huggingface/transformers

Pull Requests Commits

Avoid pipeline test failing related to Hub call (#37170)

ydshieh committed 341 days ago

Verified 35253076

Fixes the inconsistency of the optionality of attention_mask (#37153)

Zephyr271828 committed 341 days ago

Verified bf41e54f

Refactor attention for SigLIP based models (#36981)

qubvel committed 341 days ago

Verified 3249c5dc

fix XPU UT error case brough by RNG difference btw XPU and CUDA (#37121)

yao-matrix committed 341 days ago

Verified 24e311f4

[`ModernBERT`] Never save 'reference_compile' config; should be set based on end user (#36305)

tomaarsen committed 341 days ago

Verified 897ff9af

Make canine model exportable by removing unncessary complicated logic (#37124)

tugsbayasgalan committed 341 days ago

Verified c0bd8048

Only count num items in batch when needed (#36867)

IlyasMoutawwakil committed 342 days ago

Verified 60b75d99

Convert `_VALID_DICT_FIELDS` to class attribute for shared dict parsing in subclasses (#36736)

Tavish9 committed 342 days ago

Verified fac70ff3

Use public export API on torch 2.5 and future (#36781)

guangy10 committed 342 days ago

Verified ae34bd75

enable `test_assisted_decoding_in_different_gpu` test on XPU (#37120)

yao-matrix committed 342 days ago

Verified 8f6b27eb

Fix llava xpu tests. (#37130)

jiqing-feng committed 342 days ago

Verified 737cbd21

add gpt2 test on XPU (#37028)

jiqing-feng committed 342 days ago

Verified 3a6ab46a

Fix std initialization in Idefics variants (#37100)

yaswanth19 committed 342 days ago

Verified 4b13a029

Fix more inefficient PT operations (#37060)

cyyever committed 342 days ago

Verified 786d9c5e

Refactor `return_dict` logic to remove complicated if/else paths (#36794)

qubvel committed 342 days ago

Verified a1e389e6

Remove low_cpu_mem_usage and _fast_init (#36963)

Cyrilvallez committed 342 days ago

Verified f304318f

[qwen3] fix generation tests (#37142)

zucchini-nlp committed 342 days ago

Verified 88056004

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

FightingZhen committed 342 days ago

Verified e686fed6

ydshieh committed 342 days ago

Verified a03cee7a

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

guangy10 committed 343 days ago

Verified 3b07ca78

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

faaany committed 343 days ago

Verified 475664e2

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

geetu040 committed 343 days ago

Verified 0710e9b1

Remove deprecated code (#37059)

cyyever committed 343 days ago

Verified f99c279d

RWKV: fix mask warning typo (#37114)

RobinKa committed 343 days ago

Verified d1efaf03

Fix Gemma3 embedding scaling (#37109)

gau-nernst committed 343 days ago

Verified 19919689

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

huismiling committed 343 days ago

Verified d0b65bb4

fix whisper re-compile (#36712)

jiqing-feng committed 343 days ago

Verified ad63d20d

enable tp on CPU (#36299)

jiqing-feng committed 343 days ago

Verified 286393fb

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

Qubitium committed 343 days ago

Verified 4705b04c

Support passing flash_attn_kwargs when gradient_checkpointing is enabled (#37037)

efsotr committed 343 days ago

Verified 2b4734bd

Newer Older