数
- Q、K、V向量的维度,其中Q与K的维度必须相等,V的维度没有限制,不过为了方便起见,我都设为64
- Encoder和Decoder的个数
- 多头注意力中head的数量
# Transformer Parameters
d_model = 512 # Embedding Size
d_ff = 2048 # FeedForward dimension
d_k = d_v = 64 # dimension of K(=Q), V
n_layers = 6 # number of Encoder of Decoder Layer
n_heads = 8 # number of heads in Multi-Head Attention
上面都比较简单,下面开始涉及到模型就比较复杂了,因此我会将模型拆分成以下几个部分进行讲解
- Positional Encoding
- Pad Mask(针对句子不够长,加了pad,因此需要对pad进行mask)
- Subsequence Mask(Decoder input不能看到未来时刻单词信息,因此需要mask)
- ScaledDotProductAttention(计算context vector)
- Multi-Head Attention
- FeedForward Layer
- Encoder Layer
- Encoder
- Decoder Layer
- Decoder
- Transformer
关于代码中的注释,如果值为src_len
或者tgt_len
的,我一定会写清楚,但是有些函数或者类,Encoder和Decoder都有可能调用,因此就不能确定究竟是src_len
还是tgt_len
,对于不确定的,我会记作seq_len
Positional Encoding
def get_sinusoid_encoding_table(n_position, d_model):
def cal_angle(position, hid_idx):
return position / np.power(10000, 2 * (hid_idx // 2) / d_model)
def get_posi_angle_vec(position):
return [cal_angle(position, hid_j) for hid_j in range(d_model)]
sinusoid_table = np.array([get_posi_angle_vec(pos_i) for pos_i in range(n_position)])
sinusoid_table[:, 0::2] = np.sin(sinusoid_table[:, 0::2]) # dim 2i
sinusoid_table[:, 1::2] = np.cos(sinusoid_table[:, 1::2]) # dim 2i+1
return torch.FloatTensor(sinusoid_table)
这段代码并不复杂,传入的参数分别是字库的大小,以及位置编码的维度。最终返回的参数,维度也即[n_position, d_model],和字编码的大小一摸一样
Pad Mask
def get_attn_pad_mask(seq_q, seq_k):
'''
seq_q: [batch_size, seq_len]
seq_k: [batch_size, seq_len]
seq_len could be src_len or it could be tgt_len
seq_len in seq_q and seq_len in seq_k maybe not equal
'''
batch_size, len_q = seq_q.size()
batch_size, len_k = seq_k.size()
# eq(zero) is PAD token
pad_attn_mask = seq_k.data.eq(0).unsqueeze(1) # [batch_size, 1, len_k], False is masked
return pad_attn_mask.expand(batch_size, len_q, len_k) # [batch_size, len_q, len_k]
由于在Encoder和Decoder中都需要进行mask操作,因此就无法确定这个函数的参数中seq_len
的值,如果是在Encoder中调用的,seq_len
就等于src_len
;如果是在Decoder中调用的,seq_len
就有可能等于src_len
,也有可能等于tgt_len
(因为Decoder有两次mask)
这个函数最核心的一句代码是seq_k.data.eq(0)
,这句的作用是返回一个大小和seq_k
一样的tensor,只不过里面的值只有True和False。如果seq_k
某个位置的值等于0,那么对应位置就是True,否则即为False。举个例子,输入为seq_data = [1, 2, 3, 4, 0]
,seq_data.data.eq(0)
就会返回[False, False, False, False, True]
剩下的代码主要是扩展维度,强烈建议读者打印出来,看看最终返回的数据是什么样子
Subsequence Mask
def get_attn_subsequence_mask(seq):
'''
seq: [batch_size, tgt_len]
'''
attn_shape = [seq.size(0), seq.size(1), seq.size(1)]
subsequence_mask = np.triu(np.ones(attn_shape), k=1) # Upper triangular matrix
subsequence_mask = torch.from_numpy(subsequence_mask).byte()
return subsequence_mask
Subsequence Mask只有Decoder会用到,主要作用是屏蔽未来时刻单词的信息。首先通过np.ones()
生成一个全1的方阵,然后通过np.triu()
生成一个上三角矩阵,下图是np.triu()
用法
ScaledDotProductAttention
class ScaledDotProductAttention(nn.Module):
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
def forward(self, Q, K, V, attn_mask):
'''
Q: [batch_size, n_heads, len_q, d_k]
K: [batch_size, n_heads, len_k, d_k]
V: [batch_size, n_heads, len_v(=len_k), d_v]
attn_mask: [batch_size, n_heads, seq_len, seq_len]
'''
scores = torch.matmul(Q, K.transpose(-1, -2)) / np.sqrt(d_k) # scores : [batch_size, n_heads, len_q, len_k]
scores.masked_fill_(attn_mask, -1e9) # Fills elements of self tensor with value where mask is True.
