| from typing import Literal |
|
|
| from huggingface_hub.dataclasses import strict |
| from transformers import AutoConfig |
| from transformers.configuration_utils import PreTrainedConfig, SubConfigSpec |
|
|
|
|
| @strict |
| class UltraVBertConfig(PreTrainedConfig): |
| model_type = "ultravbert" |
| sub_configs_defaults = { |
| "text_config": SubConfigSpec(config_class=AutoConfig, model_type="ultrabert"), |
| "vision_config": SubConfigSpec(config_class=AutoConfig, model_type="siglip_vision_model"), |
| } |
|
|
| text_config: PreTrainedConfig | dict | None = None |
| vision_config: PreTrainedConfig | dict | None = None |
|
|
| image_token_id: int = 49191 |
| pixel_shuffle_factor: int = 4 |
| initializer_range: float = 0.02 |
| initializer_cutoff_factor: float = 2.0 |
| classifier_pooling: Literal["cls", "mean"] = "cls" |
| classifier_dropout: float | int = 0.0 |
| tie_word_embeddings: bool = False |
| connector_norm: bool = True |
|
|
|
|
| __all__ = ["UltraVBertConfig"] |
|
|
