Transformer简版实战教程
Transformer简版实战教程
至于Transformer的理论内容可以参考Transformer 与 Attention和Transformer 与 Attention的一些Trick
本文主要实战， 这是一个简单版本的Transformer实现，也便于大家理解。
准备
需要准备的是翻译的语料集sentences以及模型参数src_vocab-输入词表， tgt_vocab目标词表，src_len 和tgt_len是句子的最大长度，d_model是hidden_size维度大小， d_ff是前馈网络中间CNN的输出维度。n_layers 是transformer block的个数，n_heads是头的个数， 因此多头的后最后维度为d_model/n_heads（64）。注意ich mochte ein bier P中的’P’是Padding值。
## 语料集准备
sentences = ['ich mochte ein bier P', 'S i want a beer', 'i want a beer E']
## 模型参数
src_vocab = {'P': 0, 'ich': 1, 'mochte': 2, 'ein': 3, 'bier': 4}
src_vocab_size = len(src_vocab)
tgt_vocab = {'P': 0, 'i': 1, 'want': 2, 'a': 3, 'beer': 4, 'S': 5, 'E': 6} ## 目标词表大小
number_dict = {i: w for i, w in enumerate(tgt_vocab)} ##
tgt_vocab_size = len(tgt_vocab)
src_len = 5 # length of source
tgt_len = 5 # length of target
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
Transfomer
Transformer主要是由encoder、decoder和Linear组成。
输入：
- enc_inputs：[batch_size, src_len, hidden_size]
- dec_inputs：[batch_size, tgt_len, hidden_size]
矩阵变换： - enc_outputs：[batch_size, src_len, hidden_size]
- dec_outputs：[batch_size, tgt_len, hidden_size]
- enc_self_attns：
- dec_self_attns：
- dec_logits [batch_size, src_len, tgt_len]
输出：
dec_logits.view(-1, dec_logits.size(-1)): [batch_size* src_len, tgt_len]
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_outputs, enc_self_attns = self.encoder(enc_inputs)
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 x src_vocab_size x tgt_vocab_size]
return dec_logits.view(-1, dec_logits.size(-1)), enc_self_attns, dec_self_attns, dec_enc_attns
Encoder
主要由src_embedding, postion_embedding和n_layer个EncoderLayer组成。
输入：
- enc_inputs：[batch_size, src_len]
- [[1,2,3,4,0]] 其实也是输入，因为是一个batch，固定写死了， 0是padding值
矩阵变换：
enc_outputs ：[batch_size, src_len, hidden_size] (开始时input embedding与position embedding相加)
enc_self_attn_mask ：[batch_size, src_len, src_len] (encoder层注意力掩码)
enc_self_attn：[batch_size, n_head, src_len, src_len]
输出：
enc_outputs：[batch_size, src_len, hidden_size]
enc_self_attns：[batch_size, n_head, src_len, src_len]
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_len+1, d_model),freeze=True)
self.layers = nn.ModuleList([EncoderLayer() for _ in range(n_layers)])
def forward(self, enc_inputs): # enc_inputs : [batch_size x source_len]
enc_outputs = self.src_emb(enc_inputs) + self.pos_emb(torch.LongTensor([[1,2,3,4,0]]))
enc_self_attn_mask = get_attn_pad_mask(enc_inputs, enc_inputs)
enc_self_attns = []
for layer in self.layers:
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
位置编码
位置编码很好理解，偶数位置sin函数，奇数位置cos函数。
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)
attention mask
mask主要是Q与K的注意力掩码，当然在Transformer中Q，K和V是相同的，都是input值。
pad_attn_mask :[batch_size,1, k_len]
pad_attn_mask.expand(batch_size, len_q, len_k): [batch_size , q_len , k_len]
def get_attn_pad_mask(seq_q, seq_k):
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 x 1 x len_k(=len_q), one is masking
return pad_attn_mask.expand(batch_size, len_q, len_k) # batch_size x len_q x len_k
EncoderLayer
主要由多头注意力层（MultiHeadAttention）和前馈层（PoswiseFeedForwardNet）组成。
