File size: 9,398 Bytes
3b2b066 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 |
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import copy
import math
import numpy as np
import torch
import torch.nn as nn
import torch.nn.functional as F
from .Generator import pack_wrapper, GenModel
def clones(module, N):
return nn.ModuleList([copy.deepcopy(module) for _ in range(N)])
def attention(query, key, value, mask=None, dropout=None):
d_k = query.size(-1)
scores = torch.matmul(query, key.transpose(-2, -1)) / math.sqrt(d_k)
if mask is not None:
scores = scores.masked_fill(mask == 0, -1e9)
p_attn = F.softmax(scores, dim=-1)
if dropout is not None:
p_attn = dropout(p_attn)
return torch.matmul(p_attn, value), p_attn
def subsequent_mask(size):
attn_shape = (1, size, size)
subsequent_mask = np.triu(np.ones(attn_shape), k=1).astype('uint8')
return torch.from_numpy(subsequent_mask) == 0
class Transformer(nn.Module):
def __init__(self, encoder, decoder, src_embed, tgt_embed):
super(Transformer, self).__init__()
self.encoder = encoder
self.decoder = decoder
self.src_embed = src_embed
self.tgt_embed = tgt_embed
def forward(self, src, tgt, src_mask, tgt_mask):
return self.decode(self.encode(src, src_mask), src_mask, tgt, tgt_mask)
def encode(self, src, src_mask):
return self.encoder(self.src_embed(src), src_mask)
def decode(self, hidden_states, src_mask, tgt, tgt_mask):
memory = None
return self.decoder(self.tgt_embed(tgt), hidden_states, src_mask, tgt_mask, memory)
class Encoder(nn.Module):
def __init__(self, layer, N):
super(Encoder, self).__init__()
self.layers = clones(layer, N)
self.norm = LayerNorm(layer.d_model)
def forward(self, x, mask):
for layer in self.layers:
x = layer(x, mask)
return self.norm(x)
class EncoderLayer(nn.Module):
def __init__(self, d_model, self_attn, feed_forward, dropout):
super(EncoderLayer, self).__init__()
self.self_attn = self_attn
self.feed_forward = feed_forward
self.sublayer = clones(SublayerConnection(d_model, dropout), 2)
self.d_model = d_model # 512
def forward(self, x, mask):
x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, mask))
return self.sublayer[1](x, self.feed_forward)
class SublayerConnection(nn.Module):
def __init__(self, d_model, dropout):
super(SublayerConnection, self).__init__()
self.norm = LayerNorm(d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x, sublayer):
return x + self.dropout(sublayer(self.norm(x)))
class LayerNorm(nn.Module):
def __init__(self, features, eps=1e-6):
super(LayerNorm, self).__init__()
self.gamma = nn.Parameter(torch.ones(features))
self.beta = nn.Parameter(torch.zeros(features))
self.eps = eps
def forward(self, x):
mean = x.mean(-1, keepdim=True)
std = x.std(-1, keepdim=True)
return self.gamma * (x - mean) / (std + self.eps) + self.beta
class Decoder(nn.Module):
def __init__(self, layer, N):
super(Decoder, self).__init__()
self.layers = clones(layer, N)
self.norm = LayerNorm(layer.d_model)
def forward(self, x, hidden_states, src_mask, tgt_mask, memory):
for layer in self.layers:
x = layer(x, hidden_states, src_mask, tgt_mask, memory)
return self.norm(x)
class DecoderLayer(nn.Module):
def __init__(self, d_model, self_attn, src_attn, feed_forward, dropout):
super(DecoderLayer, self).__init__()
self.d_model = d_model # 512
self.self_attn = self_attn
self.src_attn = src_attn
self.feed_forward = feed_forward
self.sublayer = clones(SublayerConnection(d_model, dropout), 3)
def forward(self, x, hidden_states, src_mask, tgt_mask, memory):
m = hidden_states
x = self.sublayer[0](x, lambda x: self.self_attn(x, x, x, tgt_mask))
x = self.sublayer[1](x, lambda x: self.src_attn(x, m, m, src_mask))
return self.sublayer[2](x, self.feed_forward)
class MultiHeadedAttention(nn.Module):
def __init__(self, h, d_model, dropout=0.1):
super(MultiHeadedAttention, self).__init__()
assert d_model % h == 0
self.d_k = d_model // h
self.h = h
self.linears = clones(nn.Linear(d_model, d_model), 4)
self.attn = None
self.dropout = nn.Dropout(p=dropout)
