huggingface/transformers

Pull Requests Commits

ydshieh committed 222 days ago

22b7f7e0

ydshieh committed 222 days ago

9a781a49

ydshieh committed 222 days ago

def8ea56

ydshieh committed 223 days ago

fc3f99f8

ydshieh committed 223 days ago

42f0074d

Fix Bad Outputs in Fast Path for GraniteMoeHybrid (#39033)

alex-jw-brooks committed 223 days ago

Verified 5995cfa0

Granite speech speedup + model saving bugfix (#39028)

avihu111 committed 223 days ago

Verified 22b0a898

[tests] remove TF tests (uses of `require_tf`) (#38944)

gante committed 224 days ago

Verified 1d45d90e

Two ReDOS fixes (#39013)

Rocketknight1 committed 224 days ago

Verified d37f7517

[Kyutai-STT] correct model type + model id (#39035)

eustlb committed 224 days ago

Verified 551e48f1

Add SmolLM3 (#38755)

anton-l committed 224 days ago

Verified dad0e87c

refactor: remove custom BarkLayerNorm (#39003)

eginhard committed 224 days ago

Verified 3233e9b7

Fix grammatical error in models documentation (#39019)

marcndo committed 224 days ago

Verified 3c1d4dfb

Remove script datasets in tests (#38940)

lhoestq committed 224 days ago

Verified 858f9b71

fix gemma3 grad acc (#37208)

SunMarc committed 224 days ago

Verified 3c322c9c

fix: astronomical loss with ModernBERT when using gradient checkpointing (#38982) (#38983)

umarbutler committed 224 days ago

Verified 860b898d

Support for Flash Attention 3 (#38972)

EduardDurech committed 224 days ago

Verified a2eb75c8

Fix the seamless_m4t cannot work on Gaudi (#38363)

yuanwu2017 committed 224 days ago

Verified de98fb25

[Model] add dots1 (#38143)

redmoe-moutain committed 224 days ago

Verified 7503cb91

Encoder-Decoder Gemma (#38332)

bzhangGo committed 224 days ago

Verified 3ef88969

GLM-4.1V Model support (#38431)

zRzRzRzRzRzRzR committed 224 days ago

Verified af987026

Drop unnecessary tokens in GPT2Model generation (#39016)

null-pointer-access committed 224 days ago

Verified 7b380738

[video processor] support torchcodec and decrease cuda memory usage (#38880)

zucchini-nlp committed 224 days ago

Verified e212ff9e

[AutoModelForMaskGeneration] Remove duplicate code (#38622)

NielsRogge committed 224 days ago

Verified 11d0feac

Fix graph break in torch.compile when using FA2 with attention_mask=None and batch size > 1 (#37332)

efsotr committed 224 days ago

Verified 3ee72af6

Add zero dim tensor check when using flash_attention (#38280)

ranzhejiang committed 224 days ago

Verified ae32f1ad

[LightGlue] Fixed attribute usage from descriptor_dim to keypoint_detector_descriptor_dim (#39021)

sbucaille committed 224 days ago

Verified ca402e21

Add Hugging Face authentication procedure for IDEs (PyCharm, VS Code,… (#38954)

marcndo committed 224 days ago

Verified 48b6ef02

[HPU][Critical Issue Fix] ThreadPool instead of Pool for parallel pre-processing (#39002)

dsmertin committed 224 days ago

Verified ea9a3092

Skip sdpa dispatch on flash test due to unsupported head dims (#39010)

ivarflakstad committed 225 days ago

Verified 995666ed

Older