[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill-huggingface-transformers":3,"mdc-52mns7-key":35,"related-repo-huggingface-transformers":775,"related-org-huggingface-transformers":784},{"slug":4,"name":4,"fn":5,"description":6,"org":7,"tags":11,"stars":24,"repoUrl":25,"updatedAt":26,"license":27,"forks":28,"topics":29,"repo":30,"sourceUrl":33,"mdContent":34},"transformers","run Transformers machine learning tasks","Agent-invokable Transformers commands. Pass `--format json` at the top level (e.g. `transformers --format json classify ...`) to receive the structured output documented in each capability's `outputs` schema.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},"huggingface","Hugging Face","https:\u002F\u002Fpexgzepcugksgbtrxkhf.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Forg-logos\u002Fhuggingface.png",[12,15,18,21],{"name":9,"slug":13,"type":14},"hugging-face","tag",{"name":16,"slug":17,"type":14},"Deep Learning","deep-learning",{"name":19,"slug":20,"type":14},"Python","python",{"name":22,"slug":23,"type":14},"AI Infrastructure","ai-infrastructure",3,"https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fagentcap","2026-06-27T06:56:08.059837",null,1,[],{"repoUrl":25,"stars":24,"forks":28,"topics":31,"description":32},[],"A framework to capture, analyse and export agentic sessions","https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fagentcap\u002Ftree\u002FHEAD\u002Fexamples\u002Ftransformers-agentic\u002Fskill\u002Fskills\u002Ftransformers","---\nname: transformers\ndescription: Agent-invokable Transformers commands. Pass `--format json` at the top level (e.g. `transformers --format json classify ...`) to receive the structured output documented in each capability's `outputs` schema.\n---\n\n# Transformers CLI\n\nFor one-off inference, training, quantization, or export, invoke the\n`transformers` command directly rather than writing Python. Run\n`transformers --help` for the full command list; run\n`transformers \u003Ccommand> --help` for flags per command.\n\n## Invocation rules\n\n**All inputs are named flags, never positional.** Wrong invocations like\n``transformers classify \"my text\"`` or ``transformers ner \"sentence\"`` will\nfail with ``Got unexpected extra argument``. The text \u002F image \u002F audio \u002F file\nargument is always a flag: ``--text``, ``--image``, ``--audio``, ``--file``.\n\n**Always invoke as `transformers \u003Ccmd> ...`.** Do not use\n``python -m transformers ...`` patterns — the console script is what the\n``transformers`` package installs.\n\n**Use `transformers --format json` for machine-readable output**:\n``transformers --format json classify --text \"...\"``.\n\n## Example invocations (copy these shapes)\n\nText (classify, ner, token-classify, summarize, translate, fill-mask):\n```\ntransformers classify --text \"I loved this movie\"\ntransformers classify --text \"...\" --model distilbert\u002Fdistilbert-base-uncased-finetuned-sst-2-english\ntransformers ner --text \"Apple CEO Tim Cook visited Paris.\" --model dslim\u002Fbert-base-NER\ntransformers summarize --file article.txt --model facebook\u002Fbart-large-cnn\ntransformers translate --text \"The weather is nice\" --model Helsinki-NLP\u002Fopus-mt-en-de\n```\n\nQuestion answering (takes `--question` and `--context`):\n```\ntransformers qa --question \"Who invented it?\" --context \"Graham Bell invented the telephone in 1876.\"\n```\n\nImage (caption, image-classify, detect, segment, depth, vqa, ocr):\n```\ntransformers caption --image photo.jpg --model llava-hf\u002Fllava-interleave-qwen-0.5b-hf\ntransformers image-classify --image photo.jpg\ntransformers vqa --image photo.jpg --question \"What color is the car?\"\n```\n\nAudio (transcribe, audio-classify, speak):\n```\ntransformers transcribe --audio clip.wav --model openai\u002Fwhisper-tiny\ntransformers audio-classify --audio clip.wav\ntransformers speak --text \"Hello\" --output hello.wav\n```\n\nTokenize \u002F inspect \u002F embed:\n```\ntransformers tokenize --model HuggingFaceTB\u002FSmolLM2-360M-Instruct --text \"tokenize me\"\ntransformers inspect meta-llama\u002FLlama-3.2-1B-Instruct\ntransformers embed --text \"some sentence\" --model BAAI\u002Fbge-small-en-v1.5\n```\n\nGenerate (text completion):\n```\ntransformers generate --prompt \"Once upon a time\" --model HuggingFaceTB\u002FSmolLM2-360M-Instruct\n```\n\n## Available commands\n\n- `transformers classify` — Classify text into categories\n- `transformers ner` — Extract named entities from text (NER)\n- `transformers token-classify` — Tag tokens with labels (POS tagging, chunking, etc.)