| | import torch |
| | from torch import nn |
| | from transformers import CLIPVisionModel, CLIPImageProcessor |
| |
|
| | class VisualToGPTMapping(nn.Module): |
| | def __init__(self, visual_emb_dim, gpt_emb_dim, num_gpt_embs, num_heads): |
| | super(VisualToGPTMapping, self).__init__() |
| | self.transformer_layer = TransformerEncoderLayer(d_model=visual_emb_dim, nhead=num_heads, batch_first=True, norm_first=False) |
| | self.linear = Linear(visual_emb_dim, gpt_emb_dim) |
| | self.n_embeddings = num_gpt_embs |
| | self.embedding_dim = gpt_emb_dim |
| | def forward(self, visual_embs): |
| | out = self.transformer_layer(visual_embs) |
| | out = self.linear(out).view(-1, self.n_embeddings, self.embedding_dim) |
| | return out |
| |
|
| | class CLIPVisionTower(nn.Module): |
| | def __init__(self, vision_tower, delay_load=False): |
| | super().__init__() |
| |
|
| | self.is_loaded = False |
| |
|
| | self.vision_tower_name = vision_tower |
| | self.select_layer = -2 |
| | self.select_feature = 'patch' |
| |
|
| | if not delay_load: |
| | self.load_model() |
| | else: |
| | self.cfg_only = CLIPVisionConfig.from_pretrained(self.vision_tower_name) |
| |
|
| | def load_model(self): |
| | self.image_processor = CLIPImageProcessor.from_pretrained(self.vision_tower_name) |
| | self.vision_tower = CLIPVisionModel.from_pretrained(self.vision_tower_name) |
| | self.vision_tower.requires_grad_(False) |
| |
|
| | self.is_loaded = True |
| |
|
| | def feature_select(self, image_forward_outs): |
| | image_features = image_forward_outs.hidden_states[self.select_layer] |
| | if self.select_feature == 'patch': |
| | image_features = image_features[:, 1:] |
| | elif self.select_feature == 'cls_patch': |
| | image_features = image_features |
| | else: |
| | raise ValueError(f'Unexpected select feature: {self.select_feature}') |
| | return image_features |
| |
|
| | @torch.no_grad() |
| | def forward(self, images): |
| | if type(images) is list: |
| | image_features = [] |
| | for image in images: |
| | image_forward_out = self.vision_tower(image.to(device=self.device, dtype=self.dtype).unsqueeze(0), output_hidden_states=True) |
| | image_feature = self.feature_select(image_forward_out).to(image.dtype) |
| | image_features.append(image_feature) |
| | else: |
| | image_forward_outs = self.vision_tower(images.to(device=self.device, dtype=self.dtype), output_hidden_states=True) |
| | image_features = self.feature_select(image_forward_outs).to(images.dtype) |
| |
|
| | return image_features |
| |
|
| | @property |
| | def dummy_feature(self): |
| | return torch.zeros(1, self.hidden_size, device=self.device, dtype=self.dtype) |
| |
|
| | @property |
| | def dtype(self): |
| | return self.vision_tower.dtype |
| |
|
| | @property |
| | def device(self): |
| | return self.vision_tower.device |
| |
|
| | @property |
| | def config(self): |
| | if self.is_loaded: |
| | return self.vision_tower.config |
| | else: |
| | return self.cfg_only |
| |
|
| | @property |
| | def hidden_size(self): |
| | return self.config.hidden_size |
