- sections: - local: index title: ๐Ÿค— Transformers - local: installation title: ์„ค์น˜๋ฐฉ๋ฒ• - local: quicktour title: ๋‘˜๋Ÿฌ๋ณด๊ธฐ title: ์‹œ์ž‘ํ•˜๊ธฐ - isExpanded: false sections: - sections: - local: models title: ๋ชจ๋ธ ๋กœ๋“œํ•˜๊ธฐ - local: custom_models title: ์‚ฌ์šฉ์ž ์ •์˜ ๋ชจ๋ธ ๊ณต์œ ํ•˜๊ธฐ - local: how_to_hack_models title: ๋ชจ๋ธ ๊ตฌ์„ฑ ์š”์†Œ ๋งž์ถค ์„ค์ •ํ•˜๊ธฐ - local: model_sharing title: ๋งŒ๋“  ๋ชจ๋ธ ๊ณต์œ ํ•˜๊ธฐ - local: modular_transformers title: transformers์—์„œ์˜ ๋ชจ๋“ˆ์„ฑ - local: add_new_model title: ๐Ÿค— Transformers์— ์ƒˆ๋กœ์šด ๋ชจ๋ธ์„ ์ถ”๊ฐ€ํ•˜๋Š” ๋ฐฉ๋ฒ• - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Documenting a model - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Customizing attention function title: ๋ชจ๋ธ - sections: - local: fast_tokenizers title: ๐Ÿค— Tokenizers ๋ผ์ด๋ธŒ๋Ÿฌ๋ฆฌ์—์„œ ํ† ํฌ๋‚˜์ด์ € ์‚ฌ์šฉํ•˜๊ธฐ - local: image_processors title: ์ด๋ฏธ์ง€ ํ”„๋กœ์„ธ์„œ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Video processors - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Backbones - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Feature extractors - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Processors - local: tokenizer_summary title: ํ† ํฌ๋‚˜์ด์ € ์š”์•ฝ - local: pad_truncation title: ํŒจ๋”ฉ๊ณผ ์ž˜๋ผ๋‚ด๊ธฐ title: ์ „์ฒ˜๋ฆฌ๊ธฐ(Preprocessors) title: Base classes - isExpanded: false sections: - sections: - local: pipeline_tutorial title: Pipeline์œผ๋กœ ์ถ”๋ก ํ•˜๊ธฐ - local: pipeline_gradio title: ๋จธ์‹ ๋Ÿฌ๋‹ ์•ฑ - local: pipeline_webserver title: ์ถ”๋ก  ์›น ์„œ๋ฒ„๋ฅผ ์œ„ํ•œ ํŒŒ์ดํ”„๋ผ์ธ - local: add_new_pipeline title: ์–ด๋–ป๊ฒŒ ๐Ÿค— Transformers์— ํŒŒ์ดํ”„๋ผ์ธ์„ ์ถ”๊ฐ€ํ•˜๋‚˜์š”? title: ํŒŒ์ดํ”„๋ผ์ธ API - sections: - local: llm_tutorial title: ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ๋กœ ์ƒ์„ฑํ•˜๊ธฐ - local: generation_strategies title: ํ…์ŠคํŠธ ์ƒ์„ฑ ์ „๋žต ์‚ฌ์šฉ์ž ์ •์˜ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Generation features - local: tasks/prompting title: ๋Œ€๊ทœ๋ชจ ์–ธ์–ด ๋ชจ๋ธ ํ”„๋กฌํ”„ํŒ… ๊ฐ€์ด๋“œ - local: llm_optims title: LLM ์ถ”๋ก  ์ตœ์ ํ™” - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Caching - local: in_translation title: (๋ฒˆ์—ญ์ค‘) KV cache strategies - local: serving title: ๋ชจ๋ธ ์„œ๋น™ํ•˜๊ธฐ - local: llm_tutorial_optimization title: LLM์„ ์ตœ๋Œ€ํ•œ ํ™œ์šฉํ•˜๊ธฐ - local: perplexity title: ๊ณ ์ • ๊ธธ์ด ๋ชจ๋ธ์˜ ํŽ„ํ”Œ๋ ‰์„œํ‹ฐ(Perplexity) title: ๊ฑฐ๋Œ€ ์–ธ์–ด ๋ชจ๋ธ(LLMs) - sections: - local: conversations title: Transformers๋กœ ์ฑ„ํŒ…ํ•˜๊ธฐ - local: chat_templating title: ์ฑ—๋ด‡ ํ…œํ”Œ๋ฆฟ ์ตํžˆ๊ธฐ - local: chat_extras title: Tools ์™€ RAG - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Multimodal templates - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Template writing - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Tools and RAG title: ๋ชจ๋ธ์„ ์‚ฌ์šฉํ•ด ๋Œ€ํ™”ํ•˜๊ธฐ - sections: - local: tiny_agents title: Tiny-Agents CLI ๋ฐ MCP ๋„๊ตฌ title: ์„œ๋น™(Serving) - sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘) torch.compile - local: perf_infer_gpu_one title: ํ•˜๋‚˜์˜ GPU๋ฅผ ํ™œ์šฉํ•œ ์ถ”๋ก  - local: perf_infer_gpu_multi title: (๋ฒˆ์—ญ์ค‘) Distributed inference - local: perf_infer_cpu title: CPU๋กœ ์ถ”๋ก ํ•˜๊ธฐ title: ์ตœ์ ํ™”(Optimization) - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Agents - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Tools title: ์ถ”๋ก (Inference) - isExpanded: false sections: - sections: - local: trainer title: ํŠธ๋ ˆ์ด๋„ˆ(Trainer) - local: training title: ์‚ฌ์ „ ํ•™์Šต๋œ ๋ชจ๋ธ ๋ฏธ์„ธ ์กฐ์ •ํ•˜๊ธฐ - local: optimizers title: ์˜ตํ‹ฐ๋งˆ์ด์ €(Optimizers) - local: hpo_train title: Trainer API๋ฅผ ์‚ฌ์šฉํ•œ ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํƒ์ƒ‰ title: Trainer API - sections: - local: accelerator_selection title: (๋ฒˆ์—ญ์ค‘) Accelerator selection - local: accelerate title: ๐Ÿค— Accelerate๋กœ ๋ถ„์‚ฐ ํ•™์Šต ๊ตฌ์„ฑํ•˜๊ธฐ - local: fsdp title: ์™„์ „ ๋ถ„ํ•  ๋ฐ์ดํ„ฐ ๋ณ‘๋ ฌ ์ฒ˜๋ฆฌ - local: deepspeed title: DeepSpeed - local: debugging title: ๋””๋ฒ„๊น… - local: perf_train_cpu_many title: ๋‹ค์ค‘ CPU์—์„œ ํ•™์Šตํ•˜๊ธฐ - local: perf_train_gpu_many title: ๋‹ค์ค‘ GPU์—์„œ ํ•™์Šต ์ง„ํ–‰ํ•˜๊ธฐ title: ๋ถ„์‚ฐ ํ•™์Šต(Distributed training) - sections: - local: perf_train_gpu_one title: GPU - local: perf_train_cpu title: CPU์—์„œ ํ›ˆ๋ จ - local: perf_train_special title: Apple ์‹ค๋ฆฌ์ฝ˜์—์„œ PyTorch ํ•™์Šต - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Intel Gaudi - local: perf_hardware title: ํ›ˆ๋ จ์šฉ ์‚ฌ์šฉ์ž ๋งž์ถคํ˜• ํ•˜๋“œ์›จ์–ด title: ํ•˜๋“œ์›จ์–ด - local: peft title: ๐Ÿค— PEFT๋กœ ์–ด๋Œ‘ํ„ฐ ๋กœ๋“œ ๋ฐ ํ•™์Šตํ•˜๊ธฐ - local: model_memory_anatomy title: ๋ชจ๋ธ ํ•™์Šต ํ•ด๋ถ€ํ•˜๊ธฐ title: ํ•™์Šต(Training) - isExpanded: false sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Overview - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Selecting a quantization method - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Quantization concepts - local: in_translation title: (๋ฒˆ์—ญ์ค‘) AQLM - local: in_translation title: (๋ฒˆ์—ญ์ค‘) AutoRound - local: quantization/awq title: AWQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) BitNet - local: quantization/bitsandbytes title: bitsandbytes - local: in_translation title: (๋ฒˆ์—ญ์ค‘) compressed-tensors - local: quantization/eetq title: EETQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) FBGEMM - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Fine-grained FP8 - local: gguf title: GGUF ํŒŒ์ผ๋“ค๊ณผ์˜ ์ƒํ˜ธ ์šด์šฉ์„ฑ - local: quantization/gptq title: GPTQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) HIGGS - local: in_translation title: (๋ฒˆ์—ญ์ค‘) HQQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Optimum - local: quantization/quanto title: Quanto - local: quantization/quark title: Quark - local: in_translation title: (๋ฒˆ์—ญ์ค‘) torchao - local: in_translation title: (๋ฒˆ์—ญ์ค‘) SpQR - local: in_translation title: (๋ฒˆ์—ญ์ค‘) VPTQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Contribute title: ์–‘์žํ™”(Quantization) - isExpanded: false sections: - local: serialization title: ONNX๋กœ ๋‚ด๋ณด๋‚ด๊ธฐ - local: executorch title: ExecuTorch title: ๋ฐฐํฌํ™˜๊ฒฝ์— ๋‚ด๋ณด๋‚ด๊ธฐ - isExpanded: false sections: - sections: - sections: - local: tasks/sequence_classification title: ํ…์ŠคํŠธ ๋ถ„๋ฅ˜ - local: tasks/token_classification title: ํ† ํฐ ๋ถ„๋ฅ˜ - local: tasks/question_answering title: ์งˆ์˜ ์‘๋‹ต(Question Answering) - local: tasks/language_modeling title: ์ธ๊ณผ์  ์–ธ์–ด ๋ชจ๋ธ๋ง(Causal language modeling) - local: tasks/masked_language_modeling title: ๋งˆ์Šคํ‚น๋œ ์–ธ์–ด ๋ชจ๋ธ๋ง(Masked language modeling) - local: tasks/translation title: ๋ฒˆ์—ญ - local: tasks/summarization title: ์š”์•ฝ - local: tasks/multiple_choice title: ๊ฐ๊ด€์‹ ๋ฌธ์ œ(Multiple Choice) title: ์ž์—ฐ์–ด์ฒ˜๋ฆฌ - sections: - local: tasks/audio_classification title: ์˜ค๋””์˜ค ๋ถ„๋ฅ˜ - local: tasks/asr title: ์ž๋™ ์Œ์„ฑ ์ธ์‹ title: ์˜ค๋””์˜ค - sections: - local: tasks/image_classification title: ์ด๋ฏธ์ง€ ๋ถ„๋ฅ˜ - local: tasks/semantic_segmentation title: ์ด๋ฏธ์ง€ ์„ธ๊ทธ๋ฉ˜ํ…Œ์ด์…˜ - local: tasks/video_classification title: ๋น„๋””์˜ค ๋ถ„๋ฅ˜ - local: tasks/object_detection title: ๊ฐ์ฒด ํƒ์ง€(Object detection) - local: tasks/zero_shot_object_detection title: ์ œ๋กœ์ƒท(zero-shot) ๊ฐ์ฒด ํƒ์ง€ - local: tasks/zero_shot_image_classification title: ์ œ๋กœ์ƒท(zero-shot) ์ด๋ฏธ์ง€ ๋ถ„๋ฅ˜ - local: tasks/monocular_depth_estimation title: ๋‹จ์ผ ์˜์ƒ ๊ธฐ๋ฐ˜ ๊นŠ์ด ์ถ”์ • - local: tasks/image_to_image title: