huggingface/transformers

Pull Requests Commits

itazap committed 51 days ago

bb75abc6

itazap committed 51 days ago

5cd3b786

itazap committed 51 days ago

4cf1cc43

itazap committed 51 days ago

649d93a0

itazap committed 51 days ago

5d1cb134

Enable mlinter findings artifact for inline PR reviews (#48117)

ydshieh committed 52 days ago

Verified decba1d2

Fix ASR pipeline mono conversion for channels-last audio (fixes #47886) (#47888)

Kayvan-Zahiri committed 52 days ago

Verified bb8f2353

Fix CpmAnt loading: size lm_head to vocab_size (#48012)

jiqing-feng committed 52 days ago

Verified e872ca25

Delete old mlinter review comments before posting new ones (#48107)

ydshieh committed 52 days ago

Verified de795d93

Fix Qwen2.5-Omni / Qwen3-Omni-MoE generation with a compilable cache (#47872)

jiqing-feng committed 52 days ago

Verified 279dbcf4

[Video] Warn and return all frames when num_frames exceeds total_num_frames (#48074)

carlszk committed 52 days ago

Verified b67f702d

Accept artifact dir as argument in post_mlinter_review.py (#48106)

ydshieh committed 52 days ago

Verified 604da442

Improve group images by shape (#47964)

guarin committed 52 days ago

Verified 0e2ce39f

Fix mlinter artifact path (#48088)

ydshieh committed 52 days ago

Verified f2ef0b94

[Video] Fix convert_to_rgb channel slicing and alpha blending for RGBA videos (#48053)

Ultron09 committed 52 days ago

Verified e7e8b7fa

[Whisper] Fix decoder position IDs for left-padded batches in longform generation (#48028)

ydshieh committed 52 days ago

Verified e12c79c5

[PE] Skip test_sdpa_can_dispatch_on_flash for TimmWrapper-backed models (#48064)

ydshieh committed 53 days ago

Verified f9b76f2d

feat: add nvfp4 quantization (#47883)

drbh committed 53 days ago

Verified 80c66744

fix(pipeline): preserve model.generation_config precedence over pipeline defaults (#47752) (#47953)

nithin42 committed 53 days ago

Verified bc777261

[Gemma3] Update integration test expected values for A10G (#48036)

ydshieh committed 53 days ago

Verified ff09c381

Fallback from 'lanczos' to 'bicubic' when on cuda (#48026)

zucchini-nlp committed 53 days ago

Verified e8018693

[serge] Fix 2 integration tests for model `generation` failing with `import_or_config` (other (2)) (#48061)

sergereview[bot] committed 53 days ago

Verified 2dcc93fe

[serge] Fix 2 integration tests regressed by commit b9090ae58cda (PR #47096) (#48060)

sergereview[bot] committed 53 days ago

Verified 02d9442a

[GPT2] Fix encoder_attention_mask being silently discarded in cross-attention (#47946)

DavidJohnQuinlan committed 53 days ago

Verified e0aeab4c

[Fix] Small FA-related test failures in CB (#47341)

remi-or committed 53 days ago

Verified e4075b9e

fix: honor empty processor_kwargs={} in multimodal pipelines (#48044)

hungnnvidia committed 53 days ago

Verified 606e5010

[CircleCI] Enable CI for private forks, no-op for public repo (#48056)

ydshieh committed 53 days ago

Verified bea0343f

:rotating_light: Leftover processors (#47924)

zucchini-nlp committed 53 days ago

Verified 5068901e

Support `BatchFeature` in length-grouped samplers (#48034)

qgallouedec committed 53 days ago

Verified 24fb8989

Fix EOS for candidate generators (#47931)

pjordanandrsn committed 53 days ago

Verified 5b90fc37

Older