Aluode/PerceptionLabPortable
0
1# coding=utf-82# Copyright 2018 The HuggingFace Inc. team.3#4# Licensed under the Apache License, Version 2.0 (the "License");5# you may not use this file except in compliance with the License.6# You may obtain a copy of the License at7#8# http://www.apache.org/licenses/LICENSE-2.09#10# Unless required by applicable law or agreed to in writing, software11# distributed under the License is distributed on an "AS IS" BASIS,12# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.13# See the License for the specific language governing permissions and14# limitations under the License.15"""Auto Model class."""16 17import os18import warnings19from collections import OrderedDict20from typing import TYPE_CHECKING, Union21 22from ...utils import logging23from .auto_factory import (24 _BaseAutoBackboneClass,25 _BaseAutoModelClass,26 _LazyAutoMapping,27 auto_class_update,28)29from .configuration_auto import CONFIG_MAPPING_NAMES30 31 32if TYPE_CHECKING:33 from ...generation import GenerationMixin34 from ...modeling_utils import PreTrainedModel35 36 # class for better type annotations37 class _BaseModelWithGenerate(PreTrainedModel, GenerationMixin):38 pass39 40 41logger = logging.get_logger(__name__)42 43MODEL_MAPPING_NAMES = OrderedDict(44 [45 # Base model mapping46 ("aimv2", "Aimv2Model"),47 ("aimv2_vision_model", "Aimv2VisionModel"),48 ("albert", "AlbertModel"),49 ("align", "AlignModel"),50 ("altclip", "AltCLIPModel"),51 ("apertus", "ApertusModel"),52 ("arcee", "ArceeModel"),53 ("aria", "AriaModel"),54 ("aria_text", "AriaTextModel"),55 ("audio-spectrogram-transformer", "ASTModel"),56 ("autoformer", "AutoformerModel"),57 ("aya_vision", "AyaVisionModel"),58 ("bamba", "BambaModel"),59 ("bark", "BarkModel"),60 ("bart", "BartModel"),61 ("beit", "BeitModel"),62 ("bert", "BertModel"),63 ("bert-generation", "BertGenerationEncoder"),64 ("big_bird", "BigBirdModel"),65 ("bigbird_pegasus", "BigBirdPegasusModel"),66 ("biogpt", "BioGptModel"),67 ("bit", "BitModel"),68 ("bitnet", "BitNetModel"),69 ("blenderbot", "BlenderbotModel"),70 ("blenderbot-small", "BlenderbotSmallModel"),71 ("blip", "BlipModel"),72 ("blip-2", "Blip2Model"),73 ("blip_2_qformer", "Blip2QFormerModel"),74 ("bloom", "BloomModel"),75 ("blt", "BltModel"),76 ("bridgetower", "BridgeTowerModel"),77 ("bros", "BrosModel"),78 ("camembert", "CamembertModel"),79 ("canine", "CanineModel"),80 ("chameleon", "ChameleonModel"),81 ("chinese_clip", "ChineseCLIPModel"),82 ("chinese_clip_vision_model", "ChineseCLIPVisionModel"),83 ("clap", "ClapModel"),84 ("clip", "CLIPModel"),85 ("clip_text_model", "CLIPTextModel"),86 ("clip_vision_model", "CLIPVisionModel"),87 ("clipseg", "CLIPSegModel"),88 ("clvp", "ClvpModelForConditionalGeneration"),89 ("code_llama", "LlamaModel"),90 ("codegen", "CodeGenModel"),91 ("cohere", "CohereModel"),92 ("cohere2", "Cohere2Model"),93 ("cohere2_vision", "Cohere2VisionModel"),94 ("conditional_detr", "ConditionalDetrModel"),95 ("convbert", "ConvBertModel"),96 ("convnext", "ConvNextModel"),97 ("convnextv2", "ConvNextV2Model"),98 ("cpmant", "CpmAntModel"),99 ("csm", "CsmForConditionalGeneration"),100 ("ctrl", "CTRLModel"),101 ("cvt", "CvtModel"),102 ("d_fine", "DFineModel"),103 ("dab-detr", "DabDetrModel"),104 ("dac", "DacModel"),105 ("data2vec-audio", "Data2VecAudioModel"),106 ("data2vec-text", "Data2VecTextModel"),107 ("data2vec-vision", "Data2VecVisionModel"),108 ("dbrx", "DbrxModel"),109 ("deberta", "DebertaModel"),110 ("deberta-v2", "DebertaV2Model"),111 ("decision_transformer", "DecisionTransformerModel"),112 ("deepseek_v2", "DeepseekV2Model"),113 ("deepseek_v3", "DeepseekV3Model"),114 ("deepseek_vl", "DeepseekVLModel"),115 ("deepseek_vl_hybrid", "DeepseekVLHybridModel"),116 ("deformable_detr", "DeformableDetrModel"),117 ("deit", "DeiTModel"),118 ("depth_pro", "DepthProModel"),119 ("deta", "DetaModel"),120 ("detr", "DetrModel"),121 ("dia", "DiaModel"),122 ("diffllama", "DiffLlamaModel"),123 ("dinat", "DinatModel"),124 ("dinov2", "Dinov2Model"),125 ("dinov2_with_registers", "Dinov2WithRegistersModel"),126 ("dinov3_convnext", "DINOv3ConvNextModel"),127 ("dinov3_vit", "DINOv3ViTModel"),128 ("distilbert", "DistilBertModel"),129 ("doge", "DogeModel"),130 ("donut-swin", "DonutSwinModel"),131 ("dots1", "Dots1Model"),132 ("dpr", "DPRQuestionEncoder"),133 ("dpt", "DPTModel"),134 ("edgetam", "EdgeTamModel"),135 ("edgetam_video", "EdgeTamVideoModel"),136 ("edgetam_vision_model", "EdgeTamVisionModel"),137 ("efficientformer", "EfficientFormerModel"),138 ("efficientloftr", "EfficientLoFTRModel"),139 ("efficientnet", "EfficientNetModel"),140 ("electra", "ElectraModel"),141 ("emu3", "Emu3Model"),142 ("encodec", "EncodecModel"),143 ("ernie", "ErnieModel"),144 ("ernie4_5", "Ernie4_5Model"),145 ("ernie4_5_moe", "Ernie4_5_MoeModel"),146 ("ernie_m", "ErnieMModel"),147 ("esm", "EsmModel"),148 ("evolla", "EvollaModel"),149 ("exaone4", "Exaone4Model"),150 ("falcon", "FalconModel"),151 ("falcon_h1", "FalconH1Model"),152 ("falcon_mamba", "FalconMambaModel"),153 ("fastspeech2_conformer", "FastSpeech2ConformerModel"),154 ("fastspeech2_conformer_with_hifigan", "FastSpeech2ConformerWithHifiGan"),155 ("flaubert", "FlaubertModel"),156 ("flava", "FlavaModel"),157 ("flex_olmo", "FlexOlmoModel"),158 ("florence2", "Florence2Model"),159 ("fnet", "FNetModel"),160 ("focalnet", "FocalNetModel"),161 ("fsmt", "FSMTModel"),162 ("funnel", ("FunnelModel", "FunnelBaseModel")),163 ("fuyu", "FuyuModel"),164 ("gemma", "GemmaModel"),165 ("gemma2", "Gemma2Model"),166 ("gemma3", "Gemma3Model"),167 ("gemma3_text", "Gemma3TextModel"),168 ("gemma3n", "Gemma3nModel"),169 ("gemma3n_audio", "Gemma3nAudioEncoder"),170 ("gemma3n_text", "Gemma3nTextModel"),171 ("gemma3n_vision", "TimmWrapperModel"),172 ("git", "GitModel"),173 ("glm", "GlmModel"),174 ("glm4", "Glm4Model"),175 ("glm4_moe", "Glm4MoeModel"),176 ("glm4v", "Glm4vModel"),177 ("glm4v_moe", "Glm4vMoeModel"),178 ("glm4v_moe_text", "Glm4vMoeTextModel"),179 ("glm4v_text", "Glm4vTextModel"),180 ("glpn", "GLPNModel"),181 ("got_ocr2", "GotOcr2Model"),182 ("gpt-sw3", "GPT2Model"),183 ("gpt2", "GPT2Model"),184 ("gpt_bigcode", "GPTBigCodeModel"),185 ("gpt_neo", "GPTNeoModel"),186 ("gpt_neox", "GPTNeoXModel"),187 ("gpt_neox_japanese", "GPTNeoXJapaneseModel"),188 ("gpt_oss", "GptOssModel"),189 ("gptj", "GPTJModel"),190 ("gptsan-japanese", "GPTSanJapaneseForConditionalGeneration"),191 ("granite", "GraniteModel"),192 ("granitemoe", "GraniteMoeModel"),193 ("granitemoehybrid", "GraniteMoeHybridModel"),194 ("granitemoeshared", "GraniteMoeSharedModel"),195 ("graphormer", "GraphormerModel"),196 ("grounding-dino", "GroundingDinoModel"),197 ("groupvit", "GroupViTModel"),198 ("helium", "HeliumModel"),199 ("hgnet_v2", "HGNetV2Backbone"),200 ("hiera", "HieraModel"),201 ("hubert", "HubertModel"),202 ("hunyuan_v1_dense", "HunYuanDenseV1Model"),203 ("hunyuan_v1_moe", "HunYuanMoEV1Model"),204 ("ibert", "IBertModel"),205 ("idefics", "IdeficsModel"),206 ("idefics2", "Idefics2Model"),207 ("idefics3", "Idefics3Model"),208 ("idefics3_vision", "Idefics3VisionTransformer"),209 ("ijepa", "IJepaModel"),210 ("imagegpt", "ImageGPTModel"),211 ("informer", "InformerModel"),212 ("instructblip", "InstructBlipModel"),213 ("instructblipvideo", "InstructBlipVideoModel"),214 ("internvl", "InternVLModel"),215 ("internvl_vision", "InternVLVisionModel"),216 ("jamba", "JambaModel"),217 ("janus", "JanusModel"),218 ("jetmoe", "JetMoeModel"),219 ("jukebox", "JukeboxModel"),220 ("kosmos-2", "Kosmos2Model"),221 ("kosmos-2.5", "Kosmos2_5Model"),222 ("kyutai_speech_to_text", "KyutaiSpeechToTextModel"),223 ("layoutlm", "LayoutLMModel"),224 ("layoutlmv2", "LayoutLMv2Model"),225 ("layoutlmv3", "LayoutLMv3Model"),226 ("led", "LEDModel"),227 ("levit", "LevitModel"),228 ("lfm2", "Lfm2Model"),229 ("lfm2_vl", "Lfm2VlModel"),230 ("lightglue", "LightGlueForKeypointMatching"),231 ("lilt", "LiltModel"),232 ("llama", "LlamaModel"),233 ("llama4", "Llama4ForConditionalGeneration"),234 ("llama4_text", "Llama4TextModel"),235 ("llava", "LlavaModel"),236 ("llava_next", "LlavaNextModel"),237 ("llava_next_video", "LlavaNextVideoModel"),238 ("llava_onevision", "LlavaOnevisionModel"),239 ("longcat_flash", "LongcatFlashModel"),240 ("longformer", "LongformerModel"),241 ("longt5", "LongT5Model"),242 ("luke", "LukeModel"),243 ("lxmert", "LxmertModel"),244 ("m2m_100", "M2M100Model"),245 ("mamba", "MambaModel"),246 ("mamba2", "Mamba2Model"),247 ("marian", "MarianModel"),248 ("markuplm", "MarkupLMModel"),249 ("mask2former", "Mask2FormerModel"),250 ("maskformer", "MaskFormerModel"),251 ("maskformer-swin", "MaskFormerSwinModel"),252 ("mbart", "MBartModel"),253 ("mctct", "MCTCTModel"),254 ("mega", "MegaModel"),255 ("megatron-bert", "MegatronBertModel"),256 ("metaclip_2", "MetaClip2Model"),257 ("mgp-str", "MgpstrForSceneTextRecognition"),258 ("mimi", "MimiModel"),259 ("minimax", "MiniMaxModel"),260 ("ministral", "MinistralModel"),261 ("mistral", "MistralModel"),262 ("mistral3", "Mistral3Model"),263 ("mixtral", "MixtralModel"),264 ("mlcd", "MLCDVisionModel"),265 ("mllama", "MllamaModel"),266 ("mm-grounding-dino", "MMGroundingDinoModel"),267 ("mobilebert", "MobileBertModel"),268 ("mobilenet_v1", "MobileNetV1Model"),269 ("mobilenet_v2", "MobileNetV2Model"),270 ("mobilevit", "MobileViTModel"),271 ("mobilevitv2", "MobileViTV2Model"),272 ("modernbert", "ModernBertModel"),273 ("modernbert-decoder", "ModernBertDecoderModel"),274 ("moonshine", "MoonshineModel"),275 ("moshi", "MoshiModel"),276 ("mpnet", "MPNetModel"),277 ("mpt", "MptModel"),278 ("mra", "MraModel"),279 ("mt5", "MT5Model"),280 ("musicgen", "MusicgenModel"),281 ("musicgen_melody", "MusicgenMelodyModel"),282 ("mvp", "MvpModel"),283 ("nat", "NatModel"),284 ("nemotron", "NemotronModel"),285 ("nezha", "NezhaModel"),286 ("nllb-moe", "NllbMoeModel"),287 ("nystromformer", "NystromformerModel"),288 ("olmo", "OlmoModel"),289 ("olmo2", "Olmo2Model"),290 ("olmo3", "Olmo3Model"),291 ("olmoe", "OlmoeModel"),292 ("omdet-turbo", "OmDetTurboForObjectDetection"),293 ("oneformer", "OneFormerModel"),294 ("open-llama", "OpenLlamaModel"),295 ("openai-gpt", "OpenAIGPTModel"),296 ("opt", "OPTModel"),297 ("ovis2", "Ovis2Model"),298 ("owlv2", "Owlv2Model"),299 ("owlvit", "OwlViTModel"),300 ("paligemma", "PaliGemmaModel"),301 ("parakeet_ctc", "ParakeetForCTC"),302 ("parakeet_encoder", "ParakeetEncoder"),303 ("patchtsmixer", "PatchTSMixerModel"),304 ("patchtst", "PatchTSTModel"),305 ("pegasus", "PegasusModel"),306 ("pegasus_x", "PegasusXModel"),307 ("perceiver", "PerceiverModel"),308 ("perception_encoder", "PerceptionEncoder"),309 ("perception_lm", "PerceptionLMModel"),310 ("persimmon", "PersimmonModel"),311 ("phi", "PhiModel"),312 ("phi3", "Phi3Model"),313 ("phi4_multimodal", "Phi4MultimodalModel"),314 ("phimoe", "PhimoeModel"),315 ("pixtral", "PixtralVisionModel"),316 ("plbart", "PLBartModel"),317 ("poolformer", "PoolFormerModel"),318 ("prophetnet", "ProphetNetModel"),319 ("pvt", "PvtModel"),320 ("pvt_v2", "PvtV2Model"),321 ("qdqbert", "QDQBertModel"),322 ("qwen2", "Qwen2Model"),323 ("qwen2_5_vl", "Qwen2_5_VLModel"),324 ("qwen2_5_vl_text", "Qwen2_5_VLTextModel"),325 ("qwen2_audio_encoder", "Qwen2AudioEncoder"),326 ("qwen2_moe", "Qwen2MoeModel"),327 ("qwen2_vl", "Qwen2VLModel"),328 ("qwen2_vl_text", "Qwen2VLTextModel"),329 ("qwen3", "Qwen3Model"),330 ("qwen3_moe", "Qwen3MoeModel"),331 ("qwen3_next", "Qwen3NextModel"),332 ("qwen3_vl", "Qwen3VLModel"),333 ("qwen3_vl_moe", "Qwen3VLMoeModel"),334 ("qwen3_vl_moe_text", "Qwen3VLMoeTextModel"),335 ("qwen3_vl_text", "Qwen3VLTextModel"),336 ("recurrent_gemma", "RecurrentGemmaModel"),337 ("reformer", "ReformerModel"),338 ("regnet", "RegNetModel"),339 ("rembert", "RemBertModel"),340 ("resnet", "ResNetModel"),341 ("retribert", "RetriBertModel"),342 ("roberta", "RobertaModel"),343 ("roberta-prelayernorm", "RobertaPreLayerNormModel"),344 ("roc_bert", "RoCBertModel"),345 ("roformer", "RoFormerModel"),346 ("rt_detr", "RTDetrModel"),347 ("rt_detr_v2", "RTDetrV2Model"),348 ("rwkv", "RwkvModel"),349 ("sam", "SamModel"),350 ("sam2", "Sam2Model"),351 ("sam2_hiera_det_model", "Sam2HieraDetModel"),352 ("sam2_video", "Sam2VideoModel"),353 ("sam2_vision_model", "Sam2VisionModel"),354 ("sam_hq", "SamHQModel"),355 ("sam_hq_vision_model", "SamHQVisionModel"),356 ("sam_vision_model", "SamVisionModel"),357 ("seamless_m4t", "SeamlessM4TModel"),358 ("seamless_m4t_v2", "SeamlessM4Tv2Model"),359 ("seed_oss", "SeedOssModel"),360 ("segformer", "SegformerModel"),361 ("seggpt", "SegGptModel"),362 ("sew", "SEWModel"),363 ("sew-d", "SEWDModel"),364 ("siglip", "SiglipModel"),365 ("siglip2", "Siglip2Model"),366 ("siglip2_vision_model", "Siglip2VisionModel"),367 ("siglip_vision_model", "SiglipVisionModel"),368 ("smollm3", "SmolLM3Model"),369 ("smolvlm", "SmolVLMModel"),370 ("smolvlm_vision", "SmolVLMVisionTransformer"),371 ("speech_to_text", "Speech2TextModel"),372 ("speecht5", "SpeechT5Model"),373 ("splinter", "SplinterModel"),374 ("squeezebert", "SqueezeBertModel"),375 ("stablelm", "StableLmModel"),376 ("starcoder2", "Starcoder2Model"),377 ("swiftformer", "SwiftFormerModel"),378 ("swin", "SwinModel"),379 ("swin2sr", "Swin2SRModel"),380 ("swinv2", "Swinv2Model"),381 ("switch_transformers", "SwitchTransformersModel"),382 ("t5", "T5Model"),383 ("t5gemma", "T5GemmaModel"),384 ("table-transformer", "TableTransformerModel"),385 ("tapas", "TapasModel"),386 ("textnet", "TextNetModel"),387 ("time_series_transformer", "TimeSeriesTransformerModel"),388 ("timesfm", "TimesFmModel"),389 ("timesformer", "TimesformerModel"),390 ("timm_backbone", "TimmBackbone"),391 ("timm_wrapper", "TimmWrapperModel"),392 ("trajectory_transformer", "TrajectoryTransformerModel"),393 ("transfo-xl", "TransfoXLModel"),394 ("tvlt", "TvltModel"),395 ("tvp", "TvpModel"),396 ("udop", "UdopModel"),397 ("umt5", "UMT5Model"),398 ("unispeech", "UniSpeechModel"),399 ("unispeech-sat", "UniSpeechSatModel"),400 ("univnet", "UnivNetModel"),401 ("van", "VanModel"),402 ("vaultgemma", "VaultGemmaModel"),403 ("video_llava", "VideoLlavaModel"),404 ("videomae", "VideoMAEModel"),405 ("vilt", "ViltModel"),406 ("vipllava", "VipLlavaModel"),407 ("vision-text-dual-encoder", "VisionTextDualEncoderModel"),408 ("visual_bert", "VisualBertModel"),409 ("vit", "ViTModel"),410 ("vit_hybrid", "ViTHybridModel"),411 ("vit_mae", "ViTMAEModel"),412 ("vit_msn", "ViTMSNModel"),413 ("vitdet", "VitDetModel"),414 ("vits", "VitsModel"),415 ("vivit", "VivitModel"),416 ("vjepa2", "VJEPA2Model"),417 ("voxtral", "VoxtralForConditionalGeneration"),418 ("voxtral_encoder", "VoxtralEncoder"),419 ("wav2vec2", "Wav2Vec2Model"),420 ("wav2vec2-bert", "Wav2Vec2BertModel"),421 ("wav2vec2-conformer", "Wav2Vec2ConformerModel"),422 ("wavlm", "WavLMModel"),423 ("whisper", "WhisperModel"),424 ("xclip", "XCLIPModel"),425 ("xcodec", "XcodecModel"),426 ("xglm", "XGLMModel"),427 ("xlm", "XLMModel"),428 ("xlm-prophetnet", "XLMProphetNetModel"),429 ("xlm-roberta", "XLMRobertaModel"),430 ("xlm-roberta-xl", "XLMRobertaXLModel"),431 ("xlnet", "XLNetModel"),432 ("xlstm", "xLSTMModel"),433 ("xmod", "XmodModel"),434 ("yolos", "YolosModel"),435 ("yoso", "YosoModel"),436 ("zamba", "ZambaModel"),437 ("zamba2", "Zamba2Model"),438 ]439)440 441MODEL_FOR_PRETRAINING_MAPPING_NAMES = OrderedDict(442 [443 # Model for pre-training mapping444 ("albert", "AlbertForPreTraining"),445 ("bart", "BartForConditionalGeneration"),446 ("bert", "BertForPreTraining"),447 ("big_bird", "BigBirdForPreTraining"),448 ("bloom", "BloomForCausalLM"),449 ("camembert", "CamembertForMaskedLM"),450 ("colpali", "ColPaliForRetrieval"),451 ("colqwen2", "ColQwen2ForRetrieval"),452 ("ctrl", "CTRLLMHeadModel"),453 ("data2vec-text", "Data2VecTextForMaskedLM"),454 ("deberta", "DebertaForMaskedLM"),455 ("deberta-v2", "DebertaV2ForMaskedLM"),456 ("distilbert", "DistilBertForMaskedLM"),457 ("electra", "ElectraForPreTraining"),458 ("ernie", "ErnieForPreTraining"),459 ("evolla", "EvollaForProteinText2Text"),460 ("exaone4", "Exaone4ForCausalLM"),461 ("falcon_mamba", "FalconMambaForCausalLM"),462 ("flaubert", "FlaubertWithLMHeadModel"),463 ("flava", "FlavaForPreTraining"),464 ("florence2", "Florence2ForConditionalGeneration"),465 ("fnet", "FNetForPreTraining"),466 ("fsmt", "FSMTForConditionalGeneration"),467 ("funnel", "FunnelForPreTraining"),468 ("gemma3", "Gemma3ForConditionalGeneration"),469 ("gpt-sw3", "GPT2LMHeadModel"),470 ("gpt2", "GPT2LMHeadModel"),471 ("gpt_bigcode", "GPTBigCodeForCausalLM"),472 ("gptsan-japanese", "GPTSanJapaneseForConditionalGeneration"),473 ("hiera", "HieraForPreTraining"),474 ("ibert", "IBertForMaskedLM"),475 ("idefics", "IdeficsForVisionText2Text"),476 ("idefics2", "Idefics2ForConditionalGeneration"),477 ("idefics3", "Idefics3ForConditionalGeneration"),478 ("janus", "JanusForConditionalGeneration"),479 ("layoutlm", "LayoutLMForMaskedLM"),480 ("llava", "LlavaForConditionalGeneration"),481 ("llava_next", "LlavaNextForConditionalGeneration"),482 ("llava_next_video", "LlavaNextVideoForConditionalGeneration"),483 ("llava_onevision", "LlavaOnevisionForConditionalGeneration"),484 ("longformer", "LongformerForMaskedLM"),485 ("luke", "LukeForMaskedLM"),486 ("lxmert", "LxmertForPreTraining"),487 ("mamba", "MambaForCausalLM"),488 ("mamba2", "Mamba2ForCausalLM"),489 ("mega", "MegaForMaskedLM"),490 ("megatron-bert", "MegatronBertForPreTraining"),491 ("mistral3", "Mistral3ForConditionalGeneration"),492 ("mllama", "MllamaForConditionalGeneration"),493 ("mobilebert", "MobileBertForPreTraining"),494 ("mpnet", "MPNetForMaskedLM"),495 ("mpt", "MptForCausalLM"),496 ("mra", "MraForMaskedLM"),497 ("mvp", "MvpForConditionalGeneration"),498 ("nezha", "NezhaForPreTraining"),499 ("nllb-moe", "NllbMoeForConditionalGeneration"),500 ("openai-gpt", "OpenAIGPTLMHeadModel"),501 ("paligemma", "PaliGemmaForConditionalGeneration"),502 ("qwen2_audio", "Qwen2AudioForConditionalGeneration"),503 ("retribert", "RetriBertModel"),504 ("roberta", "RobertaForMaskedLM"),505 ("roberta-prelayernorm", "RobertaPreLayerNormForMaskedLM"),506 ("roc_bert", "RoCBertForPreTraining"),507 ("rwkv", "RwkvForCausalLM"),508 ("splinter", "SplinterForPreTraining"),509 ("squeezebert", "SqueezeBertForMaskedLM"),510 ("switch_transformers", "SwitchTransformersForConditionalGeneration"),511 ("t5", "T5ForConditionalGeneration"),512 ("t5gemma", "T5GemmaForConditionalGeneration"),513 ("tapas", "TapasForMaskedLM"),514 ("transfo-xl", "TransfoXLLMHeadModel"),515 ("tvlt", "TvltForPreTraining"),516 ("unispeech", "UniSpeechForPreTraining"),517 ("unispeech-sat", "UniSpeechSatForPreTraining"),518 ("video_llava", "VideoLlavaForConditionalGeneration"),519 ("videomae", "VideoMAEForPreTraining"),520 ("vipllava", "VipLlavaForConditionalGeneration"),521 ("visual_bert", "VisualBertForPreTraining"),522 ("vit_mae", "ViTMAEForPreTraining"),523 ("voxtral", "VoxtralForConditionalGeneration"),524 ("wav2vec2", "Wav2Vec2ForPreTraining"),525 ("wav2vec2-conformer", "Wav2Vec2ConformerForPreTraining"),526 ("xlm", "XLMWithLMHeadModel"),527 ("xlm-roberta", "XLMRobertaForMaskedLM"),528 ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),529 ("xlnet", "XLNetLMHeadModel"),530 ("xlstm", "xLSTMForCausalLM"),531 ("xmod", "XmodForMaskedLM"),532 ]533)534 535MODEL_WITH_LM_HEAD_MAPPING_NAMES = OrderedDict(536 [537 # Model with LM heads mapping538 ("albert", "AlbertForMaskedLM"),539 ("bart", "BartForConditionalGeneration"),540 ("bert", "BertForMaskedLM"),541 ("big_bird", "BigBirdForMaskedLM"),542 ("bigbird_pegasus", "BigBirdPegasusForConditionalGeneration"),543 ("blenderbot-small", "BlenderbotSmallForConditionalGeneration"),544 ("bloom", "BloomForCausalLM"),545 ("camembert", "CamembertForMaskedLM"),546 ("codegen", "CodeGenForCausalLM"),547 ("convbert", "ConvBertForMaskedLM"),548 ("cpmant", "CpmAntForCausalLM"),549 ("ctrl", "CTRLLMHeadModel"),550 ("data2vec-text", "Data2VecTextForMaskedLM"),551 ("deberta", "DebertaForMaskedLM"),552 ("deberta-v2", "DebertaV2ForMaskedLM"),553 ("dia", "DiaForConditionalGeneration"),554 ("distilbert", "DistilBertForMaskedLM"),555 ("electra", "ElectraForMaskedLM"),556 ("encoder-decoder", "EncoderDecoderModel"),557 ("ernie", "ErnieForMaskedLM"),558 ("esm", "EsmForMaskedLM"),559 ("exaone4", "Exaone4ForCausalLM"),560 ("falcon_mamba", "FalconMambaForCausalLM"),561 ("flaubert", "FlaubertWithLMHeadModel"),562 ("fnet", "FNetForMaskedLM"),563 ("fsmt", "FSMTForConditionalGeneration"),564 ("funnel", "FunnelForMaskedLM"),565 ("git", "GitForCausalLM"),566 ("gpt-sw3", "GPT2LMHeadModel"),567 ("gpt2", "GPT2LMHeadModel"),568 ("gpt_bigcode", "GPTBigCodeForCausalLM"),569 ("gpt_neo", "GPTNeoForCausalLM"),570 ("gpt_neox", "GPTNeoXForCausalLM"),571 ("gpt_neox_japanese", "GPTNeoXJapaneseForCausalLM"),572 ("gptj", "GPTJForCausalLM"),573 ("gptsan-japanese", "GPTSanJapaneseForConditionalGeneration"),574 ("ibert", "IBertForMaskedLM"),575 ("layoutlm", "LayoutLMForMaskedLM"),576 ("led", "LEDForConditionalGeneration"),577 ("longformer", "LongformerForMaskedLM"),578 ("longt5", "LongT5ForConditionalGeneration"),579 ("luke", "LukeForMaskedLM"),580 ("m2m_100", "M2M100ForConditionalGeneration"),581 ("mamba", "MambaForCausalLM"),582 ("mamba2", "Mamba2ForCausalLM"),583 ("marian", "MarianMTModel"),584 ("mega", "MegaForMaskedLM"),585 ("megatron-bert", "MegatronBertForCausalLM"),586 ("mobilebert", "MobileBertForMaskedLM"),587 ("moonshine", "MoonshineForConditionalGeneration"),588 ("mpnet", "MPNetForMaskedLM"),589 ("mpt", "MptForCausalLM"),590 ("mra", "MraForMaskedLM"),591 ("mvp", "MvpForConditionalGeneration"),592 ("nezha", "NezhaForMaskedLM"),593 ("nllb-moe", "NllbMoeForConditionalGeneration"),594 ("nystromformer", "NystromformerForMaskedLM"),595 ("openai-gpt", "OpenAIGPTLMHeadModel"),596 ("pegasus_x", "PegasusXForConditionalGeneration"),597 ("plbart", "PLBartForConditionalGeneration"),598 ("pop2piano", "Pop2PianoForConditionalGeneration"),599 ("qdqbert", "QDQBertForMaskedLM"),600 ("reformer", "ReformerModelWithLMHead"),601 ("rembert", "RemBertForMaskedLM"),602 ("roberta", "RobertaForMaskedLM"),603 ("roberta-prelayernorm", "RobertaPreLayerNormForMaskedLM"),604 ("roc_bert", "RoCBertForMaskedLM"),605 ("roformer", "RoFormerForMaskedLM"),606 ("rwkv", "RwkvForCausalLM"),607 ("speech_to_text", "Speech2TextForConditionalGeneration"),608 ("squeezebert", "SqueezeBertForMaskedLM"),609 ("switch_transformers", "SwitchTransformersForConditionalGeneration"),610 ("t5", "T5ForConditionalGeneration"),611 ("t5gemma", "T5GemmaForConditionalGeneration"),612 ("tapas", "TapasForMaskedLM"),613 ("transfo-xl", "TransfoXLLMHeadModel"),614 ("wav2vec2", "Wav2Vec2ForMaskedLM"),615 ("whisper", "WhisperForConditionalGeneration"),616 ("xlm", "XLMWithLMHeadModel"),617 ("xlm-roberta", "XLMRobertaForMaskedLM"),618 ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),619 ("xlnet", "XLNetLMHeadModel"),620 ("xmod", "XmodForMaskedLM"),621 ("yoso", "YosoForMaskedLM"),622 ]623)624 625MODEL_FOR_CAUSAL_LM_MAPPING_NAMES = OrderedDict(626 [627 # Model for Causal LM mapping628 ("apertus", "ApertusForCausalLM"),629 ("arcee", "ArceeForCausalLM"),630 ("aria_text", "AriaTextForCausalLM"),631 ("bamba", "BambaForCausalLM"),632 ("bart", "BartForCausalLM"),633 ("bert", "BertLMHeadModel"),634 ("bert-generation", "BertGenerationDecoder"),635 ("big_bird", "BigBirdForCausalLM"),636 ("bigbird_pegasus", "BigBirdPegasusForCausalLM"),637 ("biogpt", "BioGptForCausalLM"),638 ("bitnet", "BitNetForCausalLM"),639 ("blenderbot", "BlenderbotForCausalLM"),640 ("blenderbot-small", "BlenderbotSmallForCausalLM"),641 ("bloom", "BloomForCausalLM"),642 ("blt", "BltForCausalLM"),643 ("camembert", "CamembertForCausalLM"),644 ("code_llama", "LlamaForCausalLM"),645 ("codegen", "CodeGenForCausalLM"),646 ("cohere", "CohereForCausalLM"),647 ("cohere2", "Cohere2ForCausalLM"),648 ("cpmant", "CpmAntForCausalLM"),649 ("ctrl", "CTRLLMHeadModel"),650 ("data2vec-text", "Data2VecTextForCausalLM"),651 ("dbrx", "DbrxForCausalLM"),652 ("deepseek_v2", "DeepseekV2ForCausalLM"),653 ("deepseek_v3", "DeepseekV3ForCausalLM"),654 ("diffllama", "DiffLlamaForCausalLM"),655 ("doge", "DogeForCausalLM"),656 ("dots1", "Dots1ForCausalLM"),657 ("electra", "ElectraForCausalLM"),658 ("emu3", "Emu3ForCausalLM"),659 ("ernie", "ErnieForCausalLM"),660 ("ernie4_5", "Ernie4_5ForCausalLM"),661 ("ernie4_5_moe", "Ernie4_5_MoeForCausalLM"),662 ("exaone4", "Exaone4ForCausalLM"),663 ("falcon", "FalconForCausalLM"),664 ("falcon_h1", "FalconH1ForCausalLM"),665 ("falcon_mamba", "FalconMambaForCausalLM"),666 ("flex_olmo", "FlexOlmoForCausalLM"),667 ("fuyu", "FuyuForCausalLM"),668 ("gemma", "GemmaForCausalLM"),669 ("gemma2", "Gemma2ForCausalLM"),670 ("gemma3", "Gemma3ForConditionalGeneration"),671 ("gemma3_text", "Gemma3ForCausalLM"),672 ("gemma3n", "Gemma3nForConditionalGeneration"),673 ("gemma3n_text", "Gemma3nForCausalLM"),674 ("git", "GitForCausalLM"),675 ("glm", "GlmForCausalLM"),676 ("glm4", "Glm4ForCausalLM"),677 ("glm4_moe", "Glm4MoeForCausalLM"),678 ("got_ocr2", "GotOcr2ForConditionalGeneration"),679 ("gpt-sw3", "GPT2LMHeadModel"),680 ("gpt2", "GPT2LMHeadModel"),681 ("gpt_bigcode", "GPTBigCodeForCausalLM"),682 ("gpt_neo", "GPTNeoForCausalLM"),683 ("gpt_neox", "GPTNeoXForCausalLM"),684 ("gpt_neox_japanese", "GPTNeoXJapaneseForCausalLM"),685 ("gpt_oss", "GptOssForCausalLM"),686 ("gptj", "GPTJForCausalLM"),687 ("granite", "GraniteForCausalLM"),688 ("granitemoe", "GraniteMoeForCausalLM"),689 ("granitemoehybrid", "GraniteMoeHybridForCausalLM"),690 ("granitemoeshared", "GraniteMoeSharedForCausalLM"),691 ("helium", "HeliumForCausalLM"),692 ("hunyuan_v1_dense", "HunYuanDenseV1ForCausalLM"),693 ("hunyuan_v1_moe", "HunYuanMoEV1ForCausalLM"),694 ("jamba", "JambaForCausalLM"),695 ("jetmoe", "JetMoeForCausalLM"),696 ("lfm2", "Lfm2ForCausalLM"),697 ("llama", "LlamaForCausalLM"),698 ("llama4", "Llama4ForCausalLM"),699 ("llama4_text", "Llama4ForCausalLM"),700 ("longcat_flash", "LongcatFlashForCausalLM"),701 ("mamba", "MambaForCausalLM"),702 ("mamba2", "Mamba2ForCausalLM"),703 ("marian", "MarianForCausalLM"),704 ("mbart", "MBartForCausalLM"),705 ("mega", "MegaForCausalLM"),706 ("megatron-bert", "MegatronBertForCausalLM"),707 ("minimax", "MiniMaxForCausalLM"),708 ("ministral", "MinistralForCausalLM"),709 ("mistral", "MistralForCausalLM"),710 ("mixtral", "MixtralForCausalLM"),711 ("mllama", "MllamaForCausalLM"),712 ("modernbert-decoder", "ModernBertDecoderForCausalLM"),713 ("moshi", "MoshiForCausalLM"),714 ("mpt", "MptForCausalLM"),715 ("musicgen", "MusicgenForCausalLM"),716 ("musicgen_melody", "MusicgenMelodyForCausalLM"),717 ("mvp", "MvpForCausalLM"),718 ("nemotron", "NemotronForCausalLM"),719 ("olmo", "OlmoForCausalLM"),720 ("olmo2", "Olmo2ForCausalLM"),721 ("olmo3", "Olmo3ForCausalLM"),722 ("olmoe", "OlmoeForCausalLM"),723 ("open-llama", "OpenLlamaForCausalLM"),724 ("openai-gpt", "OpenAIGPTLMHeadModel"),725 ("opt", "OPTForCausalLM"),726 ("pegasus", "PegasusForCausalLM"),727 ("persimmon", "PersimmonForCausalLM"),728 ("phi", "PhiForCausalLM"),729 ("phi3", "Phi3ForCausalLM"),730 ("phi4_multimodal", "Phi4MultimodalForCausalLM"),731 ("phimoe", "PhimoeForCausalLM"),732 ("plbart", "PLBartForCausalLM"),733 ("prophetnet", "ProphetNetForCausalLM"),734 ("qdqbert", "QDQBertLMHeadModel"),735 ("qwen2", "Qwen2ForCausalLM"),736 ("qwen2_moe", "Qwen2MoeForCausalLM"),737 ("qwen3", "Qwen3ForCausalLM"),738 ("qwen3_moe", "Qwen3MoeForCausalLM"),739 ("qwen3_next", "Qwen3NextForCausalLM"),740 ("recurrent_gemma", "RecurrentGemmaForCausalLM"),741 ("reformer", "ReformerModelWithLMHead"),742 ("rembert", "RemBertForCausalLM"),743 ("roberta", "RobertaForCausalLM"),744 ("roberta-prelayernorm", "RobertaPreLayerNormForCausalLM"),745 ("roc_bert", "RoCBertForCausalLM"),746 ("roformer", "RoFormerForCausalLM"),747 ("rwkv", "RwkvForCausalLM"),748 ("seed_oss", "SeedOssForCausalLM"),749 ("smollm3", "SmolLM3ForCausalLM"),750 ("speech_to_text_2", "Speech2Text2ForCausalLM"),751 ("stablelm", "StableLmForCausalLM"),752 ("starcoder2", "Starcoder2ForCausalLM"),753 ("transfo-xl", "TransfoXLLMHeadModel"),754 ("trocr", "TrOCRForCausalLM"),755 ("vaultgemma", "VaultGemmaForCausalLM"),756 ("whisper", "WhisperForCausalLM"),757 ("xglm", "XGLMForCausalLM"),758 ("xlm", "XLMWithLMHeadModel"),759 ("xlm-prophetnet", "XLMProphetNetForCausalLM"),760 ("xlm-roberta", "XLMRobertaForCausalLM"),761 ("xlm-roberta-xl", "XLMRobertaXLForCausalLM"),762 ("xlnet", "XLNetLMHeadModel"),763 ("xlstm", "xLSTMForCausalLM"),764 ("xmod", "XmodForCausalLM"),765 ("zamba", "ZambaForCausalLM"),766 ("zamba2", "Zamba2ForCausalLM"),767 ]768)769 770MODEL_FOR_IMAGE_MAPPING_NAMES = OrderedDict(771 [772 # Model for Image