|
from dataclasses import dataclass |
|
import math |
|
import torch |
|
import torch.nn as nn |
|
|
|
@dataclass |
|
class Args: |
|
source_vocab_size: int |
|
target_vocab_size: int |
|
source_sequence_length: int |
|
target_sequence_length: int |
|
d_model: int = 512 |
|
Layers: int = 6 |
|
heads: int = 8 |
|
dropout: float = 0.1 |
|
d_ff: int = 2048 |
|
|
|
class InputEmbeddingLayer(nn.Module): |
|
def __init__(self, d_model: int, vocab_size: int) -> None: |
|
super().__init__() |
|
self.d_model = d_model |
|
self.vocab_size = vocab_size |
|
self.embedding = nn.Embedding(vocab_size, d_model) |
|
|
|
def forward(self, x): |
|
return self.embedding(x) * math.sqrt(self.d_model) |
|
|
|
class PositionalEncodingLayer(nn.Module): |
|
def __init__(self, d_model: int, sequence_length: int, dropout: float) -> None: |
|
super().__init__() |
|
self.d_model = d_model |
|
self.sequence_length = sequence_length |
|
self.dropout = nn.Dropout(dropout) |
|
|
|
PE = torch.zeros(sequence_length, d_model) |
|
Position = torch.arange(0, sequence_length, dtype=torch.float).unsqueeze(1) |
|
deviation_term = torch.exp(torch.arange(0, d_model, 2).float() * (-math.log(10000.0) / d_model)) |
|
|
|
PE[:, 0::2] = torch.sin(Position * deviation_term) |
|
PE[:, 1::2] = torch.cos(Position * deviation_term) |
|
PE = PE.unsqueeze(0) |
|
self.register_buffer('PE', PE) |
|
|
|
def forward(self, x): |
|
x = x + (self.PE[:, :x.shape[1], :]).requires_grad(False) |
|
return self.dropout(x) |
|
|
|
class NormalizationLayer(nn.Module): |
|
def __init__(self, Epsilon: float = 10**-4) -> None: |
|
super().__init__() |
|
self.Epsilon = Epsilon |
|
self.Alpha = nn.Parameter(torch.ones(1)) |
|
self.Bias = nn.Parameter(torch.ones(1)) |
|
|
|
def forward(self, x): |
|
mean = x.mean(dim = -1, keepdim = True) |
|
std = x.std(dim = -1, keepdim = True) |
|
return self.Alpha * (x - mean) / (std + self.Epsilon) + self.Bias |
|
|
|
class FeedForwardBlock(nn.Module): |
|
def __init__(self, d_model: int, d_ff: int, dropout: float) -> None: |
|
super().__init__() |
|
self.Linear_1 = nn.Linear(d_model, d_ff) |
|
self.dropout = nn.Dropout(dropout) |
|
self.Linear_2 = nn.Linear(d_ff, d_model) |
|
|
|
def forward(self, x): |
|
return self.Linear_2(self.dropout(torch.relu(self.Linear_1(x)))) |
|
|
|
class MultiHeadAttentionBlock(nn.Module): |
|
def __init__(self, d_model: int, heads: int, dropout: float) -> None: |
|
super().__init__() |
|
self.d_model = d_model |
|
self.heads = heads |
|
assert d_model % heads == 0, "d_model is not divisable by heads" |
|
|
|
self.d_k = d_model // heads |
|
|
|
self.W_Q = nn.Linear(d_model, d_model) |
|
self.W_K = nn.Linear(d_model, d_model) |
|
self.W_V = nn.Linear(d_model, d_model) |
|
self.W_O = nn.Linear(d_model, d_model) |
|
|
|
self.dropout = nn.Dropout(dropout) |
|
|
|
@staticmethod |
|
def Attention(Query, Key, Value, mask, dropout): |
|
d_k = Query.shape[-1] |
|
self_attention_scores = (Query @ Key.traspose(-2, -1)) / math.sqrt(d_k) |
|
|
|
if mask is not None: |
|
self_attention_scores.masked_fill(mask == 0, -1e9) |
|
self_attention_scores = self_attention_scores.Softmax(dim = -1) |
|
|
|
if dropout is not None: |
|
self_attention_scores = dropout(self_attention_scores) |