Image-to-Image - local: tasks/image_feature_extraction title: ์ด๋ฏธ์ง€ ํŠน์ง• ์ถ”์ถœ - local: tasks/mask_generation title: ๋งˆ์Šคํฌ ์ƒ์„ฑ - local: tasks/keypoint_detection title: ํ‚คํฌ์ธํŠธ ํƒ์ง€ - local: tasks/knowledge_distillation_for_image_classification title: ์ปดํ“จํ„ฐ ๋น„์ „(์ด๋ฏธ์ง€ ๋ถ„๋ฅ˜)๋ฅผ ์œ„ํ•œ ์ง€์‹ ์ฆ๋ฅ˜(knowledge distillation) title: ์ปดํ“จํ„ฐ ๋น„์ „ - sections: - local: tasks/image_captioning title: ์ด๋ฏธ์ง€ ์บก์…”๋‹ - local: tasks/document_question_answering title: ๋ฌธ์„œ ์งˆ์˜ ์‘๋‹ต(Document Question Answering) - local: tasks/visual_question_answering title: ์‹œ๊ฐ์  ์งˆ์˜์‘๋‹ต (Visual Question Answering) - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Text to speech - local: tasks/idefics title: IDEFICS๋ฅผ ์ด์šฉํ•œ ์ด๋ฏธ์ง€ ์ž‘์—… - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Image-text-to-text - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Video-text-to-text - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Visual Document Retrieval title: ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ title: ํƒœ์Šคํฌ ๋ ˆ์‹œํ”ผ - local: run_scripts title: ์Šคํฌ๋ฆฝํŠธ๋กœ ํ•™์Šตํ•˜๊ธฐ - local: glossary title: Glossary - local: philosophy title: ์ด๋…๊ณผ ๋ชฉํ‘œ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Notebooks with examples - local: community title: ์ปค๋ฎค๋‹ˆํ‹ฐ ๋ฆฌ์†Œ์Šค - local: troubleshooting title: ๋ฌธ์ œ ํ•ด๊ฒฐ - local: gguf title: GGUF ํŒŒ์ผ๋“ค๊ณผ์˜ ์ƒํ˜ธ ์šด์šฉ์„ฑ - local: modular_transformers title: transformers์—์„œ์˜ ๋ชจ๋“ˆ์„ฑ title: (๋ฒˆ์—ญ์ค‘) ๊ฐœ๋ฐœ์ž ๊ฐ€์ด๋“œ - sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Getting started - local: quantization/bitsandbytes title: bitsandbytes - local: quantization/gptq title: GPTQ - local: quantization/awq title: AWQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) AQLM - local: in_translation title: (๋ฒˆ์—ญ์ค‘) VPTQ - local: quantization/quanto title: Quanto - local: quantization/quark title: Quark - local: quantization/eetq title: EETQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) HQQ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Optimum - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Contribute new quantization method title: (๋ฒˆ์—ญ์ค‘) ๊ฒฝ๋Ÿ‰ํ™” ๋ฉ”์†Œ๋“œ - sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Quantization - local: llm_optims title: LLM ์ถ”๋ก  ์ตœ์ ํ™” - local: cache_explanation title: ์–ดํ…์…˜ ํ–‰๋ ฌ ์บ์‹ฑ - sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Methods and tools for efficient training on a single GPU - local: perf_train_gpu_many title: ๋‹ค์ค‘ GPU์—์„œ ํ›ˆ๋ จ ์ง„ํ–‰ํ•˜๊ธฐ - local: deepspeed title: DeepSpeed - local: fsdp title: ์™„์ „ ๋ถ„ํ•  ๋ฐ์ดํ„ฐ ๋ณ‘๋ ฌ ์ฒ˜๋ฆฌ - local: perf_train_cpu title: CPU์—์„œ ํ›ˆ๋ จ - local: perf_train_cpu_many title: ๋‹ค์ค‘ CPU์—์„œ ํ›ˆ๋ จํ•˜๊ธฐ - local: perf_train_special title: Apple ์‹ค๋ฆฌ์ฝ˜์—์„œ PyTorch ํ•™์Šต - local: perf_hardware title: ํ›ˆ๋ จ์šฉ ์‚ฌ์šฉ์ž ๋งž์ถคํ˜• ํ•˜๋“œ์›จ์–ด - local: hpo_train title: Trainer API๋ฅผ ์‚ฌ์šฉํ•œ ํ•˜์ดํผํŒŒ๋ผ๋ฏธํ„ฐ ํƒ์ƒ‰ title: (๋ฒˆ์—ญ์ค‘) ํšจ์œจ์ ์ธ ํ•™์Šต ๊ธฐ์ˆ ๋“ค - sections: - local: perf_infer_cpu title: CPU๋กœ ์ถ”๋ก ํ•˜๊ธฐ - local: perf_infer_gpu_one title: ํ•˜๋‚˜์˜ GPU๋ฅผ ํ™œ์šฉํ•œ ์ถ”๋ก  title: ์ถ”๋ก  ์ตœ์ ํ™”ํ•˜๊ธฐ - local: debugging title: ๋””๋ฒ„๊น… - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Optimize inference using `torch.compile()` title: (๋ฒˆ์—ญ์ค‘) ์„ฑ๋Šฅ ๋ฐ ํ™•์žฅ์„ฑ - sections: - local: contributing title: ๐Ÿค— Transformers์— ๊ธฐ์—ฌํ•˜๋Š” ๋ฐฉ๋ฒ• - local: add_new_model title: ๐Ÿค— Transformers์— ์ƒˆ๋กœ์šด ๋ชจ๋ธ์„ ์ถ”๊ฐ€ํ•˜๋Š” ๋ฐฉ๋ฒ• - local: add_new_pipeline title: ์–ด๋–ป๊ฒŒ ๐Ÿค— Transformers์— ํŒŒ์ดํ”„๋ผ์ธ์„ ์ถ”๊ฐ€ํ•˜๋‚˜์š”? - local: testing title: ํ…Œ์ŠคํŠธ - local: pr_checks title: Pull Request์— ๋Œ€ํ•œ ๊ฒ€์‚ฌ title: ๋ฆฌ์†Œ์Šค - isExpanded: false sections: - local: contributing title: ๐Ÿค— Transformers์— ๊ธฐ์—ฌํ•˜๋Š” ๋ฐฉ๋ฒ• - local: testing title: Transformers ๋ชจ๋ธ ํ…Œ์ŠคํŠธํ•˜๊ธฐ - local: pr_checks title: Pull request ๊ฒ€์‚ฌํ•˜๊ธฐ title: ๊ธฐ์—ฌํ•˜๊ธฐ - isExpanded: false sections: - sections: - local: model_doc/auto title: Auto Classes - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Backbones - local: main_classes/callback title: Callbacks - local: main_classes/configuration title: Configuration - local: main_classes/data_collator title: Data Collator - local: main_classes/logging title: Logging - local: main_classes/model title: Models - local: main_classes/text_generation title: ํ…์ŠคํŠธ ์ƒ์„ฑ - local: main_classes/onnx title: ONNX - local: main_classes/optimizer_schedules title: ์ตœ์ ํ™” - local: main_classes/output title: ๋ชจ๋ธ ์ถœ๋ ฅ - local: main_classes/peft title: PEFT - local: main_classes/pipelines title: ํŒŒ์ดํ”„๋ผ์ธ - local: main_classes/processors title: ํ”„๋กœ์„ธ์„œ - local: main_classes/quantization title: ์–‘์žํ™” - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Tokenizer - local: main_classes/tokenizer title: ํ† ํฌ๋‚˜์ด์ € - local: main_classes/trainer title: Trainer - local: in_translation title: (๋ฒˆ์—ญ์ค‘) DeepSpeed - local: in_translation title: ExecuTorch - local: main_classes/feature_extractor title: ํ”ผ์ณ ์ถ”์ถœ๊ธฐ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Image Processor - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Video Processor title: ๋ฉ”์ธ ํด๋ž˜์Šค - sections: - sections: - local: model_doc/albert title: ALBERT - local: in_translation title: Arcee - local: in_translation title: Bamba - local: model_doc/bart title: BART - local: model_doc/barthez title: BARThez - local: model_doc/bartpho title: BARTpho - local: model_doc/bert title: BERT - local: in_translation title: BertGeneration - local: model_doc/bert-japanese title: BertJapanese - local: model_doc/bertweet title: BERTweet - local: model_doc/big_bird title: BigBird - local: in_translation title: BigBirdPegasus - local: model_doc/biogpt title: BioGpt - local: in_translation title: BitNet - local: in_translation title: Blenderbot - local: in_translation title: Blenderbot Small - local: in_translation title: BLOOM - local: in_translation title: BORT - local: in_translation title: ByT5 - local: in_translation title: CamemBERT - local: in_translation title: CANINE - local: model_doc/codegen title: CodeGen - local: model_doc/code_llama title: CodeLlama - local: model_doc/cohere title: Cohere - local: in_translation title: Cohere2 - local: model_doc/convbert title: ConvBERT - local: in_translation title: CPM - local: in_translation title: CPMANT - local: in_translation title: CTRL - local: model_doc/dbrx title: DBRX - local: model_doc/deberta title: DeBERTa - local: model_doc/deberta-v2 title: DeBERTa-v2 - local: model_doc/deepseek_v3 title: DeepSeek-V3 - local: in_translation title: DialoGPT - local: in_translation title: DiffLlama - local: in_translation title: DistilBERT - local: in_translation title: Doge - local: in_translation title: dots1 - local: in_translation title: DPR - local: model_doc/electra title: ELECTRA - local: model_doc/encoder-decoder title: Encoder Decoder Models - local: in_translation title: ERNIE - local: in_translation title: ErnieM - local: model_doc/esm title: ESM - local: model_doc/exaone4 title: EXAONE-4.0 - local: in_translation title: Falcon - local: in_translation title: Falcon3 - local: in_translation title: FalconH1 - local: in_translation title: FalconMamba - local: in_translation title: FLAN-T5 - local: in_translation title: FLAN-UL2 - local: in_translation title: FlauBERT - local: in_translation title: FNet - local: in_translation title: FSMT - local: in_translation title: Funnel Transformer - local: in_translation title: Fuyu - local: model_doc/gemma title: Gemma - local: model_doc/gemma2 title: Gemma2 - local: in_translation title: GLM - local: in_translation title: glm4 - local: model_doc/openai-gpt title: GPT - local: in_translation title: GPT Neo - local: in_translation title: GPT NeoX - local: model_doc/gpt_neox_japanese title: GPT NeoX Japanese - local: in_translation title: GPT-J - local: model_doc/gpt2 title: GPT-2 - local: in_translation title: GPTBigCode - local: in_translation title: GPTSAN Japanese - local: in_translation title: GPTSw3 - local: in_translation title: Granite - local: in_translation title: GraniteMoe - local: in_translation title: GraniteMoeHybrid - local: in_translation title: GraniteMoeShared - local: in_translation title: Helium - local: in_translation title: HerBERT - local: in_translation title: HGNet-V2 - local: in_translation title: I-BERT - local: model_doc/jamba title: Jamba - local: in_translation title: JetMoe - local: in_translation title: Jukebox - local: in_translation title: LED - local: model_doc/lfm2 title: LFM2 - local: in_translation title: LFM2-VL - local: model_doc/llama title: LLaMA - local: model_doc/llama2 title: Llama2 - local: model_doc/llama3 title: Llama3 - local: in_translation title: Longformer - local: in_translation title: LongT5 - local: in_translation title: LUKE - local: in_translation title: M2M100 - local: in_translation title: MADLAD-400 - local: model_doc/mamba title: Mamba - local: model_doc/mamba2 title: Mamba2 - local: model_doc/marian title: MarianMT - local: in_translation title: MarkupLM - local: in_translation title: MBart and MBart-50 - local: in_translation title: MEGA - local: in_translation title: MegatronBERT - local: in_translation title: MegatronGPT2 - local: in_translation title: MiniMax - local: model_doc/mistral title: Mistral - local: in_translation title: Mixtral - local: in_translation title: mLUKE - local: in_translation title: MobileBERT - local: in_translation title: ModernBert - local: in_translation title: ModernBERTDecoder - local: in_translation title: MPNet - local: in_translation title: MPT - local: in_translation title: MRA - local: in_translation title: MT5 - local: in_translation title: MVP - local: in_translation title: myt5 - local: in_translation title: Nemotron - local: in_translation title: NEZHA - local: in_translation title: NLLB - local: in_translation title: NLLB-MoE - local: in_translation title: Nystrรถmformer - local: in_translation title: OLMo - local: in_translation title: OLMo2 - local: in_translation title: OLMoE - local: in_translation title: Open-Llama - local: in_translation title: OPT - local: in_translation title: Pegasus - local: in_translation title: PEGASUS-X - local: in_translation title: Persimmon - local: in_translation title: Phi - local: in_translation title: Phi-3 - local: in_translation title: PhiMoE - local: in_translation title: PhoBERT - local: in_translation title: PLBart - local: in_translation title: ProphetNet - local: in_translation title: QDQBert - local: in_translation title: Qwen2 - local: in_translation title: Qwen2MoE - local: in_translation title: Qwen3 - local: in_translation title: Qwen3MoE - local: model_doc/rag title: RAG - local: in_translation title: REALM - local: in_translation title: RecurrentGemma - local: in_translation title: Reformer - local: in_translation title: RemBERT - local: in_translation title: RetriBERT - local: model_doc/roberta title: RoBERTa - local: in_translation title: RoBERTa-PreLayerNorm - local: in_translation title: RoCBert - local: in_translation title: RoFormer - local: in_translation title: RWKV - local: in_translation title: Splinter - local: in_translation title: SqueezeBERT - local: in_translation title: StableLm - local: in_translation title: Starcoder2 - local: in_translation title: SwitchTransformers - local: in_translation title: T5 - local: in_translation title: T5Gemma - local: in_translation title: T5v1.1 - local: in_translation title: TAPEX - local: in_translation title: Transformer XL - local: in_translation title: UL2 - local: in_translation title: UMT5 - local: in_translation title: X-MOD - local: in_translation title: XGLM - local: in_translation title: XLM - local: in_translation title: XLM-ProphetNet - local: in_translation title: XLM-RoBERTa - local: in_translation title: XLM-RoBERTa-XL - local: in_translation title: XLM-V - local: in_translation title: XLNet - local: in_translation title: YOSO - local: in_translation title: Zamba - local: in_translation title: Zamba2 title: ํ…์ŠคํŠธ ๋ชจ๋ธ - sections: - local: in_translation title: Aimv2 - local: in_translation title: BEiT - local: in_translation title: BiT - local: in_translation title: Conditional DETR - local: in_translation title: ConvNeXT - local: in_translation title: ConvNeXTV2 - local: in_translation title: CvT - local: in_translation title: D-FINE - local: in_translation title: DAB-DETR - local: in_translation title: DeepSeek-V2 - local: in_translation title: Deformable DETR - local: in_translation title: DeiT - local: in_translation title: Depth Anything - local: in_translation title: Depth Anything V2 - local: in_translation title: DepthPro - local: in_translation title: DETA - local: in_translation title: DETR - local: in_translation title: DiNAT - local: in_translation title: DINOV2 - local: in_translation title: DINOv2 with Registers - local: in_translation title: DiT - local: in_translation title: DPT - local: in_translation title: EfficientFormer - local: in_translation title: EfficientNet - local: in_translation title: EoMT - local: in_translation title: FocalNet - local: in_translation title: GLPN - local: in_translation title: Hiera - local: in_translation title: I-JEPA - local: in_translation title: ImageGPT - local: in_translation title: LeViT - local: in_translation title: LightGlue - local: in_translation title: Mask2Former - local: in_translation title: MaskFormer - local: in_translation title: MLCD - local: in_translation title: MobileNetV1 - local: in_translation title: MobileNetV2 - local: in_translation title: MobileViT - local: in_translation title: MobileViTV2 - local: in_translation title: NAT - local: in_translation title: PoolFormer - local: in_translation title: Prompt Depth Anything - local: in_translation title: Pyramid Vision Transformer (PVT) - local: in_translation title: Pyramid Vision Transformer v2 (PVTv2) - local: in_translation title: RegNet - local: in_translation title: ResNet - local: in_translation title: RT-DETR - local: in_translation title: RT-DETRv2 - local: in_translation title: SegFormer - local: in_translation title: SegGpt - local: model_doc/sam_hq title: Segment Anything High Quality (SAM-HQ) - local: in_translation title: SuperGlue - local: in_translation title: SuperPoint - local: in_translation title: SwiftFormer - local: model_doc/swin title: Swin Transformer - local: model_doc/swinv2 title: Swin Transformer V2 - local: model_doc/swin2sr title: Swin2SR - local: in_translation title: Table Transformer - local: in_translation title: TextNet - local: in_translation title: Timm Wrapper - local: in_translation title: UperNet - local: in_translation title: VAN - local: model_doc/vit title: Vision Transformer (ViT) - local: in_translation title: ViT Hybrid - local: in_translation title: ViTDet - local: