huggingface/transformers

Pull Requests Commits

[Feature] Support using FlashAttention2 on Ascend NPU (#36696)

FightingZhen committed 333 days ago

Verified e686fed6

ydshieh committed 333 days ago

Verified a03cee7a

Export T5 (encoder-decoder) to ExecuTorch (#36486)

guangy10 committed 333 days ago

Verified 3b07ca78

[tests] remove cuda-only test marker in `AwqConfigTest` (#37032)

faaany committed 333 days ago

Verified 475664e2

Create and Expose SamVisionModel as public for better accessibility (#36493)

geetu040 committed 333 days ago

Verified 0710e9b1

Remove deprecated code (#37059)

cyyever committed 333 days ago

Verified f99c279d

RWKV: fix mask warning typo (#37114)

RobinKa committed 333 days ago

Verified d1efaf03

Fix Gemma3 embedding scaling (#37109)

gau-nernst committed 333 days ago

Verified 19919689

[MLU] Fix FA2 check error, remove deepspeed-mlu deps. (#36159)

huismiling committed 333 days ago

Verified d0b65bb4

fix whisper re-compile (#36712)

jiqing-feng committed 333 days ago

Verified ad63d20d

enable tp on CPU (#36299)

jiqing-feng committed 333 days ago

Verified 286393fb

Fix 4090/ada not detected as having FP8 support (#37067)

Qubitium committed 333 days ago

Verified 4705b04c

Support passing flash_attn_kwargs when gradient_checkpointing is enabled (#37037)

efsotr committed 333 days ago

Verified 2b4734bd

Gaudi: Fix the pipeline failed issue with hpu device (#36990)

yuanwu2017 committed 333 days ago

Verified bd41b9c1

Adding Qwen3 and Qwen3MoE (#36878)

bozheng-hit committed 333 days ago

Verified 6acd5aec

🌐 [i18n-KO] Translated `qwen2_vl.md` to Korean (#36750)

MinJu-Ha committed 333 days ago

Verified 0d6a60fe

ydshieh committed 335 days ago

Verified b7fc2daf

[Cache] rename dtype attribute 🚨 🚨 (#37044)

gante committed 335 days ago

Verified bab605dd

[generate] beam search -- fix output cropping (#37080)

gante committed 335 days ago

Verified 9fd94760

fixed typo. (#37057)

zhanluxianshen committed 335 days ago

Verified 257bc670

Fix AttentionInterface following feedback (#37010)

Cyrilvallez committed 335 days ago

Verified 2bea6bf2

Fix state_dict map location when quantized (#37086)

Cyrilvallez committed 335 days ago

Verified a86dad56

Update w/ new account (#37084)

muellerzr committed 335 days ago

Verified d6064754

fix tied weigths issue (#37031)

ydshieh committed 335 days ago

Verified 581cf96e

[WIP] add deepseek-v3 (#35926)

bzantium committed 336 days ago

Verified eca74d13

[blip-2] Fix dtype mismatch when keep in fp32 (#37068)

zucchini-nlp committed 336 days ago

Verified 52cc204d

Change deprecated PT functions (#37041)

cyyever committed 336 days ago

Verified aa3778af

Fix some typos about benchmark scripts. (#37027)

zhanluxianshen committed 336 days ago

Verified c90e6e96

Use `lru_cache` for tokenization tests (#36818)

ydshieh committed 336 days ago

Verified 1fcaad6d

fix: AttributeError: 'LlavaProcessor' object has no attribute 'image_token_id' (#37026)

jp1924 committed 336 days ago

Verified 3af425d4

Newer Older