输入：
- enc_inputs：[batch_size, src_len, hidden_size]
- enc_self_attn_mask: [batch_size, src_len, src_len] (即 [batch_size, q_len, k_len])
矩阵变换： - enc_outputs：[batch_size, src_len, hidden_size]
- attn：[batch_size, n_head, src_len, src_len]
输出：
enc_outputs和attn
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_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 x len_q x d_model]
return enc_outputs, attn
MultiHeadAttention
主要由Q K V线性层和LayerNorm组成
输入：
- Q， K ， V : [batch_size, src_len, hidden_size]
- attn_mask [batch_size, src_len, src_len]
矩阵变换：
q_s = self.W_Q(Q).view(batch_size, -1, n_heads, d_k).transpose(1,2)[batch_size, , src_len, hidden_size] 转为 [batch, n_head, src_len, hidden_size//n_head] ，这里每个头是64。
attn_mask.unsqueeze(1).repeat(1, n_heads, 1, 1)将维度变为 [batch, n_head, src_len, hidden_size//n_head] 便于注意力掩码计算。
ScaledDotProductAttention()(q_s, k_s, v_s, attn_mask)注意力计算，下面会介绍。
context.transpose(1, 2).contiguous().view(batch_size, -1, n_heads * d_v)维度变为原来的[batch_size, , src_len, hidden_size]
self.layer_norm(output + residual)残差+layerNorm ->[batch_size, , src_len, hidden_size]
输出：
class MultiHeadAttention(nn.Module):
def __init__(self):
super(MultiHeadAttention, self).__init__()
self.W_Q = nn.Linear(d_model, d_k * n_heads)
self.W_K = nn.Linear(d_model, d_k * n_heads)
self.W_V = nn.Linear(d_model, d_v * n_heads)
self.linear = nn.Linear(n_heads * d_v, d_model)
self.layer_norm = nn.LayerNorm(d_model)
def forward(self, Q, K, V, attn_mask):
# q: [batch_size x len_q x d_model], k: [batch_size x len_k x d_model], v: [batch_size x len_k x d_model]
residual, batch_size = Q, Q.size(0)
# (B, S, D) -proj-> (B, S, D) -split-> (B, S, H, W) -trans-> (B, H, S, W)
q_s = self.W_Q(Q).view(batch_size, -1, n_heads, d_k).transpose(1,2) # q_s: [batch_size x n_heads x len_q x d_k]
k_s = self.W_K(K).view(batch_size, -1, n_heads, d_k).transpose(1,2) # k_s: [batch_size x n_heads x len_k x d_k]
v_s = self.W_V(V).view(batch_size, -1, n_heads, d_v).transpose(1,2) # v_s: [batch_size x n_heads x len_k x d_v]
attn_mask = attn_mask.unsqueeze(1).repeat(1, n_heads, 1, 1) # attn_mask : [batch_size x n_heads x len_q x len_k]
# context: [batch_size x n_heads x len_q x d_v], attn: [batch_size x n_heads x len_q(=len_k) x len_k(=len_q)]
context, attn = ScaledDotProductAttention()(q_s, k_s, v_s, attn_mask)
context = context.transpose(1, 2).contiguous().view(batch_size, -1, n_heads * d_v) # context: [batch_size x len_q x n_heads * d_v]
output = self.linear(context)
return self.layer_norm(output + residual), attn # output: [batch_size x len_q x d_model]
ScaledDotProductAttention
每个head的Q与响应的K相乘得到注意力得分， attn_mask的作用是padding部分填充-1e9，使其在注意力中不起作用， 权重得分变小。然后进行Softmax，得到权重分布，最后乘以V 得到加权后的得分。维度一致保持[batch_size, n_head, src_len, d_k]
class ScaledDotProductAttention(nn.Module):
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
def forward(self, Q, K, V, attn_mask):
scores = torch.matmul(Q, K.transpose(-1, -2)) / np.sqrt(d_k) # scores:[batch_size x n_heads x len_q(=len_k) x len_k(=len_q)]
scores.masked_fill_(attn_mask, -1e9) # Fills elements of self tensor with value where mask is one.
attn = nn.Softmax(dim=-1)(scores)
context = torch.matmul(attn, V)
return context, attn
PoswiseFeedForwardNet
主要由两层CNN、一层激活层和一层LayerNorm组成。
输入：
- inputs：[batch_size, src_len, hidden_size]
矩阵变换： - 第一层CNN：
inputs.transpose(1, 2)改变维度位置,因为torch dim=1的位置是channel，输出为[batch_size, 2048, hidden_size] - 激活函数RELU：
- 第二层CNN：[batch_size, src_len, hidden_size]
self.layer_norm(output + residual)残差+layerNorm ->[batch_size, , src_len, hidden_size]
class PoswiseFeedForwardNet(nn.Module):
def __init__(self):
super(PoswiseFeedForwardNet, self).__init__()
self.conv1 = nn.Conv1d(in_channels=d_model, out_channels=d_ff, kernel_size=1)
self.conv2 = nn.Conv1d(in_channels=d_ff, out_channels=d_model, kernel_size=1)
self.layer_norm = nn.LayerNorm(d_model)
def forward(self, inputs):
residual = inputs # inputs : [batch_size, len_q, d_model]
output = nn.ReLU()(self.conv1(inputs.transpose(1, 2)))
output = self.conv2(output).transpose(1, 2)
return self.layer_norm(output + residual)
------------------------------------------ Encoder的部分结束，下面我们讲解Decoder部分 ----------------------------------------------
Decoder（也是encdder输出与decoder输入交互层，重点）
主要由tgt_input embedding、tgt_postion embedding和n_layer个DecoderLayer组成。
输入：
- dec_inputs: [batch_size, tgt_len, hidden_size]
- enc_inputs:[batch_size, src_len, hidden_size]
- enc_outputs:[batch_size, src_len, hidden_size]
矩阵变换
dec_self_attn_pad_maskdecoder开始的padding 掩码部分，与encoder掩码一样，只是输入的是decoder的输入。
dec_self_attn_subsequent_mask解码掩码上三角，因为在翻译的时候，当前词是看不到后面词的信息。get_attn_subsequent_mask函数中np.triu方法返回的是上三角矩阵[batch_size, src_len, src_len]，如下图
torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequent_mask), 0) -> torch.gt(a,b)函数比较a中元素大于（这里是严格大于）b中对应元素目的是转为bool值。
get_attn_pad_mask(dec_inputs, enc_inputs) 最终的encoder-decoder padding掩码，与单纯的encoder或者decoderpadding掩码不一样， 输入的是decoder input和encoder input， 其中decoder 部分最为Q， decoder部分最为K， 最中掩码维度为[batch_size, q_len, k_len]
输出：
- dec_outputs：[batch_size, tgt_len, hidden_size]
- dec_self_attns：[batch_size, tgt_len, tgt_len]
- dec_enc_attns：[batch_size, tgt_len, src_len]
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_len+1, 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 x target_len]
dec_outputs = self.tgt_emb(dec_inputs) + self.pos_emb(torch.LongTensor([[5,1,2,3,4]]))
dec_self_attn_pad_mask = get_attn_pad_mask(dec_inputs, dec_inputs)
dec_self_attn_subsequent_mask = get_attn_subsequent_mask(dec_inputs)
dec_self_attn_mask = torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequent_mask), 0)
dec_enc_attn_mask = get_attn_pad_mask(dec_inputs, enc_inputs)
dec_self_attns, dec_enc_attns = [], []
for layer in self.layers:
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
def get_attn_subsequent_mask(seq):
attn_shape = [seq.size(0), seq.size(1), seq.size(1)]
subsequent_mask = np.triu(np.ones(attn_shape), k=1)
subsequent_mask = torch.from_numpy(subsequent_mask).byte()
return subsequent_mask
DecoderLayer
主要由自注意力MultiHeadAttention,交互注意力MultiHeadAttention(输入的是dec_enc_attn_mask掩码)，和前馈神经PoswiseFeedForwardNet组成（与encoder一样）
输入：
- dec_inputs： [batch_size, tgt_len, hidden_size]
- enc_outputs：[batch_size, src_len, hidden_size]
- dec_self_attn_mask：[batch_size, tgt_len, tgt_len]
- dec_enc_attn_mask：[batch_size, tgt_len, src_len]
矩阵变化：
对头注意力层和前馈层与encoder的结构一样， 这里不一一介绍。这里需要注意的是交互注意力层输入不一样。
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_outputs, dec_self_attn = self.dec_self_attn(dec_inputs, dec_inputs, dec_inputs, dec_self_attn_mask)
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)
return dec_outputs, dec_self_attn, dec_enc_attn
最后是一层线性层，输出logit,维度为[batch_size, src_len, tgt_vocab_size]