|
return self_attention_scores @ Value |
|
|
|
def forward(self, query, key, value, mask): |
|
Query = self.W_Q(query) |
|
Key = self.W_K(key) |
|
Value = self.W_V(value) |
|
|
|
Query = Query.view(Query.shape[0], Query.shape[1], self.heads, self.d_k).transpose(1,2) |
|
Key = Key.view(Key.shape[0], Key.shape[1], self.heads, self.d_k).transpose(1,2) |
|
Value = Value.view(Value.shape[0], Value.shape[1], self.heads, self.d_k).transpose(1,2) |
|
|
|
x, self.self_attention_scores = MultiHeadAttentionBlock.Attention(Query, Key, Value, mask, self.dropout) |
|
x = x.transpose().contiguous().view(x.shape[0], -1, self.heads * self.d_k) |
|
return self.W_O(x) |
|
|
|
class ResidualConnection(nn.Module): |
|
def __init__(self, dropout: float) -> None: |
|
super().__init__() |
|
self.dropout = nn.Dropout(dropout) |
|
self.normalization_layer = NormalizationLayer() |
|
|
|
def forward(self, x, subLayer): |
|
return self.dropout(subLayer(self.normalization_layer)) |
|
|
|
class EncoderBlock(nn.Module): |
|
def __init__(self, self_attetion_block: MultiHeadAttentionBlock, feed_forward_block: FeedForwardBlock, dropout: float) -> None: |
|
super().__init__() |
|
self.self_attention_block = self_attetion_block |
|
self.feed_forward_block = feed_forward_block |
|
self.residual_connection = nn.ModuleList([ResidualConnection(dropout) for _ in range(2)]) |
|
|
|
def forward(self, x, source_mask): |
|
x = self.residual_connection[0](x, lambda x: self.self_attention_block(x, x, x, source_mask)) |
|
x = self.residual_connection[1](x, self.feed_forward_block) |
|
return x |
|
|
|
class Encoder(nn.Module): |
|
def __init__(self, Layers: nn.ModuleList) -> None: |
|
super().__init__() |
|
self.Layers = Layers |
|
self.normalization_layer = NormalizationLayer() |
|
|
|
def forward(self, x, source_mask): |
|
for layer in self.Layers: |
|
x = layer(x, source_mask) |
|
return self.normalization_layer(x) |
|
|
|
class DecoderBlock(nn.Module): |
|
def __init__(self, masked_self_attention_block: MultiHeadAttentionBlock, self_attention_block: MultiHeadAttentionBlock, feedforwardblock: FeedForwardBlock, dropout: float) -> None: |
|
super().__init__() |
|
self.masked_self_attention_block = masked_self_attention_block |
|
self.self_attention_block = self_attention_block |
|
self.feedforwardblock = feedforwardblock |
|
self.residual_connection = nn.ModuleList([ResidualConnection(dropout) for _ in range(3)]) |
|
|
|
def forward(self, x, Encoder_output, source_mask, target_mask): |
|
x = self.residual_connection[0](x, lambda x: self.masked_self_attention_block(x, x, x, source_mask)) |
|
x = self.residual_connection[1](x, lambda x: self.self_attention_block(x, Encoder_output, Encoder_output, target_mask)) |
|
x = self.residual_connection[1](x, self.feedforwardblock) |
|
return x |
|
|
|
class Decoder(nn.Module): |
|
def __init__(self, Layers: nn.ModuleList) -> None: |
|
super().__init__() |
|
self.Layers = Layers |
|
self.normalization_layer = NormalizationLayer() |
|
|
|
def forward(self, x, Encoder_output, source_mask, target_mask): |
|
for layer in self.Layers: |
|
x = layer(x, Encoder_output, source_mask, target_mask) |
|
return self.normalization_layer(x) |
