huggingface/transformers

Pull Requests Commits

abidlabs committed 19 hours ago

73531629

abidlabs committed 19 hours ago

fc7e1b23

Fix formatting of trackio model tag

abidlabs committed 20 hours ago

Verified ed34208a

Fix tests trainer again (#42933)

SunMarc committed 1 day ago

Verified 70179949

[`Ernie 4.5`] Ernie VL models (#39585)

vasqu committed 1 day ago

Verified a8a22624

[`Generation`] Fix default overwrite for non-`None` defaults (#42958)

vasqu committed 1 day ago

Verified f218ed21

Fix infinity in JSON serialized files (#42959)

LysandreJik committed 1 day ago

Verified d14d99ea

Add runner specification to CodeQL workflow (#42955)

paulinebm committed 1 day ago

Verified d54d78f7

[loading] Really initialize on meta device for huge perf gains (#42941)

Cyrilvallez committed 2 days ago

Verified bb9357ff

fix error: 'BlockMask' object has no attribute 'dtype' for lasr model (#42802)

kaixuanliu committed 2 days ago

Verified 789226c1

fix concat order (#42946)

eustlb committed 2 days ago

Verified a4d62291

[kernels] Fix failling tests (#42953)

MekkCyber committed 2 days ago

Verified cf0f071e

[Quantization] CI green by end of year (#42951)

MekkCyber committed 2 days ago

Verified b5eea347

Updated `backbone_config` docstrings and type annotations (#42927)

simonreise committed 2 days ago

Verified 3e4baf8e

[docs] dtype (#42883)

stevhliu committed 2 days ago

Verified b93f2e3a

Load generation config from nested configs (#42922)

zucchini-nlp committed 2 days ago

Verified f2c6d2ad

🚨 Fix ConvNeXt image processor default interpolation to BICUBIC (#42934)

lukepayyapilli committed 2 days ago

Verified 60634caa

fix(tvp): add missing type_vocab_size parameter to TvpConfig (#42928)

majiayu000 committed 2 days ago

Verified b62e5b3e

[Quantization] Removing misleading int8 quantization in Finegrained FP8 (#42945)

MekkCyber committed 2 days ago

Verified 4dc60fd3

Fp8 dq (#42926)

SunMarc committed 2 days ago

Verified af91c0ba

Add buffers to `_init_weights` for ALL models (#42309)

Cyrilvallez committed 2 days ago

Verified 537c2e3d

rewrite _process_parameter_type in auto_docstring.py to improve usability (#42431)

Yacklin committed 2 days ago

Verified 728f34c3

Update `param_element_size` (#42818)

SunMarc committed 2 days ago

Verified dd8057af

[Tests] Fix CompressedTensors tests (#42935)

kylesayrs committed 3 days ago

Verified 0a846542

[CB] Allow block sharing in hybrid models (#42877)

remi-or committed 3 days ago

Verified 04e78e67

[Quantization] Misc tests fixes (#42940)

MekkCyber committed 3 days ago

Verified d7dd443a

[Quantization] rm _pre_quantization_dtype from quantization tests (#42939)

MekkCyber committed 3 days ago

Verified d3d4b629

🚨 Generation config defaults are now `None` (#42702)

zucchini-nlp committed 3 days ago

Verified a81e04a9

Document new default shard size + dropped unsafe serialization (#42904)

Wauplin committed 3 days ago

Verified bdaddb6f

Improve BatchFeature (.to() works on lists/nested lists of tensors, auto-skip non array conversion) (#42884)

yonigozlan committed 3 days ago

Verified 9f583b1b

Older