attn = nn.Softmax(dim=-1)(scores)
context = torch.matmul(attn, V) # [batch_size, n_heads, len_q, d_v]
return context, attn
这里要做的是,通过Q
和K
计算出scores
,然后将scores
和V
相乘,得到每个单词的context vector
第一步是将Q
和K
的转置相乘没什么好说的,相乘之后得到的scores
还不能立刻进行softmax,需要和attn_mask
相加,把一些需要屏蔽的信息屏蔽掉,attn_mask
是一个仅由True和False组成的tensor,并且一定会保证attn_mask
和scores
的维度四个值相同(不然无法做对应位置相加)
mask完了之后,就可以对scores
进行softmax了。然后再与V
相乘,得到context
MultiHeadAttention
class MultiHeadAttention(nn.Module):
def __init__(self):
super(MultiHeadAttention, self).__init__()
self.W_Q = nn.Linear(d_model, d_k * n_heads, bias=False)
self.W_K = nn.Linear(d_model, d_k * n_heads, bias=False)
self.W_V = nn.Linear(d_model, d_v * n_heads, bias=False)
self.fc = nn.Linear(n_heads * d_v, d_model, bias=False)
def forward(self, input_Q, input_K, input_V, attn_mask):
'''
input_Q: [batch_size, len_q, d_model]
input_K: [batch_size, len_k, d_model]
input_V: [batch_size, len_v(=len_k), d_model]
attn_mask: [batch_size, seq_len, seq_len]
'''
residual, batch_size = input_Q, input_Q.size(0)
# (B, S, D) -proj-> (B, S, D_new) -split-> (B, S, H, W) -trans-> (B, H, S, W)
Q = self.W_Q(input_Q).view(batch_size, -1, n_heads, d_k).transpose(1,2) # Q: [batch_size, n_heads, len_q, d_k]
K = self.W_K(input_K).view(batch_size, -1, n_heads, d_k).transpose(1,2) # K: [batch_size, n_heads, len_k, d_k]
V = self.W_V(input_V).view(batch_size, -1, n_heads, d_v).transpose(1,2) # V: [batch_size, n_heads, len_v(=len_k), d_v]
attn_mask = attn_mask.unsqueeze(1).repeat(1, n_heads, 1, 1) # attn_mask : [batch_size, n_heads, seq_len, seq_len]
# context: [batch_size, n_heads, len_q, d_v], attn: [batch_size, n_heads, len_q, len_k]
context, attn = ScaledDotProductAttention()(Q, K, V, attn_mask)
context = context.transpose(1, 2).reshape(batch_size, -1, n_heads * d_v) # context: [batch_size, len_q, n_heads * d_v]
output = self.fc(context) # [batch_size, len_q, d_model]
return nn.LayerNorm(d_model)(output + residual), attn
完整代码中一定会有三处地方调用MultiHeadAttention()
,Encoder Layer调用一次,传入的input_Q
、input_K
、input_V
全部都是enc_inputs
;Decoder Layer中两次调用,第一次传入的全是dec_inputs
,第二次传入的分别是dec_outputs
,enc_outputs
,enc_outputs
FeedForward Layer
class PoswiseFeedForwardNet(nn.Module):
def __init__(self):
super(PoswiseFeedForwardNet, self).__init__()
self.fc = nn.Sequential(
nn.Linear(d_model, d_ff, bias=False),
nn.ReLU(),
nn.Linear(d_ff, d_model, bias=False)
)
def forward(self, inputs):
'''
inputs: [batch_size, seq_len, d_model]
'''
residual = inputs
output = self.fc(inputs)
return nn.LayerNorm(d_model)(output + residual) # [batch_size, seq_len, d_model]
这段代码非常简单,就是做两次线性变换,残差连接后再跟一个Layer Norm
Encoder Layer
class EncoderLayer(nn.Module):
def __init__(self):
super(EncoderLayer, self).__init__()
self.enc_self_attn = MultiHeadAttention()
self.pos_ffn = PoswiseFeedForwardNet()
def forward(self, enc_inputs, enc_self_attn_mask):
'''
enc_inputs: [batch_size, src_len, d_model]
enc_self_attn_mask: [batch_size, src_len, src_len]
'''
# enc_outputs: [batch_size, src_len, d_model], attn: [batch_size, n_heads, src_len, src_len]
enc_outputs, attn = self.enc_self_attn(enc_inputs, enc_inputs, enc_inputs, enc_self_attn_mask) # enc_inputs to same Q,K,V
enc_outputs = self.pos_ffn(enc_outputs) # enc_outputs: [batch_size, src_len, d_model]
return enc_outputs, attn
将上述组件拼起来,就是一个完整的Encoder Layer
Encoder
class Encoder(nn.Module):
def __init__(self):
super(Encoder, self).__init__()