\n- `transformers qa` — Answer a question given a context paragraph (extractive QA)\n- `transformers table-qa` — Answer a question about tabular data (CSV)\n- `transformers summarize` — Summarize text\n- `transformers translate` — Translate text between languages\n- `transformers fill-mask` — Predict the masked token in a sentence\n- `transformers image-classify` — Classify an image\n- `transformers detect` — Detect objects in an image\n- `transformers segment` — Segment an image\n- `transformers depth` — Estimate a depth map from an image\n- `transformers keypoints` — Match keypoints between two images\n- `transformers video-classify` — Classify a video\n- `transformers transcribe` — Transcribe speech from an audio file\n- `transformers audio-classify` — Classify an audio file into categories\n- `transformers speak` — Synthesize speech from text and save to a WAV file\n- `transformers audio-generate` — Generate audio (e.g. music) from a text description and save to a WAV file\n- `transformers vqa` — Visual question answering using ``AutoModelForImageTextToText``\n- `transformers document-qa` — Extractive document question answering using\n- `transformers caption` — Generate a caption for an image using ``AutoModelForImageTextToText``\n- `transformers ocr` — Extract text from an image using ``AutoModelForImageTextToText``\n- `transformers multimodal-chat` — Single-turn conversation with a model that accepts mixed inputs\n- `transformers generate` — Generate text from a prompt with full control over decoding\n- `transformers detect-watermark` — Detect whether text contains a watermark\n- `transformers embed` — Compute embeddings for text or images\n- `transformers tokenize` — Tokenize text and display the resulting tokens\n- `transformers inspect` — Inspect a model's configuration without downloading weights\n- `transformers inspect-forward` — Examine attention weights and hidden states from a forward pass\n- `transformers benchmark-quantization` — Compare quality and performance across quantization methods\n- `transformers train` — Fine-tune or pretrain a model on a dataset\n- `transformers quantize` — Quantize a model and save it\n- `transformers export` — Export a model to a deployment-friendly format\n\n## When to use what\n\n- **Atomic task** (single inference \u002F training \u002F export): use the CLI.\n- **Composed workflow** (chain models, custom logic): write Python.\n  The CLI commands' source in `transformers.cli.agentic.*` is the\n  canonical template — each file loads a model with `AutoModel*` +\n  `AutoProcessor`\u002F`AutoTokenizer`, runs a forward pass, and\n  post-processes. Copy that pattern rather than reaching for\n  `pipeline(...)`.\n",{"data":36,"body":37},{"name":4,"description":6},{"type":38,"children":39},"root",[40,49,79,86,151,183,208,214,219,231,252,261,266,275,280,289,294,303,308,317,323,707,713],{"type":41,"tag":42,"props":43,"children":45},"element","h1",{"id":44},"transformers-cli",[46],{"type":47,"value":48},"text","Transformers CLI",{"type":41,"tag":50,"props":51,"children":52},"p",{},[53,55,61,63,69,71,77],{"type":47,"value":54},"For one-off inference, training, quantization, or export, invoke the\n",{"type":41,"tag":56,"props":57,"children":59},"code",{"className":58},[],[60],{"type":47,"value":4},{"type":47,"value":62}," command directly rather than writing Python. Run\n",{"type":41,"tag":56,"props":64,"children":66},{"className":65},[],[67],{"type":47,"value":68},"transformers --help",{"type":47,"value":70}," for the full command list; run\n",{"type":41,"tag":56,"props":72,"children":74},{"className":73},[],[75],{"type":47,"value":76},"transformers \u003Ccommand> --help",{"type":47,"value":78}," for flags per command.",{"type":41,"tag":80,"props":81,"children":83},"h2",{"id":82},"invocation-rules",[84],{"type":47,"value":85},"Invocation rules",{"type":41,"tag":50,"props":87,"children":88},{},[89,95,97,103,105,111,113,119,121,127,129,135,136,142,143,149],{"type":41,"tag":90,"props":91,"children":92},"strong",{},[93],{"type":47,"value":94},"All inputs are named flags, never positional.",{"type":47,"value":96}," Wrong invocations like\n",{"type":41,"tag":56,"props":98,"children":100},{"className":99},[],[101],{"type":47,"value":102},"transformers classify \"my text\"",{"type":47,"value":104}," or ",{"type":41,"tag":56,"props":106,"children":108},{"className":107},[],[109],{"type":47,"value":110},"transformers ner \"sentence\"",{"type":47,"value":112}," will\nfail with ",{"type":41,"tag":56,"props":114,"children":116},{"className":115},[],[117],{"type":47,"value":118},"Got unexpected extra argument",{"type":47,"value":120},". The text \u002F image \u002F audio \u002F file\nargument is always a flag: ",{"type":41,"tag":56,"props":122,"children":124},{"className":123},[],[125],{"type":47,"value":126},"--text",{"type":47,"value":128},", ",{"type":41,"tag":56,"props":130,"children":132},{"className":131},[],[133],{"type":47,"value":134},"--image",{"type":47,"value":128},{"type":41,"tag":56,"props":137,"children":139},{"className":138},[],[140],{"type":47,"value":141},"--audio",{"type":47,"value":128},{"type":41,"tag":56,"props":144,"children":146},{"className":145},[],[147],{"type":47,"value":148},"--file",{"type":47,"value":150},".",{"type":41,"tag":50,"props":152,"children":153},{},[154,166,168,174,176,181],{"type":41,"tag":90,"props":155,"children":156},{},[157,159,165],{"type":47,"value":158},"Always invoke as ",{"type":41,"tag":56,"props":160,"children":162},{"className":161},[],[163],{"type":47,"value":164},"transformers \u003Ccmd> ...",{"type":47,"value":150},{"type":47,"value":167}," Do not use\n",{"type":41,"tag":56,"props":169,"children":171},{"className":170},[],[172],{"type":47,"value":173},"python -m transformers ...",{"type":47,"value":175}," patterns — the console script is what the\n",{"type":41,"tag":56,"props":177,"children":179},{"className":178},[],[180],{"type":47,"value":4},{"type":47,"value":182}," package installs.",{"type":41,"tag":50,"props":184,"children":185},{},[186,199,201,207],{"type":41,"tag":90,"props":187,"children":188},{},[189,191,197],{"type":47,"value":190},"Use ",{"type":41,"tag":56,"props":192,"children":194},{"className":193},[],[195],{"type":47,"value":196},"transformers --format json",{"type":47,"value":198}," for machine-readable output",{"type":47,"value":200},":\n",{"type":41,"tag":56,"props":202,"children":204},{"className":203},[],[205],{"type":47,"value":206},"transformers --format json classify --text \"...\"",{"type":47,"value":150},{"type":41,"tag":80,"props":209,"children":211},{"id":210},"example-invocations-copy-these-shapes",[212],{"type":47,"value":213},"Example invocations (copy these shapes)",{"type":41,"tag":50,"props":215,"children":216},{},[217],{"type":47,"value":218},"Text (classify, ner, token-classify, summarize, translate, fill-mask):",{"type":41,"tag":220,"props":221,"children":225},"pre",{"className":222,"code":224,"language":47},[223],"language-text","transformers classify --text \"I loved this movie\"\ntransformers classify --text \"...\" --model distilbert\u002Fdistilbert-base-uncased-finetuned-sst-2-english\ntransformers ner --text \"Apple CEO Tim Cook visited Paris.\" --model dslim\u002Fbert-base-NER\ntransformers summarize --file article.txt --model facebook\u002Fbart-large-cnn\ntransformers translate --text \"The weather is nice\" --model Helsinki-NLP\u002Fopus-mt-en-de\n",[226],{"type":41,"tag":56,"props":227,"children":229},{"__ignoreMap":228},"",[230],{"type":47,"value":224},{"type":41,"tag":50,"props":232,"children":233},{},[234,236,242,244,250],{"type":47,"value":235},"Question answering (takes ",{"type":41,"tag":56,"props":237,"children":239},{"className":238},[],[240],{"type":47,"value":241},"--question",{"type":47,"value":243}," and ",{"type":41,"tag":56,"props":245,"children":247},{"className":246},[],[248],{"type":47,"value":249},"--context",{"type":47,"value":251},"):",{"type":41,"tag":220,"props":253,"children":256},{"className":254,"code":255,"language":47},[223],"transformers qa --question \"Who invented it?\" --context \"Graham Bell invented the telephone in 1876.\"\n",[257],{"type":41,"tag":56,"props":258,"children":259},{"__ignoreMap":228},[260],{"type":47,"value":255},{"type":41,"tag":50,"props":262,"children":263},{},[264],{"type":47,"value":265},"Image (caption, image-classify, detect, segment, depth, vqa, ocr):",{"type":41,"tag":220,"props":267,"children":270},{"className":268,"code":269,"language":47},[223],"transformers caption --image photo.jpg --model llava-hf\u002Fllava-interleave-qwen-0.5b-hf\ntransformers image-classify --image photo.jpg\ntransformers vqa --image photo.jpg --question \"What color is the car?\"\n",[271],{"type":41,"tag":56,"props":272,"children":273},{"__ignoreMap":228},[274],{"type":47,"value":269},{"type":41,"tag":50,"props":276,"children":277},{},[278],{"type":47,"value":279},"Audio (transcribe, audio-classify, speak):",{"type":41,"tag":220,"props":281,"children":284},{"className":282,"code":283,"language":47},[223],"transformers transcribe --audio clip.wav --model openai\u002Fwhisper-tiny\ntransformers audio-classify --audio clip.wav\ntransformers speak --text \"Hello\" --output hello.wav\n",[285],{"type":41,"tag":56,"props":286,"children":287},{"__ignoreMap":228},[288],{"type":47,"value":283},{"type":41,"tag":50,"props":290,"children":291},{},[292],{"type":47,"value":293},"Tokenize \u002F inspect \u002F embed:",{"type":41,"tag":220,"props":295,"children":298},{"className":296,"code":297,"language":47},[223],"transformers tokenize --model HuggingFaceTB\u002FSmolLM2-360M-Instruct --text \"tokenize me\"\ntransformers inspect meta-llama\u002FLlama-3.2-1B-Instruct\ntransformers embed --text \"some sentence\" --model BAAI\u002Fbge-small-en-v1.5\n",[299],{"type":41,"tag":56,"props":300,"children":301},{"__ignoreMap":228},[302],{"type":47,"value":297},{"type":41,"tag":50,"props":304,"children":305},{},[306],{"type":47,"value":307},"Generate (text completion):",{"type":41,"tag":220,"props":309,"children":312},{"className":310,"code":311,"language":47},[223],"transformers generate --prompt \"Once upon a time\" --model HuggingFaceTB\u002FSmolLM2-360M-Instruct\n",[313],{"type":41,"tag":56,"props":314,"children":315},{"__ignoreMap":228},[316],{"type":47,"value":311},{"type":41,"tag":80,"props":318,"children":320},{"id":319},"available-commands",[321],{"type":47,"value":322},"Available commands",{"type":41,"tag":324,"props":325,"children":326},"ul",{},[327,339,350,361,372,383,394,405,416,427,438,449,460,471,482,493,504,515,526,543,554,570,586,597,608,619,630,641,652,663,674,685,696],{"type":41,"tag":328,"props":329,"children":330},"li",{},[331,337],{"type":41,"tag":56,"props":332,"children":334},{"className":333},[],[335],{"type":47,"value":336},"transformers classify",{"type":47,"value":338}," — Classify text into categories",{"type":41,"tag":328,"props":340,"children":341},{},[342,348],{"type":41,"tag":56,"props":343,"children":345},{"className":344},[],[346],{"type":47,"value":347},"transformers ner",{"type":47,"value":349}," — Extract named entities from text (NER)",{"type":41,"tag":328,"props":351,"children":352},{},[353,359],{"type":41,"tag":56,"props":354,"children":356},{"className":355},[],[357],{"type":47,"value":358},"transformers token-classify",{"type":47,"value":360}," — Tag tokens with labels (POS tagging, chunking, etc.)",{"type":41,"tag":328,"props":362,"children":363},{},[364,370],{"type":41,"tag":56,"props":365,"children":367},{"className":366},[],[368],{"type":47,"value":369},"transformers qa",{"type":47,"value":371}," — Answer a question given a context paragraph (extractive QA)",{"type":41,"tag":328,"props":373,"children":374},{},[375,381],{"type":41,"tag":56,"props":376,"children":378},{"className":377},[],[379],{"type":47,"value":380},"transformers table-qa",{"type":47,"value":382}," — Answer a question about tabular data (CSV)",{"type":41,"tag":328,"props":384,"children":385},{},[386,392],{"type":41,"tag":56,"props":387,"children":389},{"className":388},[],[390],{"type":47,"value":391},"transformers summarize",{"type":47,"value":393}," — Summarize text",{"type":41,"tag":328,"props":395,"children":396},{},[397,403],{"type":41,"tag":56,"props":398,"children":400},{"className":399},[],[401],{"type":47,"value":402},"transformers translate",{"type":47,"value":404}," — Translate text between languages",{"type":41,"tag":328,"props":406,"children":407},{},[408,414],{"type":41,"tag":56,"props":409,"children":411},{"className":410},[],[412],{"type":47,"value":413},"transformers fill-mask",{"type":47,"value":415}," — Predict the masked token in a sentence",{"type":41,"tag":328,"props":417,"children":418},{},[419,425],{"type":41,"tag":56,"props":420,"children":422},{"className":421},[],[423],{"type":47,"value":424},"transformers image-classify",{"type":47,"value":426}," — Classify an image",{"type":41,"tag":328,"props":428,"children":429},{},[430,436],{"type":41,"tag":56,"props":431,"children":433},{"className":432},[],[434],{"type":47,"value":435},"transformers detect",{"type":47,"value":437}," — Detect objects in an image",{"type":41,"tag":328,"props":439,"children":440},{},[441,447],{"type":41,"tag":56,"props":442,"children":444},{"className":443},[],[445],{"type":47,"value":446},"transformers segment",{"type":47,"value":448}," — Segment an image",{"type":41,"tag":328,"props":450,"children":451},{},[452,458],{"type":41,"tag":56,"props":453,"children":455},{"className":454},[],[456],{"type":47,"value":457},"transformers depth",{"type":47,"value":459}," — Estimate a depth map from an image",{"type":41,"tag":328,"props":461,"children":462},{},[463,469],{"type":41,"tag":56,"props":464,"children":466},{"className":465},[],[467],{"type":47,"value":468},"transformers keypoints",{"type":47,"value":470}," — Match keypoints between two images",{"type":41,"tag":328,"props":472,"children":473},{},[474,480],{"type":41,"tag":56,"props":475,"children":477},{"className":476},[],[478],{"type":47,"value":479},"transformers video-classify",{"type":47,"value":481}," — Classify a video",{"type":41,"tag":328,"props":483,"children":484},{},[485,491],{"type":41,"tag":56,"props":486,"children":488},{"className":487},[],[489],{"type":47,"value":490},"transformers transcribe",{"type":47,"value":492}," — Transcribe speech from an audio file",{"type":41,"tag":328,"props":494,"children":495},{},[496,502],{"type":41,"tag":56,"props":497,"children":499},{"className":498},[],[500],{"type":47,"value":501},"transformers audio-classify",{"type":47,"value":503}," — Classify an audio file into categories",{"type":41,"tag":328,"props":505,"children":506},{},[507,513],{"type":41,"tag":56,"props":508,"children":510},{"className":509},[],[511],{"type":47,"value":512},"transformers