File size: 2,952 Bytes
bbfa6f6 |
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 |
import torch
import torch.nn as nn
# from transformers import CLIPVisionModel, CLIPImageProcessor, CLIPVisionConfig
from .siglip.modeling_siglip import SiglipVisionModel
from .siglip.configuration_siglip import SiglipVisionConfig
from transformers import AutoImageProcessor
class SiglipVisionTower(nn.Module):
def __init__(self, vision_tower, args, delay_load=False):
super().__init__()
self.is_loaded = False
self.vision_tower_name = vision_tower
self.select_layer = args.mm_vision_select_layer
self.args = args
if not delay_load:
self.load_model()
elif getattr(args, 'unfreeze_mm_vision_tower', False):
self.load_model()
else:
self.cfg_only = SiglipVisionConfig.from_pretrained(self.vision_tower_name)
def load_model(self, device_map=None):
if self.is_loaded:
print('{} is already loaded, `load_model` called again, skipping.'.format(self.vision_tower_name))
return
self.image_processor = AutoImageProcessor.from_pretrained("google/siglip-so400m-patch14-384")
self.vision_tower = SiglipVisionModel.from_pretrained(self.vision_tower_name, device_map=device_map, _flash_attn_2_enabled=True)
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]
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
@property
def num_patches_per_side(self):
return self.image_processor.size['width'] // self.config.patch_size
@property
def num_patches(self):
return (self.image_processor.size['width'] // self.config.patch_size) ** 2
|