Spaces:
Runtime error
Runtime error
| from transformers.models.qwen2_vl.configuration_qwen2_vl import Qwen2VLConfig, Qwen2VLVisionConfig | |
| from univa.models.configuration_univa_denoise_tower import UnivaDenoiseTowerConfig | |
| from typing import Optional | |
| class UnivaQwen2VLConfig(Qwen2VLConfig): | |
| model_type = "univa_qwen2vl" | |
| sub_configs = { | |
| "vision_config": Qwen2VLVisionConfig, | |
| # "vision_tower": UnivaQwen2VLVisionTowerConfig, | |
| "denoise_tower": UnivaDenoiseTowerConfig, | |
| } | |
| def __init__( | |
| self, | |
| # vision_tower: UnivaQwen2VLVisionTowerConfig = None, | |
| denoise_tower: UnivaDenoiseTowerConfig = None, | |
| image_token_length: Optional[int] = None, | |
| shortcut_image_embeds: bool = False, | |
| shortcut_image_embeds_scale: float = 0.5, | |
| shortcut_projector_type: Optional[str] = "mlp2x_gelu", | |
| **kwargs, | |
| ): | |
| super().__init__(**kwargs) | |
| self.image_token_length = image_token_length | |
| self.shortcut_image_embeds = shortcut_image_embeds | |
| self.shortcut_image_embeds_scale = shortcut_image_embeds_scale | |
| if not shortcut_image_embeds: | |
| shortcut_projector_type = None | |
| self.shortcut_projector_type = shortcut_projector_type | |
| # if isinstance(vision_tower, dict): | |
| # vision_tower["shortcut_projector_type"] = shortcut_projector_type | |
| # self.vision_tower = UnivaQwen2VLVisionTowerConfig(**vision_tower) | |
| # elif vision_tower is None: | |
| # self.vision_tower = UnivaQwen2VLVisionTowerConfig( | |
| # shortcut_projector_type=shortcut_projector_type | |
| # ) | |
| # else: | |
| # self.vision_tower = vision_tower | |
| print(denoise_tower) | |
| if isinstance(denoise_tower, dict): | |
| denoise_tower["input_hidden_size"] = self.hidden_size | |
| self.denoise_tower = UnivaDenoiseTowerConfig(**denoise_tower) | |
| elif denoise_tower is None: | |
| self.denoise_tower = UnivaDenoiseTowerConfig( | |
| input_hidden_size=self.hidden_size | |
| ) | |
| else: | |
| self.denoise_tower = denoise_tower | |