in_translation title: ViTMAE - local: in_translation title: ViTMatte - local: in_translation title: ViTMSN - local: in_translation title: ViTPose - local: in_translation title: YOLOS - local: in_translation title: ZoeDepth title: ๋น„์ „ ๋ชจ๋ธ - sections: - local: in_translation title: Audio Spectrogram Transformer - local: in_translation title: Bark - local: in_translation title: CLAP - local: in_translation title: CSM - local: in_translation title: dac - local: in_translation title: Dia - local: in_translation title: EnCodec - local: in_translation title: FastSpeech2Conformer - local: in_translation title: GraniteSpeech - local: in_translation title: Hubert - local: in_translation title: Kyutai Speech-To-Text - local: in_translation title: MCTCT - local: in_translation title: Mimi - local: in_translation title: MMS - local: in_translation title: Moonshine - local: in_translation title: Moshi - local: in_translation title: MusicGen - local: in_translation title: MusicGen Melody - local: in_translation title: Pop2Piano - local: in_translation title: Seamless-M4T - local: in_translation title: SeamlessM4T-v2 - local: in_translation title: SEW - local: in_translation title: SEW-D - local: in_translation title: Speech2Text - local: in_translation title: Speech2Text2 - local: in_translation title: SpeechT5 - local: in_translation title: UniSpeech - local: in_translation title: UniSpeech-SAT - local: in_translation title: UnivNet - local: in_translation title: VITS - local: in_translation title: Wav2Vec2 - local: in_translation title: Wav2Vec2-BERT - local: in_translation title: Wav2Vec2-Conformer - local: in_translation title: Wav2Vec2Phoneme - local: in_translation title: WavLM - local: model_doc/whisper title: Whisper - local: model_doc/xclip title: xclip - local: in_translation title: XLS-R - local: in_translation title: XLSR-Wav2Vec2 title: ์˜ค๋””์˜ค ๋ชจ๋ธ - sections: - local: model_doc/timesformer title: TimeSformer - local: in_translation title: V-JEPA 2 - local: in_translation title: VideoMAE - local: model_doc/vivit title: ViViT title: ๋น„๋””์˜ค ๋ชจ๋ธ - sections: - local: in_translation title: ALIGN - local: model_doc/altclip title: AltCLIP - local: in_translation title: Aria - local: in_translation title: AyaVision - local: model_doc/blip title: BLIP - local: model_doc/blip-2 title: BLIP-2 - local: in_translation title: BridgeTower - local: in_translation title: BROS - local: model_doc/chameleon title: Chameleon - local: in_translation title: Chinese-CLIP - local: model_doc/clip title: CLIP - local: model_doc/clipseg title: CLIPSeg - local: in_translation title: CLVP - local: in_translation title: ColPali - local: in_translation title: ColQwen2 - local: in_translation title: Data2Vec - local: in_translation title: DePlot - local: in_translation title: Donut - local: in_translation title: Emu3 - local: in_translation title: FLAVA - local: model_doc/gemma3 title: Gemma3 - local: model_doc/gemma3n title: Gemma3n - local: in_translation title: GIT - local: in_translation title: glm4v - local: in_translation title: GOT-OCR2 - local: in_translation title: GraniteVision - local: in_translation title: (๋ฒˆ์—ญ์ค‘) GIT - local: model_doc/grounding-dino title: Grounding DINO - local: in_translation title: GroupViT - local: in_translation title: IDEFICS - local: in_translation title: Idefics2 - local: in_translation title: Idefics3 - local: in_translation title: InstructBLIP - local: in_translation title: InstructBlipVideo - local: in_translation title: InternVL - local: in_translation title: Janus - local: in_translation title: KOSMOS-2 - local: in_translation title: LayoutLM - local: in_translation title: LayoutLMV2 - local: in_translation title: LayoutLMV3 - local: in_translation title: LayoutXLM - local: in_translation title: LiLT - local: model_doc/llama4 title: Llama4 - local: in_translation title: Llava - local: in_translation title: LLaVA-NeXT - local: in_translation title: LLaVa-NeXT-Video - local: in_translation title: LLaVA-Onevision - local: in_translation title: LXMERT - local: in_translation title: MatCha - local: in_translation title: MGP-STR - local: in_translation title: Mistral3 - local: in_translation title: mllama - local: in_translation title: Nougat - local: in_translation title: OmDet-Turbo - local: in_translation title: OneFormer - local: in_translation title: OWL-ViT - local: in_translation title: OWLv2 - local: model_doc/paligemma title: PaliGemma - local: in_translation title: Perceiver - local: in_translation title: PerceptionLM - local: in_translation title: Phi4 Multimodal - local: in_translation title: Pix2Struct - local: in_translation title: Pixtral - local: in_translation title: Qwen2.5-Omni - local: in_translation title: Qwen2.5-VL - local: in_translation title: Qwen2Audio - local: model_doc/qwen2_vl title: Qwen2VL - local: in_translation title: Segment Anything - local: in_translation title: Segment Anything High Quality - local: in_translation title: ShieldGemma2 - local: model_doc/siglip title: SigLIP - local: in_translation title: SigLIP2 - local: in_translation title: SmolLM3 - local: model_doc/smolvlm title: ์†Œํ˜• ๋น„์ „ ์–ธ์–ด ๋ชจ๋ธ - local: in_translation title: Speech Encoder Decoder Models - local: in_translation title: TAPAS - local: in_translation title: TrOCR - local: in_translation title: TVLT - local: model_doc/tvp title: TVP - local: in_translation title: UDOP - local: in_translation title: VideoLlava - local: in_translation title: ViLT - local: in_translation title: VipLlava - local: in_translation title: Vision Encoder Decoder Models - local: in_translation title: Vision Text Dual Encoder - local: in_translation title: VisualBERT - local: in_translation title: Voxtral - local: in_translation title: X-CLIP title: ๋ฉ€ํ‹ฐ๋ชจ๋‹ฌ ๋ชจ๋ธ - sections: - local: in_translation title: Decision Transformer - local: model_doc/trajectory_transformer title: Trajectory Transformer title: ๊ฐ•ํ™”ํ•™์Šต ๋ชจ๋ธ - sections: - local: model_doc/autoformer title: Autoformer - local: model_doc/informer title: Informer - local: model_doc/patchtsmixer title: PatchTSMixer - local: model_doc/patchtst title: PatchTST - local: model_doc/time_series_transformer title: Time Series Transformer - local: in_translation title: TimesFM title: ์‹œ๊ฒŒ์—ด ๋ชจ๋ธ - sections: - local: model_doc/graphormer title: Graphormer title: ๊ทธ๋ž˜ํ”„ ๋ชจ๋ธ title: ๋ชจ๋ธ - sections: - local: internal/modeling_utils title: ์‚ฌ์šฉ์ž ์ •์˜ ๋ ˆ์ด์–ด ๋ฐ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Utilities for Model Debugging - local: internal/pipelines_utils title: ํŒŒ์ดํ”„๋ผ์ธ์„ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/tokenization_utils title: ํ† ํฌ๋‚˜์ด์ €๋ฅผ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/trainer_utils title: Trainer๋ฅผ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/generation_utils title: ์ƒ์„ฑ์„ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/image_processing_utils title: ์ด๋ฏธ์ง€ ํ”„๋กœ์„ธ์„œ๋ฅผ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/audio_utils title: ์˜ค๋””์˜ค ์ฒ˜๋ฆฌ๋ฅผ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: internal/file_utils title: ์ผ๋ฐ˜ ์œ ํ‹ธ๋ฆฌํ‹ฐ - local: in_translation title: (๋ฒˆ์—ญ์ค‘) Importing Utilities - local: internal/time_series_utils title: ์‹œ๊ณ„์—ด์„ ์œ„ํ•œ ์œ ํ‹ธ๋ฆฌํ‹ฐ title: ๋‚ด๋ถ€ ํ—ฌํผ(Internal helpers) - sections: - local: in_translation title: (๋ฒˆ์—ญ์ค‘)Environment Variables title: Reference title: API