#most of the codes below are copied from Query2label and DETR import torch import torch.nn.functional as F from torch import nn, Tensor from torch.nn import MultiheadAttention from typing import Optional, List import copy def _get_activation_fn(activation): if activation == "relu": return F.relu if activation == "gelu": return F.gelu if activation == "glu": return F.glu raise RuntimeError(F"activation should be relu/gelu, not {activation}.") class TransformerEncoderLayer(nn.Module): def __init__(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, activation="relu"): super().__init__() self.self_attn = MultiheadAttention(d_model, nhead, dropout=dropout) self.linear1 = nn.Linear(d_model, dim_feedforward) self.dropout = nn.Dropout(dropout) self.linear2 = nn.Linear(dim_feedforward, d_model) self.norm1 = nn.LayerNorm(d_model) self.norm2 = nn.LayerNorm(d_model) self.dropout1 = nn.Dropout(dropout) self.dropout2 = nn.Dropout(dropout) self.activation = _get_activation_fn(activation) def with_pos_embed(self, tensor, pos: Optional[Tensor]): return tensor if pos is None else tensor + pos def forward(self, src, src_mask: Optional[Tensor] = None, src_key_padding_mask: Optional[Tensor] = None, pos: Optional[Tensor] = None): src2 = self.norm1(src) q = k = self.with_pos_embed(src2, pos) src2 = self.self_attn(q, k, value=src2, attn_mask=src_mask, key_padding_mask=src_key_padding_mask)[0] src = src + self.dropout1(src2) src2 = self.norm2(src) src2 = self.linear2(self.dropout(self.activation(self.linear1(src2)))) src = src + self.dropout2(src2) return src class TransformerDecoderLayer(nn.Module): def __init__(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, activation="relu"): super().__init__() self.self_attn = MultiheadAttention(d_model, nhead, dropout=dropout) self.multihead_attn = MultiheadAttention(d_model, nhead, dropout=dropout) # Implementation of Feedforward model self.linear1 = nn.Linear(d_model, dim_feedforward) self.dropout = nn.Dropout(dropout) self.linear2 = nn.Linear(dim_feedforward, d_model) self.norm1 = nn.LayerNorm(d_model) self.norm2 = nn.LayerNorm(d_model) self.norm3 = nn.LayerNorm(d_model) self.dropout1 = nn.Dropout(dropout) self.dropout2 = nn.Dropout(dropout) self.dropout3 = nn.Dropout(dropout) self.activation = _get_activation_fn(activation) def with_pos_embed(self, tensor, pos: Optional[Tensor]): return tensor if pos is None else tensor + pos def forward(self, tgt, memory, tgt_mask: Optional[Tensor] = None, memory_mask: Optional[Tensor] = None, tgt_key_padding_mask: Optional[Tensor] = None, memory_key_padding_mask: Optional[Tensor] = None, pos: Optional[Tensor] = None, query_pos: Optional[Tensor] = None): tgt2 = self.norm1(tgt) q = k = self.with_pos_embed(tgt2, query_pos) tgt2 = self.self_attn(q, k, value=tgt2, attn_mask=tgt_mask, key_padding_mask=tgt_key_padding_mask)[0] tgt = tgt + self.dropout1(tgt2) tgt2 = self.norm2(tgt) tgt2 = self.multihead_attn(query=self.with_pos_embed(tgt2, query_pos), key=self.with_pos_embed(memory, pos), value=memory, attn_mask=memory_mask, key_padding_mask=memory_key_padding_mask)[0] tgt = tgt + self.dropout2(tgt2) tgt2 = self.norm3(tgt) tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt2)))) tgt = tgt + self.dropout3(tgt2) return tgt def _get_clones(module, N): return nn.ModuleList([copy.deepcopy(module) for i in range(N)]) class TransformerEncoder(nn.Module): def __init__(self, encoder_layer, num_layers, norm=None): super().__init__() self.layers = _get_clones(encoder_layer, num_layers) self.num_layers = num_layers self.norm = norm def forward(self, src, mask: Optional[Tensor] = None, src_key_padding_mask: Optional[Tensor] = None, pos: Optional[Tensor] = None): output = src for layer in self.layers: output = layer(output, src_mask=mask, src_key_padding_mask=src_key_padding_mask, pos=pos) if self.norm is not None: output = self.norm(output) return output class TransformerDecoder(nn.Module): def __init__(self, decoder_layer, num_layers, norm=None, return_intermediate=False): super().__init__() self.layers = _get_clones(decoder_layer, num_layers) self.num_layers = num_layers self.norm = norm self.return_intermediate = return_intermediate def forward(self, tgt, memory, tgt_mask: Optional[Tensor] = None, memory_mask: Optional[Tensor] = None, tgt_key_padding_mask: Optional[Tensor] = None, memory_key_padding_mask: Optional[Tensor] = None, pos: Optional[Tensor] = None, query_pos: Optional[Tensor] = None): output = tgt for layer in self.layers: output = layer(output, memory, tgt_mask=tgt_mask, memory_mask=memory_mask, tgt_key_padding_mask=tgt_key_padding_mask, memory_key_padding_mask=memory_key_padding_mask, pos=pos, query_pos=query_pos) if self.norm is not None: output = self.norm(output) return output class Transformer(nn.Module): def __init__(self, d_model=512, nhead=8, num_encoder_layers=6, num_decoder_layers=6, dim_feedforward=2048, dropout=0.1, activation="relu" ): super().__init__() self.num_encoder_layers = num_encoder_layers if num_decoder_layers > 0: encoder_layer = TransformerEncoderLayer(d_model, nhead, dim_feedforward, dropout, activation) encoder_norm = nn.LayerNorm(d_model) self.encoder = TransformerEncoder(encoder_layer, num_encoder_layers, encoder_norm) decoder_layer = TransformerDecoderLayer(d_model, nhead, dim_feedforward, dropout, activation) decoder_norm = nn.LayerNorm(d_model) self.decoder = TransformerDecoder(decoder_layer, num_decoder_layers, decoder_norm) self._reset_parameters() self.d_model = d_model self.nhead = nhead def _reset_parameters(self): for p in self.parameters(): if p.dim() > 1: nn.init.xavier_uniform_(p) def forward(self, src, query_embed, pos_embed=None, mask=None): bs, c, w = src.shape src = src.permute(2, 0, 1) query_embed = query_embed.transpose(0,1) # .unsqueeze(1).repeat(1, bs, 1) if mask is not None: mask = mask.flatten(1) memory = self.encoder(src, src_key_padding_mask=mask, pos=pos_embed) hs = self.decoder(query_embed, memory, memory_key_padding_mask=mask, pos=pos_embed, query_pos=None) return hs.transpose(0,1) # return hs.transpose(1, 2), memory[:h * w].permute(1, 2, 0).view(bs, c, h, w)