mapping773 ("aimv2_vision_model", "Aimv2VisionModel"),774 ("beit", "BeitModel"),775 ("bit", "BitModel"),776 ("cohere2_vision", "Cohere2VisionModel"),777 ("conditional_detr", "ConditionalDetrModel"),778 ("convnext", "ConvNextModel"),779 ("convnextv2", "ConvNextV2Model"),780 ("dab-detr", "DabDetrModel"),781 ("data2vec-vision", "Data2VecVisionModel"),782 ("deformable_detr", "DeformableDetrModel"),783 ("deit", "DeiTModel"),784 ("depth_pro", "DepthProModel"),785 ("deta", "DetaModel"),786 ("detr", "DetrModel"),787 ("dinat", "DinatModel"),788 ("dinov2", "Dinov2Model"),789 ("dinov2_with_registers", "Dinov2WithRegistersModel"),790 ("dinov3_convnext", "DINOv3ConvNextModel"),791 ("dinov3_vit", "DINOv3ViTModel"),792 ("dpt", "DPTModel"),793 ("efficientformer", "EfficientFormerModel"),794 ("efficientnet", "EfficientNetModel"),795 ("focalnet", "FocalNetModel"),796 ("glpn", "GLPNModel"),797 ("hiera", "HieraModel"),798 ("ijepa", "IJepaModel"),799 ("imagegpt", "ImageGPTModel"),800 ("levit", "LevitModel"),801 ("llama4", "Llama4VisionModel"),802 ("mlcd", "MLCDVisionModel"),803 ("mllama", "MllamaVisionModel"),804 ("mobilenet_v1", "MobileNetV1Model"),805 ("mobilenet_v2", "MobileNetV2Model"),806 ("mobilevit", "MobileViTModel"),807 ("mobilevitv2", "MobileViTV2Model"),808 ("nat", "NatModel"),809 ("poolformer", "PoolFormerModel"),810 ("pvt", "PvtModel"),811 ("regnet", "RegNetModel"),812 ("resnet", "ResNetModel"),813 ("segformer", "SegformerModel"),814 ("siglip_vision_model", "SiglipVisionModel"),815 ("swiftformer", "SwiftFormerModel"),816 ("swin", "SwinModel"),817 ("swin2sr", "Swin2SRModel"),818 ("swinv2", "Swinv2Model"),819 ("table-transformer", "TableTransformerModel"),820 ("timesformer", "TimesformerModel"),821 ("timm_backbone", "TimmBackbone"),822 ("timm_wrapper", "TimmWrapperModel"),823 ("van", "VanModel"),824 ("videomae", "VideoMAEModel"),825 ("vit", "ViTModel"),826 ("vit_hybrid", "ViTHybridModel"),827 ("vit_mae", "ViTMAEModel"),828 ("vit_msn", "ViTMSNModel"),829 ("vitdet", "VitDetModel"),830 ("vivit", "VivitModel"),831 ("yolos", "YolosModel"),832 ]833)834 835MODEL_FOR_MASKED_IMAGE_MODELING_MAPPING_NAMES = OrderedDict(836 [837 ("deit", "DeiTForMaskedImageModeling"),838 ("focalnet", "FocalNetForMaskedImageModeling"),839 ("swin", "SwinForMaskedImageModeling"),840 ("swinv2", "Swinv2ForMaskedImageModeling"),841 ("vit", "ViTForMaskedImageModeling"),842 ]843)844 845 846MODEL_FOR_CAUSAL_IMAGE_MODELING_MAPPING_NAMES = OrderedDict(847 # Model for Causal Image Modeling mapping848 [849 ("imagegpt", "ImageGPTForCausalImageModeling"),850 ]851)852 853MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING_NAMES = OrderedDict(854 [855 # Model for Image Classification mapping856 ("beit", "BeitForImageClassification"),857 ("bit", "BitForImageClassification"),858 ("clip", "CLIPForImageClassification"),859 ("convnext", "ConvNextForImageClassification"),860 ("convnextv2", "ConvNextV2ForImageClassification"),861 ("cvt", "CvtForImageClassification"),862 ("data2vec-vision", "Data2VecVisionForImageClassification"),863 (864 "deit",865 ("DeiTForImageClassification", "DeiTForImageClassificationWithTeacher"),866 ),867 ("dinat", "DinatForImageClassification"),868 ("dinov2", "Dinov2ForImageClassification"),869 ("dinov2_with_registers", "Dinov2WithRegistersForImageClassification"),870 ("donut-swin", "DonutSwinForImageClassification"),871 (872 "efficientformer",873 (874 "EfficientFormerForImageClassification",875 "EfficientFormerForImageClassificationWithTeacher",876 ),877 ),878 ("efficientnet", "EfficientNetForImageClassification"),879 ("focalnet", "FocalNetForImageClassification"),880 ("hgnet_v2", "HGNetV2ForImageClassification"),881 ("hiera", "HieraForImageClassification"),882 ("ijepa", "IJepaForImageClassification"),883 ("imagegpt", "ImageGPTForImageClassification"),884 (885 "levit",886 ("LevitForImageClassification", "LevitForImageClassificationWithTeacher"),887 ),888 ("metaclip_2", "MetaClip2ForImageClassification"),889 ("mobilenet_v1", "MobileNetV1ForImageClassification"),890 ("mobilenet_v2", "MobileNetV2ForImageClassification"),891 ("mobilevit", "MobileViTForImageClassification"),892 ("mobilevitv2", "MobileViTV2ForImageClassification"),893 ("nat", "NatForImageClassification"),894 (895 "perceiver",896 (897 "PerceiverForImageClassificationLearned",898 "PerceiverForImageClassificationFourier",899 "PerceiverForImageClassificationConvProcessing",900 ),901 ),902 ("poolformer", "PoolFormerForImageClassification"),903 ("pvt", "PvtForImageClassification"),904 ("pvt_v2", "PvtV2ForImageClassification"),905 ("regnet", "RegNetForImageClassification"),906 ("resnet", "ResNetForImageClassification"),907 ("segformer", "SegformerForImageClassification"),908 ("shieldgemma2", "ShieldGemma2ForImageClassification"),909 ("siglip", "SiglipForImageClassification"),910 ("siglip2", "Siglip2ForImageClassification"),911 ("swiftformer", "SwiftFormerForImageClassification"),912 ("swin", "SwinForImageClassification"),913 ("swinv2", "Swinv2ForImageClassification"),914 ("textnet", "TextNetForImageClassification"),915 ("timm_wrapper", "TimmWrapperForImageClassification"),916 ("van", "VanForImageClassification"),917 ("vit", "ViTForImageClassification"),918 ("vit_hybrid", "ViTHybridForImageClassification"),919 ("vit_msn", "ViTMSNForImageClassification"),920 ]921)922 923MODEL_FOR_IMAGE_SEGMENTATION_MAPPING_NAMES = OrderedDict(924 [925 # Do not add new models here, this class will be deprecated in the future.926 # Model for Image Segmentation mapping927 ("detr", "DetrForSegmentation"),928 ]929)930 931MODEL_FOR_SEMANTIC_SEGMENTATION_MAPPING_NAMES = OrderedDict(932 [933 # Model for Semantic Segmentation mapping934 ("beit", "BeitForSemanticSegmentation"),935 ("data2vec-vision", "Data2VecVisionForSemanticSegmentation"),936 ("dpt", "DPTForSemanticSegmentation"),937 ("mobilenet_v2", "MobileNetV2ForSemanticSegmentation"),938 ("mobilevit", "MobileViTForSemanticSegmentation"),939 ("mobilevitv2", "MobileViTV2ForSemanticSegmentation"),940 ("segformer", "SegformerForSemanticSegmentation"),941 ("upernet", "UperNetForSemanticSegmentation"),942 ]943)944 945MODEL_FOR_INSTANCE_SEGMENTATION_MAPPING_NAMES = OrderedDict(946 [947 # Model for Instance Segmentation mapping948 # MaskFormerForInstanceSegmentation can be removed from this mapping in v5949 ("maskformer", "MaskFormerForInstanceSegmentation"),950 ]951)952 953MODEL_FOR_UNIVERSAL_SEGMENTATION_MAPPING_NAMES = OrderedDict(954 [955 # Model for Universal Segmentation mapping956 ("detr", "DetrForSegmentation"),957 ("eomt", "EomtForUniversalSegmentation"),958 ("mask2former", "Mask2FormerForUniversalSegmentation"),959 ("maskformer", "MaskFormerForInstanceSegmentation"),960 ("oneformer", "OneFormerForUniversalSegmentation"),961 ]962)963 964MODEL_FOR_VIDEO_CLASSIFICATION_MAPPING_NAMES = OrderedDict(965 [966 ("timesformer", "TimesformerForVideoClassification"),967 ("videomae", "VideoMAEForVideoClassification"),968 ("vivit", "VivitForVideoClassification"),969 ("vjepa2", "VJEPA2ForVideoClassification"),970 ]971)972 973MODEL_FOR_VISION_2_SEQ_MAPPING_NAMES = OrderedDict(974 [975 ("blip", "BlipForConditionalGeneration"),976 ("blip-2", "Blip2ForConditionalGeneration"),977 ("chameleon", "ChameleonForConditionalGeneration"),978 ("git", "GitForCausalLM"),979 ("idefics2", "Idefics2ForConditionalGeneration"),980 ("idefics3", "Idefics3ForConditionalGeneration"),981 ("instructblip", "InstructBlipForConditionalGeneration"),982 ("instructblipvideo", "InstructBlipVideoForConditionalGeneration"),983 ("kosmos-2", "Kosmos2ForConditionalGeneration"),984 ("kosmos-2.5", "Kosmos2_5ForConditionalGeneration"),985 ("llava", "LlavaForConditionalGeneration"),986 ("llava_next", "LlavaNextForConditionalGeneration"),987 ("llava_next_video", "LlavaNextVideoForConditionalGeneration"),988 ("llava_onevision", "LlavaOnevisionForConditionalGeneration"),989 ("mistral3", "Mistral3ForConditionalGeneration"),990 ("mllama", "MllamaForConditionalGeneration"),991 ("ovis2", "Ovis2ForConditionalGeneration"),992 ("paligemma", "PaliGemmaForConditionalGeneration"),993 ("pix2struct", "Pix2StructForConditionalGeneration"),994 ("qwen2_5_vl", "Qwen2_5_VLForConditionalGeneration"),995 ("qwen2_vl", "Qwen2VLForConditionalGeneration"),996 ("qwen3_vl", "Qwen3VLForConditionalGeneration"),997 ("qwen3_vl_moe", "Qwen3VLMoeForConditionalGeneration"),998 ("video_llava", "VideoLlavaForConditionalGeneration"),999 ("vipllava", "VipLlavaForConditionalGeneration"),1000 ("vision-encoder-decoder", "VisionEncoderDecoderModel"),1001 ]1002)1003 1004MODEL_FOR_RETRIEVAL_MAPPING_NAMES = OrderedDict(1005 [1006 ("colpali", "ColPaliForRetrieval"),1007 ]1008)1009 1010MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES = OrderedDict(1011 [1012 ("aria", "AriaForConditionalGeneration"),1013 ("aya_vision", "AyaVisionForConditionalGeneration"),1014 ("blip", "BlipForConditionalGeneration"),1015 ("blip-2", "Blip2ForConditionalGeneration"),1016 ("chameleon", "ChameleonForConditionalGeneration"),1017 ("cohere2_vision", "Cohere2VisionForConditionalGeneration"),1018 ("deepseek_vl", "DeepseekVLForConditionalGeneration"),1019 ("deepseek_vl_hybrid", "DeepseekVLHybridForConditionalGeneration"),1020 ("emu3", "Emu3ForConditionalGeneration"),1021 ("evolla", "EvollaForProteinText2Text"),1022 ("florence2", "Florence2ForConditionalGeneration"),1023 ("fuyu", "FuyuForCausalLM"),1024 ("gemma3", "Gemma3ForConditionalGeneration"),1025 ("gemma3n", "Gemma3nForConditionalGeneration"),1026 ("git", "GitForCausalLM"),1027 ("glm4v", "Glm4vForConditionalGeneration"),1028 ("glm4v_moe", "Glm4vMoeForConditionalGeneration"),1029 ("got_ocr2", "GotOcr2ForConditionalGeneration"),1030 ("idefics", "IdeficsForVisionText2Text"),1031 ("idefics2", "Idefics2ForConditionalGeneration"),1032 ("idefics3", "Idefics3ForConditionalGeneration"),1033 ("instructblip", "InstructBlipForConditionalGeneration"),1034 ("internvl", "InternVLForConditionalGeneration"),1035 ("janus", "JanusForConditionalGeneration"),1036 ("kosmos-2", "Kosmos2ForConditionalGeneration"),1037 ("kosmos-2.5", "Kosmos2_5ForConditionalGeneration"),1038 ("lfm2_vl", "Lfm2VlForConditionalGeneration"),1039 ("llama4", "Llama4ForConditionalGeneration"),1040 ("llava", "LlavaForConditionalGeneration"),1041 ("llava_next", "LlavaNextForConditionalGeneration"),1042 ("llava_next_video", "LlavaNextVideoForConditionalGeneration"),1043 ("llava_onevision", "LlavaOnevisionForConditionalGeneration"),1044 ("mistral3", "Mistral3ForConditionalGeneration"),1045 ("mllama", "MllamaForConditionalGeneration"),1046 ("ovis2", "Ovis2ForConditionalGeneration"),1047 ("paligemma", "PaliGemmaForConditionalGeneration"),1048 ("perception_lm", "PerceptionLMForConditionalGeneration"),1049 ("pix2struct", "Pix2StructForConditionalGeneration"),1050 ("pixtral", "LlavaForConditionalGeneration"),1051 ("qwen2_5_vl", "Qwen2_5_VLForConditionalGeneration"),1052 ("qwen2_vl", "Qwen2VLForConditionalGeneration"),1053 ("qwen3_vl", "Qwen3VLForConditionalGeneration"),1054 ("qwen3_vl_moe", "Qwen3VLMoeForConditionalGeneration"),1055 ("shieldgemma2", "Gemma3ForConditionalGeneration"),1056 ("smolvlm", "SmolVLMForConditionalGeneration"),1057 ("udop", "UdopForConditionalGeneration"),1058 ("vipllava", "VipLlavaForConditionalGeneration"),1059 ("vision-encoder-decoder", "VisionEncoderDecoderModel"),1060 ]1061)1062 1063MODEL_FOR_MASKED_LM_MAPPING_NAMES = OrderedDict(1064 [1065 # Model for Masked LM mapping1066 ("albert", "AlbertForMaskedLM"),1067 ("bart", "BartForConditionalGeneration"),1068 ("bert", "BertForMaskedLM"),1069 ("big_bird", "BigBirdForMaskedLM"),1070 ("camembert", "CamembertForMaskedLM"),1071 ("convbert", "ConvBertForMaskedLM"),1072 ("data2vec-text", "Data2VecTextForMaskedLM"),1073 ("deberta", "DebertaForMaskedLM"),1074 ("deberta-v2", "DebertaV2ForMaskedLM"),1075 ("distilbert", "DistilBertForMaskedLM"),1076 ("electra", "ElectraForMaskedLM"),1077 ("ernie", "ErnieForMaskedLM"),1078 ("esm", "EsmForMaskedLM"),1079 ("flaubert", "FlaubertWithLMHeadModel"),1080 ("fnet", "FNetForMaskedLM"),1081 ("funnel", "FunnelForMaskedLM"),1082 ("ibert", "IBertForMaskedLM"),1083 ("layoutlm", "LayoutLMForMaskedLM"),1084 ("longformer", "LongformerForMaskedLM"),1085 ("luke", "LukeForMaskedLM"),1086 ("mbart", "MBartForConditionalGeneration"),1087 ("mega", "MegaForMaskedLM"),1088 ("megatron-bert", "MegatronBertForMaskedLM"),1089 ("mobilebert", "MobileBertForMaskedLM"),1090 ("modernbert", "ModernBertForMaskedLM"),1091 ("mpnet", "MPNetForMaskedLM"),1092 ("mra", "MraForMaskedLM"),1093 ("mvp", "MvpForConditionalGeneration"),1094 ("nezha", "NezhaForMaskedLM"),1095 ("nystromformer", "NystromformerForMaskedLM"),1096 ("perceiver", "PerceiverForMaskedLM"),1097 ("qdqbert", "QDQBertForMaskedLM"),1098 ("reformer", "ReformerForMaskedLM"),1099 ("rembert", "RemBertForMaskedLM"),1100 ("roberta", "RobertaForMaskedLM"),1101 ("roberta-prelayernorm", "RobertaPreLayerNormForMaskedLM"),1102 ("roc_bert", "RoCBertForMaskedLM"),1103 ("roformer", "RoFormerForMaskedLM"),1104 ("squeezebert", "SqueezeBertForMaskedLM"),1105 ("tapas", "TapasForMaskedLM"),1106 ("wav2vec2", "Wav2Vec2ForMaskedLM"),1107 ("xlm", "XLMWithLMHeadModel"),1108 ("xlm-roberta", "XLMRobertaForMaskedLM"),1109 ("xlm-roberta-xl", "XLMRobertaXLForMaskedLM"),1110 ("xmod", "XmodForMaskedLM"),1111 ("yoso", "YosoForMaskedLM"),1112 ]1113)1114 1115MODEL_FOR_OBJECT_DETECTION_MAPPING_NAMES = OrderedDict(1116 [1117 # Model for Object Detection mapping1118 ("conditional_detr", "ConditionalDetrForObjectDetection"),1119 ("d_fine", "DFineForObjectDetection"),1120 ("dab-detr", "DabDetrForObjectDetection"),1121 ("deformable_detr", "DeformableDetrForObjectDetection"),1122 ("deta", "DetaForObjectDetection"),1123 ("detr", "DetrForObjectDetection"),1124 ("rt_detr", "RTDetrForObjectDetection"),1125 ("rt_detr_v2", "RTDetrV2ForObjectDetection"),1126 ("table-transformer", "TableTransformerForObjectDetection"),1127 ("yolos", "YolosForObjectDetection"),1128 ]1129)1130 1131MODEL_FOR_ZERO_SHOT_OBJECT_DETECTION_MAPPING_NAMES = OrderedDict(1132 [1133 # Model for Zero Shot Object Detection mapping1134 ("grounding-dino", "GroundingDinoForObjectDetection"),1135 ("mm-grounding-dino", "MMGroundingDinoForObjectDetection"),1136 ("omdet-turbo", "OmDetTurboForObjectDetection"),1137 ("owlv2", "Owlv2ForObjectDetection"),1138 ("owlvit", "OwlViTForObjectDetection"),1139 ]1140)1141 1142MODEL_FOR_DEPTH_ESTIMATION_MAPPING_NAMES = OrderedDict(1143 [1144 # Model for depth estimation mapping1145 ("depth_anything", "DepthAnythingForDepthEstimation"),1146 ("depth_pro", "DepthProForDepthEstimation"),1147 ("dpt", "DPTForDepthEstimation"),1148 ("glpn", "GLPNForDepthEstimation"),1149 ("prompt_depth_anything", "PromptDepthAnythingForDepthEstimation"),1150 ("zoedepth", "ZoeDepthForDepthEstimation"),1151 ]1152)1153MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING_NAMES = OrderedDict(1154 [1155 # Model for Seq2Seq Causal LM mapping1156 ("bart", "BartForConditionalGeneration"),1157 ("bigbird_pegasus", "BigBirdPegasusForConditionalGeneration"),1158 ("blenderbot", "BlenderbotForConditionalGeneration"),1159 ("blenderbot-small", "BlenderbotSmallForConditionalGeneration"),1160 ("encoder-decoder", "EncoderDecoderModel"),1161 ("fsmt", "FSMTForConditionalGeneration"),1162 ("gptsan-japanese", "GPTSanJapaneseForConditionalGeneration"),1163 ("granite_speech", "GraniteSpeechForConditionalGeneration"),1164 ("led", "LEDForConditionalGeneration"),1165 ("longt5", "LongT5ForConditionalGeneration"),1166 ("m2m_100", "M2M100ForConditionalGeneration"),1167 ("marian", "MarianMTModel"),1168 ("mbart", "MBartForConditionalGeneration"),1169 ("mt5", "MT5ForConditionalGeneration"),1170 ("mvp", "MvpForConditionalGeneration"),1171 ("nllb-moe", "NllbMoeForConditionalGeneration"),1172 ("pegasus", "PegasusForConditionalGeneration"),1173 ("pegasus_x", "PegasusXForConditionalGeneration"),1174 ("plbart", "PLBartForConditionalGeneration"),1175 ("prophetnet", "ProphetNetForConditionalGeneration"),1176 ("qwen2_audio", "Qwen2AudioForConditionalGeneration"),1177 ("seamless_m4t", "SeamlessM4TForTextToText"),1178 ("seamless_m4t_v2", "SeamlessM4Tv2ForTextToText"),1179 ("switch_transformers", "SwitchTransformersForConditionalGeneration"),1180 ("t5", "T5ForConditionalGeneration"),1181 ("t5gemma", "T5GemmaForConditionalGeneration"),1182 ("umt5", "UMT5ForConditionalGeneration"),1183 ("voxtral", "VoxtralForConditionalGeneration"),1184 ("xlm-prophetnet", "XLMProphetNetForConditionalGeneration"),1185 ]1186)1187 1188MODEL_FOR_SPEECH_SEQ_2_SEQ_MAPPING_NAMES = OrderedDict(1189 [1190 ("dia", "DiaForConditionalGeneration"),1191 ("granite_speech", "GraniteSpeechForConditionalGeneration"),1192 ("kyutai_speech_to_text", "KyutaiSpeechToTextForConditionalGeneration"),1193 ("moonshine", "MoonshineForConditionalGeneration"),1194 ("pop2piano", "Pop2PianoForConditionalGeneration"),1195 ("seamless_m4t", "SeamlessM4TForSpeechToText"),1196 ("seamless_m4t_v2", "SeamlessM4Tv2ForSpeechToText"),1197 ("speech-encoder-decoder", "SpeechEncoderDecoderModel"),1198 ("speech_to_text", "Speech2TextForConditionalGeneration"),1199 ("speecht5", "SpeechT5ForSpeechToText"),1200 ("whisper", "WhisperForConditionalGeneration"),