完整代码
# -*- coding:UTF-8 -*-
# author:user
# contact: [email protected]
# datetime:2021/12/22 16:32
# software: PyCharm
"""
文件说明：
"""
import numpy as np
import torch
import torch.nn as nn
import torch.optim as optim
import matplotlib.pyplot as plt
# 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
def make_batch(sentences):
input_batch = [[src_vocab[n] for n in sentences[0].split()]]
output_batch = [[tgt_vocab[n] for n in sentences[1].split()]]
target_batch = [[tgt_vocab[n] for n in sentences[2].split()]]
return torch.LongTensor(input_batch), torch.LongTensor(output_batch), torch.LongTensor(target_batch)
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)
def get_attn_pad_mask(seq_q, seq_k):
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 x 1 x len_k(=len_q), one is masking
return pad_attn_mask.expand(batch_size, len_q, len_k) # batch_size x len_q x len_k
def get_attn_subsequent_mask(seq):
attn_shape = [seq.size(0), seq.size(1), seq.size(1)]
subsequent_mask = np.triu(np.ones(attn_shape), k=1)
subsequent_mask = torch.from_numpy(subsequent_mask).byte()
return subsequent_mask
class ScaledDotProductAttention(nn.Module):
def __init__(self):
super(ScaledDotProductAttention, self).__init__()
def forward(self, Q, K, V, attn_mask):
scores = torch.matmul(Q, K.transpose(-1, -2)) / np.sqrt(d_k) # scores : [batch_size x n_heads x len_q(=len_k) x len_k(=len_q)]
scores.masked_fill_(attn_mask, -1e9) # Fills elements of self tensor with value where mask is one.
attn = nn.Softmax(dim=-1)(scores)
context = torch.matmul(attn, V)
return context, attn
class MultiHeadAttention(nn.Module):
def __init__(self):
super(MultiHeadAttention, self).__init__()
self.W_Q = nn.Linear(d_model, d_k * n_heads)
self.W_K = nn.Linear(d_model, d_k * n_heads)
self.W_V = nn.Linear(d_model, d_v * n_heads)
self.linear = nn.Linear(n_heads * d_v, d_model)
self.layer_norm = nn.LayerNorm(d_model)
def forward(self, Q, K, V, attn_mask):
# q: [batch_size x len_q x d_model], k: [batch_size x len_k x d_model], v: [batch_size x len_k x d_model]
residual, batch_size = Q, Q.size(0)
# (B, S, D) -proj-> (B, S, D) -split-> (B, S, H, W) -trans-> (B, H, S, W)
q_s = self.W_Q(Q).view(batch_size, -1, n_heads, d_k).transpose(1,2) # q_s: [batch_size x n_heads x len_q x d_k]
k_s = self.W_K(K).view(batch_size, -1, n_heads, d_k).transpose(1,2) # k_s: [batch_size x n_heads x len_k x d_k]
v_s = self.W_V(V).view(batch_size, -1, n_heads, d_v).transpose(1,2) # v_s: [batch_size x n_heads x len_k x d_v]
attn_mask = attn_mask.unsqueeze(1).repeat(1, n_heads, 1, 1) # attn_mask : [batch_size x n_heads x len_q x len_k]
# context: [batch_size x n_heads x len_q x d_v], attn: [batch_size x n_heads x len_q(=len_k) x len_k(=len_q)]
context, attn = ScaledDotProductAttention()(q_s, k_s, v_s, attn_mask)
context = context.transpose(1, 2).contiguous().view(batch_size, -1, n_heads * d_v) # context: [batch_size x len_q x n_heads * d_v]
output = self.linear(context)
return self.layer_norm(output + residual), attn # output: [batch_size x len_q x d_model]
class PoswiseFeedForwardNet(nn.Module):
def __init__(self):
super(PoswiseFeedForwardNet, self).__init__()
self.conv1 = nn.Conv1d(in_channels=d_model, out_channels=d_ff, kernel_size=1)
self.conv2 = nn.Conv1d(in_channels=d_ff, out_channels=d_model, kernel_size=1)
self.layer_norm = nn.LayerNorm(d_model)
def forward(self, inputs):
residual = inputs # inputs : [batch_size, len_q, d_model]
output = nn.ReLU()(self.conv1(inputs.transpose(1, 2)))
output = self.conv2(output).transpose(1, 2)
return self.layer_norm(output + residual)