def forward(self, query, key, value, mask=None):
if mask is not None:
mask = mask.unsqueeze(1)
nbatches = query.size(0)
query, key, value = [l(x).view(nbatches, -1, self.h, self.d_k).transpose(1, 2)
for l, x in zip(self.linears, (query, key, value))]
x, self.attn = attention(query, key, value, mask=mask, dropout=self.dropout)
x = x.transpose(1, 2).contiguous().view(nbatches, -1, self.h * self.d_k)
return self.linears[-1](x)
class PositionwiseFeedForward(nn.Module):
def __init__(self, d_model, d_ff, dropout=0.1):
super(PositionwiseFeedForward, self).__init__()
self.w_1 = nn.Linear(d_model, d_ff)
self.w_2 = nn.Linear(d_ff, d_model)
self.dropout = nn.Dropout(dropout)
def forward(self, x):
return self.w_2(self.dropout(F.relu(self.w_1(x))))
class Embeddings(nn.Module):
def __init__(self, d_model, vocab):
super(Embeddings, self).__init__()
self.lut = nn.Embedding(vocab, d_model)
self.d_model = d_model
def forward(self, x):
return self.lut(x) * math.sqrt(self.d_model)
class PositionalEncoding(nn.Module):
def __init__(self, d_model, dropout, max_len=5000):
super(PositionalEncoding, self).__init__()
self.dropout = nn.Dropout(p=dropout)
pe = torch.zeros(max_len, d_model)
position = torch.arange(0, max_len).unsqueeze(1).float()
div_term = torch.exp(torch.arange(0, d_model, 2).float() *
-(math.log(10000.0) / d_model))
pe[:, 0::2] = torch.sin(position * div_term)
pe[:, 1::2] = torch.cos(position * div_term)
pe = pe.unsqueeze(0)
self.register_buffer('pe', pe)
def forward(self, x):
x = x + self.pe[:, :x.size(1)]
return self.dropout(x)
class EncoderDecoder(GenModel):
def make_model(self, tgt_vocab):
c = copy.deepcopy
attn = MultiHeadedAttention(self.num_heads, self.d_model)
ff = PositionwiseFeedForward(self.d_model, self.d_ff, self.dropout)
position = PositionalEncoding(self.d_model, self.dropout)
model = Transformer(
Encoder(EncoderLayer(self.d_model, c(attn), c(ff), self.dropout), self.num_layers),
Decoder(
DecoderLayer(self.d_model, c(attn), c(attn), c(ff), self.dropout),
self.num_layers),
lambda x: x,
nn.Sequential(Embeddings(self.d_model, tgt_vocab), c(position)))
for p in model.parameters():
if p.dim() > 1:
nn.init.xavier_uniform_(p)
return model
def __init__(self, args, tokenizer):
super(EncoderDecoder, self).__init__(args, tokenizer)
self.args = args
self.num_layers = args.num_layers
self.d_model = args.d_model
self.d_ff = args.d_ff
self.num_heads = args.num_heads
self.dropout = args.dropout
self.vocab_size = tokenizer.get_vocab_size()
tgt_vocab = self.vocab_size + 1
self.model = self.make_model(tgt_vocab)
self.logit = nn.Linear(args.d_model, tgt_vocab)
def _prepare_feature(self, fc_feats, att_feats, att_masks):
att_feats, seq, att_masks, seq_mask = self._prepare_feature_forward(att_feats, att_masks)
memory = self.model.encode(att_feats, att_masks)
return fc_feats[..., :1], att_feats[..., :1], memory, att_masks
def _prepare_feature_forward(self, att_feats, att_masks=None, seq=None):
att_feats, att_masks = self.clip_att(att_feats, att_masks)
att_feats = pack_wrapper(self.att_embed, att_feats, att_masks)
if att_masks is None:
att_masks = att_feats.new_ones(att_feats.shape[:2], dtype=torch.long)
att_masks = att_masks.unsqueeze(-2)
if seq is not None:
seq = seq[:, :-1]
seq_mask = (seq.data > 0)
seq_mask[:, 0] += True
seq_mask = seq_mask.unsqueeze(-2)
seq_mask = seq_mask & subsequent_mask(seq.size(-1)).to(seq_mask)
else:
seq_mask = None
return att_feats, seq, att_masks, seq_mask
def _forward(self, fc_feats, att_feats, seq, att_masks=None):
att_feats, seq, att_masks, seq_mask = self._prepare_feature_forward(att_feats, att_masks, seq)
out = self.model(att_feats, seq, att_masks, seq_mask)
outputs = F.log_softmax(self.logit(out), dim=-1)
return outputs, out
def core(self, it, memory, state, mask):
if len(state) == 0:
ys = it.unsqueeze(1)
else:
ys = torch.cat([state[0][0], it.unsqueeze(1)], dim=1)
out = self.model.decode(memory, mask, ys, subsequent_mask(ys.size(1)).to(memory.device))
return out[:, -1], [ys.unsqueeze(0)]
|