huggingface/transformers

Pull Requests Commits

file not needed anymore

tarekziade committed 2 hours ago

510a1860

tarekziade committed 3 hours ago

ad9b4a4c

added CI trigger for utils/ changes cc @ydshieh

tarekziade committed 3 hours ago

049d29f9

tarekziade committed 3 hours ago

643af719

use dict to hold values so we can di a single pass

tarekziade committed 4 hours ago

43585a11

APLv2 headers everywhere

tarekziade committed 4 hours ago

ba0275dc

tarekziade committed 4 hours ago

8d8c0730

refactored module linter

tarekziade committed 4 hours ago

fef84921

Fix backward compatibility for full path imports of Fast Image Processors (#44926)

yonigozlan committed 4 hours ago

Verified d03945fb

chore(typing): added rule 11 (#44865)

tarekziade committed 4 hours ago

Verified 29db503c

fix: improve processor loading performance by avoiding redundant tokenizer parsing (#44927)

ydshieh committed 6 hours ago

Verified b1527a32

fix(camembert): add tie_word_embeddings=True to CamembertConfig (#44931)

r266-tech committed 6 hours ago

Verified 9dc8d8aa

Support SizeDict import in get_size_dict (#44903)

yonigozlan committed 6 hours ago

Verified 687a70da

fix `processing_utils.py`: avoid deepcopying tokenizer in `ProcessorMixin` to improve performance (#44894)

ydshieh committed 7 hours ago

Verified a8683756

fix: set `clean_up_tokenization_spaces=False` in Llama 3 tokenizer conversion (#44914)

maxsloef-goodfire committed 8 hours ago

Verified 55cc1a7f

Add big angry code agent warnings! (#44890)

Rocketknight1 committed 22 hours ago

Verified 394b1a0e

[docs] model cards (#44837)

stevhliu committed 2 days ago

Verified 3a3b59cb

[Model] Add UVDoc Model Support (#43385)

XingweiDeng committed 2 days ago

Verified 52bc9b7e

Add backward compatibility for direct imports from legacy `image_processing_utils_fast` (#44897)

yonigozlan committed 2 days ago

Verified e168f86e

[`FA4`] Add kernels fallback (#44797)

vasqu committed 2 days ago

Verified e6ed96c7

Bump kernels version dependency to avoid crashes (#44887)

Cyrilvallez committed 2 days ago

Verified a269c990

[Model] Add SLANeXt Model Support (#43707)

liu-jiaxuan committed 2 days ago

Verified d4f88c20

Fix core dumped when `NemotronH` is torch compiled (#44854)

ydshieh committed 3 days ago

Verified e8712550

Fix several based models' pipeline parallel support (#44699)

hmellor committed 3 days ago

Verified 20a233bd

fix(testing): Fix PaliGemma 2 and PaddleOCR-VL test failures on main (#44765)

harshaljanjani committed 3 days ago

Verified 81128509

Fix dtype guessing from state dict (#44883)

Cyrilvallez committed 3 days ago

Verified 82db888e

Add missing dunder methods to `SizeDict` (#44884)

hmellor committed 3 days ago

Verified 7cd9b985

Fix VL model rope_deltas batch size mismatch in online RL training (#44873)

sergiopaniego committed 3 days ago

Verified b7164eca

Fix `layer_types` type hint for `AFMoE` and `Llama4` (#44874)

hmellor committed 3 days ago

Verified 1229e90d

Align lfm2 cache to other mamba caches (#44866)

Cyrilvallez committed 3 days ago

Verified cef28300

Older