self.src_emb = nn.Embedding(src_vocab_size, d_model)
self.pos_emb = nn.Embedding.from_pretrained(get_sinusoid_encoding_table(src_vocab_size, d_model),freeze=True)
self.layers = nn.ModuleList([EncoderLayer() for _ in range(n_layers)])
def forward(self, enc_inputs):
'''
enc_inputs: [batch_size, src_len]
'''
word_emb = self.src_emb(enc_inputs) # [batch_size, src_len, d_model]
pos_emb = self.pos_emb(enc_inputs) # [batch_size, src_len, d_model]
enc_outputs = word_emb + pos_emb
enc_self_attn_mask = get_attn_pad_mask(enc_inputs, enc_inputs) # [batch_size, src_len, src_len]
enc_self_attns = []
for layer in self.layers:
# enc_outputs: [batch_size, src_len, d_model], enc_self_attn: [batch_size, n_heads, src_len, src_len]
enc_outputs, enc_self_attn = layer(enc_outputs, enc_self_attn_mask)
enc_self_attns.append(enc_self_attn)
return enc_outputs, enc_self_attns
使用nn.ModuleList()
里面的参数是列表,列表里面存了n_layers
个Encoder Layer
由于我们控制好了Encoder Layer的输入和输出维度相同,所以可以直接用个for循环以嵌套的方式,将上一次Encoder Layer的输出作为下一次Encoder Layer的输入
Decoder Layer
class DecoderLayer(nn.Module):
def __init__(self):
super(DecoderLayer, self).__init__()
self.dec_self_attn = MultiHeadAttention()
self.dec_enc_attn = MultiHeadAttention()
self.pos_ffn = PoswiseFeedForwardNet()
def forward(self, dec_inputs, enc_outputs, dec_self_attn_mask, dec_enc_attn_mask):
'''
dec_inputs: [batch_size, tgt_len, d_model]
enc_outputs: [batch_size, src_len, d_model]
dec_self_attn_mask: [batch_size, tgt_len, tgt_len]
dec_enc_attn_mask: [batch_size, tgt_len, src_len]
'''
# dec_outputs: [batch_size, tgt_len, d_model], dec_self_attn: [batch_size, n_heads, tgt_len, tgt_len]
dec_outputs, dec_self_attn = self.dec_self_attn(dec_inputs, dec_inputs, dec_inputs, dec_self_attn_mask)
# dec_outputs: [batch_size, tgt_len, d_model], dec_enc_attn: [batch_size, h_heads, tgt_len, src_len]
dec_outputs, dec_enc_attn = self.dec_enc_attn(dec_outputs, enc_outputs, enc_outputs, dec_enc_attn_mask)
dec_outputs = self.pos_ffn(dec_outputs) # [batch_size, tgt_len, d_model]
return dec_outputs, dec_self_attn, dec_enc_attn
在Decoder Layer中会调用两次MultiHeadAttention
,第一次是计算Decoder Input的self-attention,得到输出dec_outputs
。然后将dec_outputs
作为生成Q的元素,enc_outputs
作为生成K和V的元素,再调用一次MultiHeadAttention
,得到的是Encoder和Decoder Layer之间的context vector。最后将dec_outptus
做一次维度变换,然后返回
Decoder
class Decoder(nn.Module):
def __init__(self):
super(Decoder, self).__init__()
self.tgt_emb = nn.Embedding(tgt_vocab_size, d_model)
self.pos_emb = nn.Embedding.from_pretrained(get_sinusoid_encoding_table(tgt_vocab_size, d_model),freeze=True)
self.layers = nn.ModuleList([DecoderLayer() for _ in range(n_layers)])
def forward(self, dec_inputs, enc_inputs, enc_outputs):
'''
dec_inputs: [batch_size, tgt_len]
enc_intpus: [batch_size, src_len]
enc_outputs: [batsh_size, src_len, d_model]
'''
word_emb = self.tgt_emb(dec_inputs) # [batch_size, tgt_len, d_model]
pos_emb = self.pos_emb(dec_inputs) # [batch_size, tgt_len, d_model]
dec_outputs = word_emb + pos_emb
dec_self_attn_pad_mask = get_attn_pad_mask(dec_inputs, dec_inputs) # [batch_size, tgt_len, tgt_len]
dec_self_attn_subsequent_mask = get_attn_subsequence_mask(dec_inputs) # [batch_size, tgt_len]
dec_self_attn_mask = torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequent_mask), 0) # [batch_size, tgt_len, tgt_len]
dec_enc_attn_mask = get_attn_pad_mask(dec_inputs, enc_inputs) # [batc_size, tgt_len, src_len]
