huggingface/transformers

Pull Requests Commits

just push all stateless functions

ArthurZucker committed 183 days ago

b25f0048

Update tiny-agents example (#39245)

Wauplin committed 183 days ago

Verified bf203aa9

adjust input and output texts for test_modeling_recurrent_gemma.py (#39190)

kaixuanliu committed 183 days ago

Verified c4e39ee5

enable xpu on kv-cache and hqq doc (#39246)

jiqing-feng committed 183 days ago

Verified 14cba7ad

Fix patch helper (#39216)

Cyrilvallez committed 183 days ago

Verified 32db48db

RotaryEmbeddings change `is not None` -> `isinstance(..., dict)` (#39145)

qubvel committed 183 days ago

Verified a3618d48

fix `fastspeech2_conformer` tests (#39229)

ydshieh committed 183 days ago

Verified 9b09fe47

[bugfix] fix flash attention 2 unavailable error on Ascend NPU (#39166)

FightingZhen committed 183 days ago

Verified 00e9efce

[modular] Simplify logic and docstring handling (#39185)

Cyrilvallez committed 183 days ago

Verified 056fa73f

Make _compute_dynamic_ntk_parameters exportable (#39171)

xadupre committed 183 days ago

Verified f16fbfb8

fix bug using FSDP V1 will lead to model device not properly set (#39177)

kaixuanliu committed 183 days ago

Verified 4243bb84

Don't send new comment if the previous one is less than 30 minutes (unless the content is changed) (#39170)

ydshieh committed 183 days ago

Verified 34c16167

fix typo in Gemma3n notes (#39196)

davanstrien committed 183 days ago

Verified b8f397e4

[modular] Follow global indexing and attribute setting, and their dependencies (#39180)

Cyrilvallez committed 183 days ago

Verified 5348fbc0

Fix missing fast tokenizer/image_processor in whisper/qwen2.5-omni processor (#39244)

Isotr0py committed 183 days ago

Verified 8570bc29

[vjepa2] replace einsum with unsqueeze (#39234)

xenova committed 183 days ago

Verified b283d52f

Expectations re-order and corrected FA3 skip (#39195)

remi-or committed 183 days ago

Verified a325409a

[video processors] Support float fps for precise frame sampling (#39134)

zrohyun committed 183 days ago

Verified b0a8e0b8

Refactor the way we handle outputs for new llamas and new models (#39120)

ArthurZucker committed 185 days ago

Verified ca7e1a37

Update expected values (after switching to A10) - part 8 - Final (#39220)

ydshieh committed 186 days ago

Verified e6a8063e

Update expected values (after switching to A10) - part 7 (#39218)

ydshieh committed 186 days ago

Verified cd8a041a

Add packed tensor format support for flex/sdpa/eager through the mask! (#39194)

Cyrilvallez committed 186 days ago

Verified 0cf27916

Update expected values (after switching to A10) - part 6 (#39207)

ydshieh committed 187 days ago

Verified 037755ed

Update expected values (after switching to A10) - part 5 (#39205)

ydshieh committed 187 days ago

Verified 1168f57a

Fix continuous batching in `transformers serve` (#39149)

LysandreJik committed 187 days ago

Verified 7d9e52f3

[serve] Cursor support, move docs into separate page, add more examples (#39133)

gante committed 187 days ago

Verified 85d93cc6

[typing] better return typehints for `from_pretrained` (#39184)

qubvel committed 187 days ago

Verified e15b06d8

Update expected values (after switching to A10) - part 4 (#39189)

ydshieh committed 187 days ago

Verified a25fc359

[`Dia`] Change ckpt path in docs (#39181)

vasqu committed 187 days ago

Verified b31e9d19

Fix many HPU failures in the CI (#39066)

IlyasMoutawwakil committed 187 days ago

Verified 18e0cae2

Older