speak",{"type":47,"value":514}," — Synthesize speech from text and save to a WAV file",{"type":41,"tag":328,"props":516,"children":517},{},[518,524],{"type":41,"tag":56,"props":519,"children":521},{"className":520},[],[522],{"type":47,"value":523},"transformers audio-generate",{"type":47,"value":525}," — Generate audio (e.g. music) from a text description and save to a WAV file",{"type":41,"tag":328,"props":527,"children":528},{},[529,535,537],{"type":41,"tag":56,"props":530,"children":532},{"className":531},[],[533],{"type":47,"value":534},"transformers vqa",{"type":47,"value":536}," — Visual question answering using ",{"type":41,"tag":56,"props":538,"children":540},{"className":539},[],[541],{"type":47,"value":542},"AutoModelForImageTextToText",{"type":41,"tag":328,"props":544,"children":545},{},[546,552],{"type":41,"tag":56,"props":547,"children":549},{"className":548},[],[550],{"type":47,"value":551},"transformers document-qa",{"type":47,"value":553}," — Extractive document question answering using",{"type":41,"tag":328,"props":555,"children":556},{},[557,563,565],{"type":41,"tag":56,"props":558,"children":560},{"className":559},[],[561],{"type":47,"value":562},"transformers caption",{"type":47,"value":564}," — Generate a caption for an image using ",{"type":41,"tag":56,"props":566,"children":568},{"className":567},[],[569],{"type":47,"value":542},{"type":41,"tag":328,"props":571,"children":572},{},[573,579,581],{"type":41,"tag":56,"props":574,"children":576},{"className":575},[],[577],{"type":47,"value":578},"transformers ocr",{"type":47,"value":580}," — Extract text from an image using ",{"type":41,"tag":56,"props":582,"children":584},{"className":583},[],[585],{"type":47,"value":542},{"type":41,"tag":328,"props":587,"children":588},{},[589,595],{"type":41,"tag":56,"props":590,"children":592},{"className":591},[],[593],{"type":47,"value":594},"transformers multimodal-chat",{"type":47,"value":596}," — Single-turn conversation with a model that accepts mixed inputs",{"type":41,"tag":328,"props":598,"children":599},{},[600,606],{"type":41,"tag":56,"props":601,"children":603},{"className":602},[],[604],{"type":47,"value":605},"transformers generate",{"type":47,"value":607}," — Generate text from a prompt with full control over decoding",{"type":41,"tag":328,"props":609,"children":610},{},[611,617],{"type":41,"tag":56,"props":612,"children":614},{"className":613},[],[615],{"type":47,"value":616},"transformers detect-watermark",{"type":47,"value":618}," — Detect whether text contains a watermark",{"type":41,"tag":328,"props":620,"children":621},{},[622,628],{"type":41,"tag":56,"props":623,"children":625},{"className":624},[],[626],{"type":47,"value":627},"transformers embed",{"type":47,"value":629}," — Compute embeddings for text or images",{"type":41,"tag":328,"props":631,"children":632},{},[633,639],{"type":41,"tag":56,"props":634,"children":636},{"className":635},[],[637],{"type":47,"value":638},"transformers tokenize",{"type":47,"value":640}," — Tokenize text and display the resulting tokens",{"type":41,"tag":328,"props":642,"children":643},{},[644,650],{"type":41,"tag":56,"props":645,"children":647},{"className":646},[],[648],{"type":47,"value":649},"transformers inspect",{"type":47,"value":651}," — Inspect a model's configuration without downloading weights",{"type":41,"tag":328,"props":653,"children":654},{},[655,661],{"type":41,"tag":56,"props":656,"children":658},{"className":657},[],[659],{"type":47,"value":660},"transformers inspect-forward",{"type":47,"value":662}," — Examine attention weights and hidden states from a forward pass",{"type":41,"tag":328,"props":664,"children":665},{},[666,672],{"type":41,"tag":56,"props":667,"children":669},{"className":668},[],[670],{"type":47,"value":671},"transformers benchmark-quantization",{"type":47,"value":673}," — Compare quality and performance across quantization methods",{"type":41,"tag":328,"props":675,"children":676},{},[677,683],{"type":41,"tag":56,"props":678,"children":680},{"className":679},[],[681],{"type":47,"value":682},"transformers train",{"type":47,"value":684}," — Fine-tune or pretrain a model on a dataset",{"type":41,"tag":328,"props":686,"children":687},{},[688,694],{"type":41,"tag":56,"props":689,"children":691},{"className":690},[],[692],{"type":47,"value":693},"transformers quantize",{"type":47,"value":695}," — Quantize a model and save it",{"type":41,"tag":328,"props":697,"children":698},{},[699,705],{"type":41,"tag":56,"props":700,"children":702},{"className":701},[],[703],{"type":47,"value":704},"transformers export",{"type":47,"value":706}," — Export a model to a deployment-friendly format",{"type":41,"tag":80,"props":708,"children":710},{"id":709},"when-to-use-what",[711],{"type":47,"value":712},"When to use what",{"type":41,"tag":324,"props":714,"children":715},{},[716,726],{"type":41,"tag":328,"props":717,"children":718},{},[719,724],{"type":41,"tag":90,"props":720,"children":721},{},[722],{"type":47,"value":723},"Atomic task",{"type":47,"value":725}," (single inference \u002F training \u002F export): use the CLI.",{"type":41,"tag":328,"props":727,"children":728},{},[729,734,736,742,744,750,752,758,760,766,768,774],{"type":41,"tag":90,"props":730,"children":731},{},[732],{"type":47,"value":733},"Composed workflow",{"type":47,"value":735}," (chain models, custom logic): write Python.