huggingface/transformers

Pull Requests Commits

fix: improve visibility of ValueError root causes in model config loading (#41972)

scottzh8 committed 62 days ago

Verified e76364d5

T5 migration to new masking interface (#41804)

Aravind-11 committed 63 days ago

Verified 2b8068c3

[`T5Gemma`] Fix cross attention cache (#41890)

vasqu committed 63 days ago

Verified 33c60a52

:rotating_light: Fix gradient checkpointing for several models and improve test robustness (#41818)

githubnemo committed 63 days ago

Verified fa22b569

Fix logic error in `prepare_inputs_for_generation` cache slicing condition (#41764)

albertvillanova committed 63 days ago

Verified f30c2250

Add dinov3 autobackbone (#41276)

vijayabhaskar-ev committed 63 days ago

Verified 496c2836

Enforce check_auto_docstring (#41635)

yonigozlan committed 63 days ago

Verified df45a92c

Avoid mutating user-provided arguments in preprocessing utils (#42126)

LeonardoEmili committed 63 days ago

Verified 3ff0e69f

Bugfix/remove emojis from print (#42091)

7amim committed 63 days ago

Verified 31839d74

DataCollatorForLanguageModeling warning error fixed (#42144)

mjaliz committed 63 days ago

Verified 2072f305

Fix T5Gemma module structure (#42145)

Cyrilvallez committed 63 days ago

Verified 3760afb2

fix: improve video processing fps assignment logic (#42009)

Xqle committed 63 days ago

Verified 3c0b2b10

update deps table (#42120)

ArthurZucker committed 63 days ago

Verified e869e9df

Remove unused functions in `image_transforms.py` (#42044)

yaswanth19 committed 64 days ago

Verified 37d48bbb

Fix MaskFormer/Mask2Former fast image processors (#41393)

yonigozlan committed 64 days ago

Verified 21913b2e

Fix model name test for compressed tensors (#42128)

SunMarc committed 64 days ago

Verified f028e934

Fix GPT-2 Flash Attention 2 generation with left-padding (#41966)

Abdennacer-Badaoui committed 64 days ago

Verified 4dd4a8fa

[`Attn Masks`] Non-vmap default for attention masks (#41852)

vasqu committed 64 days ago

Verified 03538a80

fix qwen2vl/qwen3vl video processor temporal padding when num_frames%temporal_patch_size!=1 (#42083)

yaogang2060 committed 64 days ago

Verified 700c48a2

📝 docs(smolvlm): fix variable name in batch inference example (#42123)

gorkachea committed 64 days ago

Verified 18a19dea

fix continuous batching issues, extend ut cases to xpu (#41830)

yao-matrix committed 64 days ago

Verified dba6aeb1

Fix base model prefix in VLMs (#42059)

zucchini-nlp committed 64 days ago

Verified 1c9077f6

[Bugfix] fix qwen3vl expand generation with video (#42089)

JJJYmmm committed 64 days ago

Verified 75674235

Correctly handle unbatched audio inputs in Gemma3nAudioFeatureExtractor (#42076)

kho committed 64 days ago

Verified 926c37aa

Fix return metadata checking logic (#42108)

Xqle committed 64 days ago

Verified f5630f9b

Revert "permissions worflows fix" (#42110)

ydshieh committed 66 days ago

Verified e8a6eb33

add xpu support in test_modeling_janus.py::JanusIntegrationTest::test… (#41986)

sywangyi committed 66 days ago

Verified 370fc65e

[v5] 🚨Refactor subprocessors handling in processors (#41633)

yonigozlan committed 67 days ago

Verified f065e402

Reinstate self.scaling in Gemma3nTextAttention (#41751)

RyanMullins committed 67 days ago

Verified 91d250ef

Fix Auto classes to support dynamically registered processors (#41865)

MilkClouds committed 67 days ago

Verified 7cb42801

Newer Older