|
|
|
class LinearLayer(nn.Module): |
|
def __init__(self, d_model: int, vocab_size: int) -> None: |
|
super().__init__() |
|
self.Linear = nn.Linear(d_model, vocab_size) |
|
|
|
def forward(self, x): |
|
return self.Linear(x) |
|
|
|
class TransformerBlock(nn.Module): |
|
def __init__(self, encoder: Encoder, |
|
decoder: Decoder, |
|
source_embedding: InputEmbeddingLayer, |
|
target_embedding: InputEmbeddingLayer, |
|
source_position: PositionalEncodingLayer, |
|
target_position: PositionalEncodingLayer, |
|
Linear: LinearLayer) -> None: |
|
super().__init__() |
|
self.encoder = encoder |
|
self.decoder = decoder |
|
self.source_embedding = source_embedding |
|
self.target_embedding = target_embedding |
|
self.source_position = source_position |
|
self.target_position = target_position |
|
self.Linear = Linear |
|
|
|
def encode(self, source_language, source_mask): |
|
source_language = self.source_embedding(source_language) |
|
source_language = self.source_position(source_language) |
|
return self.encoder(source_language, source_mask) |
|
|
|
def decode(self, Encoder_output, source_mask, target_language, target_mask): |
|
target_language = self.target_embedding(target_language) |
|
target_language = self.target_position(target_language) |
|
return self.decoder(target_language, Encoder_output, source_mask, target_mask) |
|
|
|
def linear(self, x): |
|
return self.Linear(x) |
|
|
|
|
|
def Transformer_model(Args: Args)->TransformerBlock: |
|
|
|
source_embedding = InputEmbeddingLayer(Args.d_model, Args.source_vocab_size) |
|
source_position = PositionalEncodingLayer(Args.d_model, Args.source_sequence_length, Args.dropout) |
|
|
|
target_embedding = InputEmbeddingLayer(Args.d_model, Args.target_vocab_size) |
|
target_position = PositionalEncodingLayer(Args.d_model, Args.target_sequence_length, Args.dropout) |
|
|
|
Encoder_Blocks = [] |
|
for _ in range(Args.Layers): |
|
encoder_self_attention_block = MultiHeadAttentionBlock(Args.d_model, Args.heads, Args.dropout) |
|
encoder_feed_forward_block = FeedForwardBlock(Args.d_model, Args.d_ff, Args.dropout) |
|
encoder_block = EncoderBlock(encoder_self_attention_block, encoder_feed_forward_block, Args.dropout) |
|
Encoder_Blocks.append(encoder_block) |
|
|
|
Decoder_Blocks = [] |
|
for _ in range(Args.Layers): |
|
decoder_self_attention_block = MultiHeadAttentionBlock(Args.d_model, Args.heads, Args.dropout) |
|
decoder_cross_attention_block = MultiHeadAttentionBlock(Args.d_model, Args.heads, Args.dropout) |
|
decoder_feed_forward_block = FeedForwardBlock(Args.d_model, Args.d_ff, Args.dropout) |
|
decoder_block = DecoderBlock(decoder_self_attention_block, decoder_cross_attention_block, decoder_feed_forward_block, Args.dropout) |
|
Decoder_Blocks.append(decoder_block) |
|
|
|
encoder = Encoder(nn.ModuleList(Encoder_Blocks)) |
|
decoder = Decoder(nn.ModuleList(Decoder_Blocks)) |
|
|
|
linear = LinearLayer(Args.d_model, Args.target_vocab_size) |
|
|
|
Transformer = TransformerBlock(encoder, decoder, source_embedding, target_embedding, source_position, target_position, linear) |
|
|
|
for t in Transformer.parameters(): |
|
if t.dim() > 1: |
|
nn.init.xavier_uniform(t) |
|
return Transformer |