\nThe CLI commands' source in ",{"type":41,"tag":56,"props":737,"children":739},{"className":738},[],[740],{"type":47,"value":741},"transformers.cli.agentic.*",{"type":47,"value":743}," is the\ncanonical template — each file loads a model with ",{"type":41,"tag":56,"props":745,"children":747},{"className":746},[],[748],{"type":47,"value":749},"AutoModel*",{"type":47,"value":751}," +\n",{"type":41,"tag":56,"props":753,"children":755},{"className":754},[],[756],{"type":47,"value":757},"AutoProcessor",{"type":47,"value":759},"\u002F",{"type":41,"tag":56,"props":761,"children":763},{"className":762},[],[764],{"type":47,"value":765},"AutoTokenizer",{"type":47,"value":767},", runs a forward pass, and\npost-processes. Copy that pattern rather than reaching for\n",{"type":41,"tag":56,"props":769,"children":771},{"className":770},[],[772],{"type":47,"value":773},"pipeline(...)",{"type":47,"value":150},{"items":776,"total":28},[777],{"slug":4,"name":4,"fn":5,"description":6,"org":778,"tags":779,"stars":24,"repoUrl":25,"updatedAt":26},{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[780,781,782,783],{"name":22,"slug":23,"type":14},{"name":16,"slug":17,"type":14},{"name":9,"slug":13,"type":14},{"name":19,"slug":20,"type":14},{"items":785,"total":956},[786,804,818,835,849,862,875,890,904,914,927,941],{"slug":787,"name":787,"fn":788,"description":789,"org":790,"tags":791,"stars":801,"repoUrl":802,"updatedAt":803},"train-sentence-transformers","train sentence-transformers models","Train or fine-tune sentence-transformers models across `SentenceTransformer` (bi-encoder; dense or static embedding model; for retrieval, similarity, clustering, classification, paraphrase mining, dedup, multimodal), `CrossEncoder` (reranker; pair scoring for two-stage retrieval \u002F pair classification), and `SparseEncoder` (SPLADE, sparse embedding model; for learned-sparse retrieval). Covers loss selection, hard-negative mining, evaluators, distillation, LoRA, Matryoshka, and Hugging Face Hub publishing. Use for any sentence-transformers training task.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[792,793,794,797,798],{"name":16,"slug":17,"type":14},{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},"LLM","llm",{"name":19,"slug":20,"type":14},{"name":799,"slug":800,"type":14},"Search","search",18914,"https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fsentence-transformers","2026-05-08T05:09:16.820066",{"slug":805,"name":805,"fn":806,"description":807,"org":808,"tags":809,"stars":815,"repoUrl":816,"updatedAt":817},"trl-training","train and fine-tune LLMs with TRL","Train and fine-tune transformer language models using TRL (Transformers Reinforcement Learning). Supports SFT, DPO, GRPO, KTO, RLOO and Reward Model training via CLI commands.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[810,811,812,813,814],{"name":22,"slug":23,"type":14},{"name":16,"slug":17,"type":14},{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},{"name":19,"slug":20,"type":14},18850,"https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Ftrl","2026-04-06T18:25:32.746828",{"slug":819,"name":819,"fn":820,"description":821,"org":822,"tags":823,"stars":832,"repoUrl":833,"updatedAt":834},"hf-cli","manage Hugging Face Hub resources via CLI","Hugging Face Hub CLI (`hf`) for downloading, uploading, and managing models, datasets, spaces, buckets, repos, papers, jobs, and more on the Hugging Face Hub. Use when: handling authentication; managing local cache; managing Hugging Face Buckets; running or scheduling jobs on Hugging Face infrastructure; managing Hugging Face repos; discussions and pull requests; browsing models, datasets and spaces; reading, searching, or browsing academic papers; managing collections; querying datasets; configuring spaces; setting up webhooks; or deploying and managing HF Inference Endpoints. Make sure to use this skill whenever the user mentions 'hf', 'huggingface', 'Hugging Face', 'huggingface-cli', or 'hugging face cli', or wants to do anything related to the Hugging Face ecosystem and to AI and ML in general. Also use for cloud storage needs like training checkpoints, data pipelines, or agent traces. Use even if the user doesn't explicitly ask for a CLI command. Replaces the deprecated `huggingface-cli`.