iddlmZddlmZddlmZmZddlmZm Z eje Z edeGd d eZ edeGd d eZd d gZy ))strict)PreTrainedConfig)auto_docstringlogging)CONFIG_MAPPING AutoConfigz$HuggingFaceTB/SmolVLM2-2.2B-Instruct) checkpointceZdZUdZdZdZdZeed<dZ eed<dZ eed <d Z eed <d Z eed <dZ eeezeeefzed<dZeeezeeefzed<dZeed<dZeed<dZeezed<dZeed<y)SmolVLMVisionConfiga Example: ```python >>> from transformers.models.smolvlm.modeling_smolvlm import SmolVLMVisionTransformer >>> from transformers.models.smolvlm.configuration_smolvlm import SmolVLMVisionConfig >>> # Initializing a SmolVLMVisionConfig with google/siglip-so400m-patch14-384 style configuration >>> configuration = SmolVLMVisionConfig() >>> # Initializing a SmolVLMVisionTransformer (with random weights) from the google/siglip-so400m-patch14-384 style configuration >>> model = SmolVLMVisionTransformer(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```smolvlm_vision vision_configi hidden_sizei intermediate_size num_hidden_layersnum_attention_headsr num_channels image_size patch_sizegelu_pytorch_tanh hidden_actgư>layer_norm_epsgattention_dropoutg{Gz?initializer_rangeN)__name__ __module__ __qualname____doc__ model_typebase_config_keyrint__annotations__rrrrrlisttuplerrstrrfloatrr/var/www/vps2.regionflexible.com/Desarrollo/venv/lib/python3.12/site-packages/transformers/models/smolvlm/configuration_smolvlm.pyr r !s ""J%OK!s!s!!L#47Jd3i%S/1746Jd3i%S/16)J) NE %(us{(#u#r-r ceZdZUdZdZeedZdZe e d<dZ e e d<dZ e e d <d Zeezd ze d <d Zeezd ze d <d Ze e d<dZe d ze d<fdZxZS) SmolVLMConfiga scale_factor (`int`, *optional*, defaults to 2): The scale factor for the image encoder. Example: ```python >>> from transformers import SmolVLMModel, SmolVLMConfig >>> # Initializing configuration >>> configuration = SmolVLMConfig() >>> # Initializing a model from the configuration >>> model = SmolVLMModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```smolvlm) text_configrT use_cacheiimage_token_idFtie_word_embeddingsNrr2r scale_factori pad_token_idc J|j%t|_tjdn4t |jt rtdi|j|_t |j t rT|j jdd|j d<t|j ddi|j |_n@|j 4tjdtdd|j|_t|,di|y)Nz2vision_config is None, using default vision configr$llamaz4text_config is None, using default Llama text configgh㈵>) rms_norm_epsr7r,) rr loggerinfo isinstancedictr2getr r7super __post_init__)selfkwargs __class__s r.rAzSmolVLMConfig.__post_init__bs    %!4!6D  KKL M **D 1!4!Jt7I7I!JD  d&& --1-=-=-A-A,PW-XD  \ *-d.>.>|.LMaPTP`P`aD     % KKN O-g6!!.. D  ''r-)r r!r"r#r$r r sub_configsr3boolr'r4r&r5rr>rr2r6r7rA __classcell__)rDs@r.r0r0Es  J",?RSKIt NC %%48M4**T1826K((4/6L#&L#*&((r-r0N)huggingface_hub.dataclassesrconfiguration_utilsrutilsrrautor r get_loggerr r;r r0__all__r,r-r.rNs./3,-   H %AB$*$C$DAB,($,(C,(^ !/ 2r-