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_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 x len_q x d_model]
return enc_outputs, attn
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_outputs, dec_self_attn = self.dec_self_attn(dec_inputs, dec_inputs, dec_inputs, dec_self_attn_mask)
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)
return dec_outputs, dec_self_attn, dec_enc_attn
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_len+1, d_model),freeze=True)
self.layers = nn.ModuleList([EncoderLayer() for _ in range(n_layers)])
def forward(self, enc_inputs): # enc_inputs : [batch_size x source_len]
enc_outputs = self.src_emb(enc_inputs) + self.pos_emb(torch.LongTensor([[1,2,3,4,0]]))
enc_self_attn_mask = get_attn_pad_mask(enc_inputs, enc_inputs)
enc_self_attns = []
for layer in self.layers:
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
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_len+1, 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 x target_len]
dec_outputs = self.tgt_emb(dec_inputs) + self.pos_emb(torch.LongTensor([[5,1,2,3,4]]))
dec_self_attn_pad_mask = get_attn_pad_mask(dec_inputs, dec_inputs)
dec_self_attn_subsequent_mask = get_attn_subsequent_mask(dec_inputs)
dec_self_attn_mask = torch.gt((dec_self_attn_pad_mask + dec_self_attn_subsequent_mask), 0)
dec_enc_attn_mask = get_attn_pad_mask(dec_inputs, enc_inputs)
dec_self_attns, dec_enc_attns = [], []
for layer in self.layers:
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
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_outputs, enc_self_attns = self.encoder(enc_inputs)
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 x src_vocab_size x tgt_vocab_size]
return dec_logits.view(-1, dec_logits.size(-1)), enc_self_attns, dec_self_attns, dec_enc_attns
def showgraph(attn):
attn = attn[-1].squeeze(0)[0]
attn = attn.squeeze(0).data.numpy()
fig = plt.figure(figsize=(n_heads, n_heads)) # [n_heads, n_heads]
ax = fig.add_subplot(1, 1, 1)
ax.matshow(attn, cmap='viridis')
ax.set_xticklabels(['']+sentences[0].split(), fontdict={'fontsize': 14}, rotation=90)
ax.set_yticklabels(['']+sentences[2].split(), fontdict={'fontsize': 14})
plt.show()
if __name__ == '__main__':
sentences = ['ich mochte ein bier P', 'S i want a beer', 'i want a beer E']
# Transformer Parameters
# Padding Should be Zero
src_vocab = {'P': 0, 'ich': 1, 'mochte': 2, 'ein': 3, 'bier': 4}
src_vocab_size = len(src_vocab)
tgt_vocab = {'P': 0, 'i': 1, 'want': 2, 'a': 3, 'beer': 4, 'S': 5, 'E': 6}
number_dict = {i: w for i, w in enumerate(tgt_vocab)}
tgt_vocab_size = len(tgt_vocab)
src_len = 5 # length of source
tgt_len = 5 # length of target
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
model = Transformer()
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=0.001)
enc_inputs, dec_inputs, target_batch = make_batch(sentences)
for epoch in range(20):
optimizer.zero_grad()
outputs, enc_self_attns, dec_self_attns, dec_enc_attns = model(enc_inputs, dec_inputs)
loss = criterion(outputs, target_batch.contiguous().view(-1))
print('Epoch:', '%04d' % (epoch + 1), 'cost =', '{:.6f}'.format(loss))
loss.backward()
optimizer.step()
# Test
predict, _, _, _ = model(enc_inputs, dec_inputs)
predict = predict.data.max(1, keepdim=True)[1]
print(sentences[0], '->', [number_dict[n.item()] for n in predict.squeeze()])
print('first head of last state enc_self_attns')
showgraph(enc_self_attns)
print('first head of last state dec_self_attns')
showgraph(dec_self_attns)
print('first head of last state dec_enc_attns')
showgraph(dec_enc_attns)