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[824,827,830,831],{"name":825,"slug":826,"type":14},"CLI","cli",{"name":828,"slug":829,"type":14},"Datasets","datasets",{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},10861,"https:\u002F\u002Fgithub.com\u002Fhuggingface\u002Fskills","2026-04-06T18:25:34.020855",{"slug":836,"name":836,"fn":837,"description":838,"org":839,"tags":840,"stars":832,"repoUrl":833,"updatedAt":848},"hf-cloud-aws-context-discovery","discover local AWS environment context","Discover the user's local AWS context (active profile, region, account ID, caller identity) at the start of any AWS task. Use this skill before any other AWS work — deploying to SageMaker, creating resources, calling AWS APIs, or anything that touches an AWS account. Use it especially when the user has not specified a region or profile explicitly, when they say things like \"use my AWS account\", \"deploy to AWS\", \"use my profile\", or when about to make any AWS CLI or SDK call. Never guess the region or account ID — always use this skill to read it from the local configuration first.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[841,844,845],{"name":842,"slug":843,"type":14},"AWS","aws",{"name":825,"slug":826,"type":14},{"name":846,"slug":847,"type":14},"Configuration","configuration","2026-07-08T05:55:33.716099",{"slug":850,"name":850,"fn":851,"description":852,"org":853,"tags":854,"stars":832,"repoUrl":833,"updatedAt":861},"hf-cloud-python-env-setup","set up Python environments for AWS","Set up an isolated Python environment for SageMaker \u002F AWS work, with the right Python version and current boto3. Use this skill whenever Python code will be executed for a SageMaker deployment, training job, or any AWS automation — including when about to run `pip install`, when about to invoke `boto3`, when creating or activating a virtualenv, or when the user asks to \"set up the environment\". Never use system Python and never `pip install` into it. Always isolate. This skill prevents the most common failure modes: wrong Python version, dependency conflicts, and stale SDKs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[855,856,857,860],{"name":22,"slug":23,"type":14},{"name":842,"slug":843,"type":14},{"name":858,"slug":859,"type":14},"Engineering","engineering",{"name":19,"slug":20,"type":14},"2026-07-08T05:55:32.505017",{"slug":863,"name":863,"fn":864,"description":865,"org":866,"tags":867,"stars":832,"repoUrl":833,"updatedAt":874},"hf-cloud-sagemaker-deployment-planner","plan model deployments to Amazon SageMaker","Plan and coordinate the deployment of a model to Amazon SageMaker AI. Use this skill whenever the user wants to deploy, host, serve, or expose a model on SageMaker or AWS — including phrases like \"deploy a model\", \"host this LLM on AWS\", \"serve this embedding model\", \"deploy a reranker\", \"deploy a text-to-image \u002F diffusion model\", \"host this for async inference\", \"create an endpoint\", \"serve my fine-tuned model\", or any request that involves making a model available for inference on AWS. Use this even when the user is vague (e.g. \"I just want to get this running on AWS, you figure it out\"). Works for text-generation LLMs, embedding models, rerankers, classifiers, text-to-image \u002F diffusion models — picks the right serving stack and chooses between real-time and async inference. This is the entry-point skill for SageMaker deployment work — it asks clarifying questions, picks a deployment pathway, and coordinates the other deployment skills.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[868,869,870,873],{"name":22,"slug":23,"type":14},{"name":842,"slug":843,"type":14},{"name":871,"slug":872,"type":14},"Deployment","deployment",{"name":9,"slug":13,"type":14},"2026-07-08T05:55:37.387689",{"slug":876,"name":876,"fn":877,"description":878,"org":879,"tags":880,"stars":832,"repoUrl":833,"updatedAt":889},"hf-cloud-sagemaker-iam-preflight","configure SageMaker IAM roles","Ensure a usable SageMaker execution role exists before deploying or training. Use this skill whenever about to create a SageMaker endpoint, model, training job, or any resource that requires an execution role. Use it especially when the user has not provided a role ARN explicitly, when scripts are about to call `iam:CreateRole`, or when an AccessDenied error mentions an IAM action. Never blindly call `iam:CreateRole` — always check for existing roles first. This skill prevents the most common SageMaker deployment failure: trying to create IAM resources from an SSO principal that has no IAM write permissions.