dec_self_attns, dec_enc_attns = [], []
for layer in self.layers:
# dec_outputs: [batch_size, tgt_len, d_model], dec_self_attn: [batch_size, n_heads, tgt_len, tgt_len], dec_enc_attn: [batch_size, h_heads, tgt_len, src_len]
dec_outputs, dec_self_attn, dec_enc_attn = layer(dec_outputs, enc_outputs, dec_self_attn_mask, dec_enc_attn_mask)
dec_self_attns.append(dec_self_attn)
dec_enc_attns.append(dec_enc_attn)
return dec_outputs, dec_self_attns, dec_enc_attns
Decoder中不仅要把"pad"mask掉,还要mask未来时刻的信息,因此就有了下面这三行代码,其中torch.gt(a, value)
的意思是,将a中各个位置上的元素和value比较,若大于value,则该位置取1,否则取0
dec_self_attn_pad_mask = get_attn_pad_mask(dec_inputs, dec_inputs) # [batch_size, tgt_len, tgt_len]
dec_self_attn_subsequence_mask = get_attn_subsequence_mask(dec_inputs) # [batch_size, tgt_len, tgt_len]
dec_self_attn_mask = torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequence_mask), 0) # [batch_size, tgt_len, tgt_len]
Transformer
class Transformer(nn.Module):
def __init__(self):
super(Transformer, self).__init__()
self.encoder = Encoder()
self.decoder = Decoder()
self.projection = nn.Linear(d_model, tgt_vocab_size, bias=False)
def forward(self, enc_inputs, dec_inputs):
'''
enc_inputs: [batch_size, src_len]
dec_inputs: [batch_size, tgt_len]
'''
# tensor to store decoder outputs
# outputs = torch.zeros(batch_size, tgt_len, tgt_vocab_size).to(self.device)
# enc_outputs: [batch_size, src_len, d_model], enc_self_attns: [n_layers, batch_size, n_heads, src_len, src_len]
enc_outputs, enc_self_attns = self.encoder(enc_inputs)
# dec_outpus: [batch_size, tgt_len, d_model], dec_self_attns: [n_layers, batch_size, n_heads, tgt_len, tgt_len], dec_enc_attn: [n_layers, batch_size, tgt_len, src_len]
dec_outputs, dec_self_attns, dec_enc_attns = self.decoder(dec_inputs, enc_inputs, enc_outputs)
dec_logits = self.projection(dec_outputs) # dec_logits: [batch_size, tgt_len, tgt_vocab_size]
return dec_logits.view(-1, dec_logits.size(-1)), enc_self_attns, dec_self_attns, dec_enc_attns
Transformer主要就是调用Encoder和Decoder。最后返回dec_logits
的维度是[batch_size * tgt_len, tgt_vocab_size],可以理解为,一个句子,这个句子有batch_size*tgt_len个单词,每个单词有tgt_vocab_size种情况,取概率最大者
模型&损失函数&优化器
model = Transformer()
criterion = nn.CrossEntropyLoss(ignore_index=0)
optimizer = optim.SGD(model.parameters(), lr=1e-3, momentum=0.99)
这里的损失函数里面我设置了一个参数ignore_index=0
,因为"pad"这个单词的索引为0,这样设置以后,就不会计算"pad"的损失(因为本来"pad"也没有意义,不需要计算),关于这个参数更详细的说明,可以看我这篇文章的最下面,稍微提了一下
训练
for epoch in range(30):
for enc_inputs, dec_inputs, dec_outputs in loader:
'''
enc_inputs: [batch_size, src_len]
dec_inputs: [batch_size, tgt_len]
dec_outputs: [batch_size, tgt_len]
'''
# enc_inputs, dec_inputs, dec_outputs = enc_inputs.to(device), dec_inputs.to(device), dec_outputs.to(device)
# outputs: [batch_size * tgt_len, tgt_vocab_size]
outputs, enc_self_attns, dec_self_attns, dec_enc_attns = model(enc_inputs, dec_inputs)
loss = criterion(outputs, dec_outputs.view(-1))
print('Epoch:', '%04d' % (epoch + 1), 'loss =', '{:.6f}'.format(loss))
optimizer.zero_grad()
loss.backward()
optimizer.step()
测试
enc_inputs, dec_inputs, _ = next(iter(loader))
predict, _, _, _ = model(enc_inputs[0].view(1, -1), dec_inputs[0].view(1, -1)) # model(enc_inputs[0].view(1, -1), greedy_dec_input)
predict = predict.data.max(1, keepdim=True)[1]
print(enc_inputs[0], '->', [idx2word[n.item()] for n in predict.squeeze()])
最后给出完整代码链接(需要科学的力量)
Github 项目地址:nlp-tutorial