huggingface/transformers

Pull Requests Commits

draft "ring_attention" in _attn_implementation

NouamaneTazi committed 230 days ago

dca6b1b3

cp compatible dataloader

NouamaneTazi committed 239 days ago

3dea815a

Fixing the example in generation strategy doc (#37598)

jeasinema committed 239 days ago

Verified e1f379bb

Deprecate modeling_utils.py classes (#37298)

qubvel committed 239 days ago

Verified 4f58fc9c

Add InternVL (2.5 MPO) (#35968)

yonigozlan committed 239 days ago

Verified a2450112

fix issue that some example with no trainer use accelerator.end_train… (#37435)

we1559 committed 239 days ago

Verified b0c6ff5e

fix 2 encoder_decoder issues on XPU (#37572)

yao-matrix committed 239 days ago

Verified 6f5014ac

[VLMs] use only `xxx_token_id` for multimodal tokens (#37573)

zucchini-nlp committed 239 days ago

Verified 2ba6b92a

Model debugger upgrades (#37391)

molbap committed 239 days ago

Verified 4afd3f48

[Gemma3] compile ✨ (#37447)

gante committed 240 days ago

Verified e5ac2308

enable 6 modeling cases on XPU (#37571)

yao-matrix committed 240 days ago

Verified a1b82563

enable 6 gemma2 cases on XPU (#37564)

yao-matrix committed 240 days ago

Verified 3cd6627c

Flag SpeechT5 flaky test (#37587)

molbap committed 240 days ago

Verified 049b75ea

[Bugfix] Fix flash-attention func param mismatch and softmax_scale default value mistake on Ascend NPU (#37575)

FightingZhen committed 240 days ago

Verified aa17cfb4

remove _run_third_party_device_tests (#37445)

jiqing-feng committed 240 days ago

Verified 14b3dbcf

Fix some GPU OOM after #37553 (#37591)

ydshieh committed 240 days ago

Verified f9742143

Gaudi: Add the bf16 support for hpu (#37568)

yuanwu2017 committed 240 days ago

Verified 438324c9

Fix Quark quantization config (#37578)

MekkCyber committed 240 days ago

Verified bb2a44ad

Update Phi4 converter (#37594)

Cyrilvallez committed 240 days ago

Verified 4acf692a

Ensure positive warm-up size (#37581)

Cyrilvallez committed 241 days ago

Verified 40cba20e

docs: fix typo (#37567)

tonyksong committed 241 days ago

Verified 346f1eeb

[phi4] update conversion (#37579)

zucchini-nlp committed 241 days ago

Verified 48dd89cf

Small fix on context manager detection (#37562)

Cyrilvallez committed 241 days ago

Verified 58e5e976

Fix qwen2audio wanr -> warn (#37559)

alex-jw-brooks committed 241 days ago

Verified c7d3cc67

[TimesFM] use the main revison instead of revision for integration test (#37558)

kashif committed 241 days ago

Verified dc06e7ce

[qwen-vl] Standardize config (#37268)

zucchini-nlp committed 241 days ago

Verified 3bc44eaa

[chat template] fix security vulnerability (#37523)

zucchini-nlp committed 241 days ago

Verified 4f96081a

Add Janus model (#36053)

yaswanth19 committed 241 days ago

Verified a2ef3cf5

All models can be initialized on meta device (#37563)

Cyrilvallez committed 241 days ago

Verified 688f4707

Bridgetower fast image processor (#37373)

rootonchair committed 241 days ago

Verified 0a83588c

Older