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[881,882,883,886],{"name":22,"slug":23,"type":14},{"name":842,"slug":843,"type":14},{"name":884,"slug":885,"type":14},"Permissions","permissions",{"name":887,"slug":888,"type":14},"Security","security","2026-07-08T05:55:34.948771",{"slug":891,"name":891,"fn":892,"description":893,"org":894,"tags":895,"stars":832,"repoUrl":833,"updatedAt":903},"hf-cloud-sagemaker-production-defaults","create production-ready SageMaker endpoints","Create a SageMaker endpoint (real-time or async) with autoscaling, CloudWatch alarms, and tagging enabled by default. Use this skill whenever about to create a SageMaker endpoint, write deployment code that calls `create_endpoint`, or finalize a deployment after the image URI and IAM role are known. Provides deploy.py for real-time endpoints and deploy_async.py for async endpoints (with genuine scale-to-zero support). This is the last step in the SageMaker deployment workflow. Never generate a bare `create_endpoint` call without these defaults — endpoints without autoscaling or alarms are demos, not deployments.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[896,897,898,899,900],{"name":22,"slug":23,"type":14},{"name":842,"slug":843,"type":14},{"name":871,"slug":872,"type":14},{"name":9,"slug":13,"type":14},{"name":901,"slug":902,"type":14},"Monitoring","monitoring","2026-07-08T05:55:38.664702",{"slug":905,"name":905,"fn":906,"description":907,"org":908,"tags":909,"stars":832,"repoUrl":833,"updatedAt":913},"hf-cloud-serving-image-selection","select SageMaker serving containers","Pick the right serving container for a SageMaker model deployment and find its current image URI. Use this skill whenever about to deploy a model to a SageMaker endpoint and an image URI needs to be chosen — including when the user says \"deploy this LLM\", \"host this HuggingFace model\", \"serve this fine-tuned model\", \"deploy this embedding model\", \"host a reranker\", \"serve a sentence-transformers model\", or when about to hardcode any container URI in deployment code. HuggingFace-curated Deep Learning Containers are ALWAYS preferred: HuggingFace vLLM (LLMs and generative rerankers), HuggingFace vLLM-Omni (multimodal), TEI (embeddings\u002Fcross-encoder rerankers), HF Inference Toolkit (other transformers). Generic images (AWS vLLM, DJL-LMI, SGLang) are used only when no HuggingFace image is compatible — never merely because they carry a newer version. Never hardcode a container URI from memory and never default to TGI. Prevents stale-image failures and wrong-region URIs.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[910,911,912],{"name":22,"slug":23,"type":14},{"name":842,"slug":843,"type":14},{"name":871,"slug":872,"type":14},"2026-07-08T05:55:36.173465",{"slug":915,"name":915,"fn":916,"description":917,"org":918,"tags":919,"stars":832,"repoUrl":833,"updatedAt":926},"hf-mcp","access Hugging Face Hub via MCP","Use Hugging Face Hub via MCP server tools. Search models, datasets, Spaces, papers. Get repo details, fetch documentation, run compute jobs, and use Gradio Spaces as AI tools. Available when connected to the HF MCP server.",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[920,921,922,923],{"name":828,"slug":829,"type":14},{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},{"name":924,"slug":925,"type":14},"MCP","mcp","2026-04-06T18:25:50.364185",{"slug":928,"name":928,"fn":929,"description":930,"org":931,"tags":932,"stars":832,"repoUrl":833,"updatedAt":940},"hf-mem","estimate memory for Hugging Face models","Hugging Face CLI to estimate the required memory to load Safetensors or GGUF model weights for inference from the Hugging Face Hub",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[933,934,935,936,937],{"name":22,"slug":23,"type":14},{"name":825,"slug":826,"type":14},{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},{"name":938,"slug":939,"type":14},"Performance","performance","2026-06-13T07:23:57.101435",{"slug":942,"name":942,"fn":943,"description":944,"org":945,"tags":946,"stars":832,"repoUrl":833,"updatedAt":955},"huggingface-best","find and compare Hugging Face models","Use when the user asks about finding the best, top, or recommended model for a task, wants to know what AI model to use, or wants to compare models by benchmark scores. Triggers on: \"best model for X\", \"what model should I use for\", \"top models for [task]\", \"which model runs on my laptop\u002Fmachine\u002Fdevice\", \"recommend a model for\", \"what LLM should I use for\", \"compare models for\", \"what's state of the art for\", or any question about choosing an AI model for a specific use case. Always use this skill when the user wants model recommendations or comparisons, even if they don't explicitly mention HuggingFace or benchmarks.\n",{"slug":8,"name":9,"logoUrl":10,"githubOrg":8},[947,950,951,952],{"name":948,"slug":949,"type":14},"Analytics","analytics",{"name":9,"slug":13,"type":14},{"name":795,"slug":796,"type":14},{"name":953,"slug":954,"type":14},"Research","research","2026-04-24T05:09:45.870658",37]