huggingface/transformers

Pull Requests Commits

serve: load GGUF models end-to-end

ArthurZucker committed 87 days ago

182d7aa3

GGUF: uniform meta-time swap, no rename, bind kernels post-load

ArthurZucker committed 104 days ago

e0aae61b

GGUF: expose header metadata without materializing tensors

ArthurZucker committed 104 days ago

e09ea879

GGUF: de-offset norms via keep-in-fp32 instead of pre-applying

ArthurZucker committed 104 days ago

b11c4d7a

GGUF: fix MPS dequant byte-corruption and make norm de-offset data-driven

ArthurZucker committed 106 days ago

8d87b3c4

GGUF: simplify quantizer plumbing and fix swap-plan rename bug

ArthurZucker committed 106 days ago

ccbcb1cf

Merge branch 'main' of github.com:huggingface/transformers into gguf-matmul-kernels

ArthurZucker committed 107 days ago

c76d77b9

Fix unhandled exception noise from background safetensors conversion thread (#45752)

dhruv7477 committed 107 days ago

Verified 855aad56

fix: cache warmup `RuntimeError` on mps (#46239)

McPatate committed 107 days ago

Verified 5390fc30

Add Expectations for pipeline token classification tests (#46151)

kaixuanliu committed 107 days ago

Verified 7b87f642

[docs] fix auto-add release dates (#46283)

zucchini-nlp committed 107 days ago

Verified 3ceb0e11

🚨 Fix float16 overflow in Gemma4 vision pooler (#46277)

Bluear7878 committed 107 days ago

Verified 834c0460

Make more tests work with read-only cache (#46299)

ydshieh committed 109 days ago

Verified 39603d0e

Separate pip command syntax for notebook and CLI tabs in Quickstart (#46243)

pvelayudhan committed 110 days ago

Verified bb0ee88a

Romanian translation of README.md, index.md, installation.md, _config.py and quicktour.md. (#46166)

filipinescu committed 110 days ago

Verified 56153cc4

fix series of bugs for model parallel beam search (#46280)

kaixuanliu committed 110 days ago

Verified 8d974fb6

Fall back to flat kwarg when modality dict is passed without it (#46195)

Ace3Z committed 110 days ago

Verified 420c80aa

Fix load_adapter OOM caused by full-model warmup sizing (#46145)

Yooniel committed 110 days ago

Verified c220ea9e

Replace assert with raise ImportError for optuna/ray dependency checks (#46263)

SebTardif committed 110 days ago

Verified e1a37d29

chore(linter): respect TRF017 modeling rule (#46260)

tarekziade committed 110 days ago

Verified e977b060

Delete dead code in qwen-vl series (#45827)

zucchini-nlp committed 110 days ago

Verified 84a6f0b9

qa: fix ty caching and align CI with local run (#46278)

tarekziade committed 110 days ago

Verified fda83a31

Guard DeviceMesh import in continuous batching (#46205)

danyalahmed1995 committed 110 days ago

Verified c3b7df0e

Processor compatibility with vLLM (#46258)

zucchini-nlp committed 110 days ago

Verified 45570925

Fix PR CI workflow cancellation condition (#46276)

ydshieh committed 110 days ago

Verified 96d3ecce

[fix] toctree (#46106)

stevhliu committed 111 days ago

Verified 7b89db28

add more generic support for distributed trainer tests (#46109)

kaixuanliu committed 111 days ago

Verified c5292db6

add XPU Expectations for florence2 and lfm2_vl model test (#46275)

kaixuanliu committed 111 days ago

Verified da78b36f

Fix model parallel issue for altclip model and ChineseClip model (#45487)

kaixuanliu committed 111 days ago

Verified 08f13097

feat: Add support for JetBrains' `Mellum` v2 code generation model (#46112)

shadeMe committed 111 days ago

Verified 9cfa266f

Older