huggingface/transformers

Pull Requests Commits

Expand npu device for KernelConfig (#42358)

zheliuyu committed 25 days ago

Verified 6b4b7bf8

Protect `torch.distributed` imports (#42361)

Cyrilvallez committed 25 days ago

Verified 0cc848f8

Fix typos (#42354)

omahs committed 25 days ago

Verified fd20cdc2

Fix reference to imagenet 1k dataset (#42348)

JuanFKurucz committed 25 days ago

Verified 1f022739

Fix documentation reference to pytorch max memory allocated (#42350)

JuanFKurucz committed 25 days ago

Verified 2e0457e6

Fix reference to yelp dataset (#42349)

JuanFKurucz committed 25 days ago

Verified acae07ab

Fix tied weight for Bart (for BC) (#42355)

Cyrilvallez committed 25 days ago

Verified 2f7747c3

Fix typo - indentation in JSON dump example (#42332)

anthonyylee committed 25 days ago

Verified 1ae3e5bb

[loading] Re-add and improve disk offloading support (#42242)

Cyrilvallez committed 25 days ago

Verified af6a36a3

Auto convert tekken.json (#42299)

ArthurZucker committed 25 days ago

Verified 6940b44d

fix(benchmarks): correct sdpa_backend inconsistency and attn_implementation for continuous batching (#42339)

engmohamedsalah committed 25 days ago

Verified 00ab75e6

Gemma3 hybrid fix (#42287)

remi-or committed 25 days ago

Verified 3410ba9b

Fix ChineseCLIPModel.get_text_features (#42351)

JiangJQ2000 committed 26 days ago

Verified f7e964e5

fix bug when gemma3n model run on multiple device (#42303)

kaixuanliu committed 26 days ago

Verified 122a6a3a

[`Attn Masks`] Lift bidirectional mask restriction on eager (#42325)

vasqu committed 28 days ago

Verified bdee0889

[XPU] Add flash_attn2 support for XPU (#41956)

YangKai0616 committed 28 days ago

Verified 07bfd2f8

[`CI`] Skip `EfficientLoFTR` test (#42327)

vasqu committed 28 days ago

Verified 9162e190

Correctly create tied key mapping in post_init, and dynamic tie weight (#42270)

Cyrilvallez committed 28 days ago

Verified ce7a5e00

Allow LayoutLMV3Processor to accept rescale_factor (#42305)

Rocketknight1 committed 28 days ago

Verified f15b95e6

[loading] Use fewer threads by default for much better performances (#42324)

Cyrilvallez committed 28 days ago

Verified 73d5c2d7

Fix gpt2 modeling tests (#42321)

Abdennacer-Badaoui committed 28 days ago

Verified 923cebd2

HF Trainer: ALST/Ulysses sequence parallelism integration via HF Accelerate (#41832)

stas00 committed 28 days ago

Verified 7e0ea699

Fix typo from side_dict to size_dict (#42319)

nihui committed 28 days ago

Verified afdc40db

[loading] Fix device detection (#42323)

Cyrilvallez committed 28 days ago

Verified 4799952d

Merge branch 'main' into add-missing-truncate

SunMarc committed 28 days ago

Verified 3848942c

fix tests/models/xcodec/test_modeling_xcodec.py::XcodecIntegrationTest (#42272)

sywangyi committed 29 days ago

Verified f2738ee3

Fix post processing methods in keypoints matching models (#42018)

yonigozlan committed 29 days ago

Verified decde58e

Add Pix2Struct fast image processor (#42020)

yonigozlan committed 29 days ago

Verified e04c7a60

[SAM3 Video] Add support for multi prompts (#42293)

yonigozlan committed 29 days ago

Verified c3fb1b1a

fix(granitemoe*): Only create block_sparse_moe if num_local_experts > 0 (#42036)

gabe-l-hart committed 29 days ago

Verified a1afecae

Newer Older