huggingface/transformers

Pull Requests Commits

[`CI`] Fix post merge ernie 4.5 (#39561)

vasqu committed 1 year ago

Verified 049a674e

[Fast image processors] Improve handling of image-like inputs other than images (segmentation_maps) (#39489)

yonigozlan committed 1 year ago

Verified b3ebc761

[`Ernie 4.5`] Add ernie text models (#39228)

vasqu committed 1 year ago

Verified b4115a42

Refactor embedding input/output getter/setter (#39339)

molbap committed 1 year ago

Verified 69b15826

🌐 [i18n-KO] Translated `perf_infer_gpu_multi.md` to Korean (#39441)

nevertmr committed 1 year ago

Verified 2da97f09

[Fast image processor] refactor fast image processor glm4v (#39490)

yonigozlan committed 1 year ago

Verified 82807e56

fix ndim check of device_mesh for TP (#39538)

winglian committed 1 year ago

Verified 4b4f04fc

Refactor `MambaCache` to `modeling_mamba.py` (#38086)

manueldeprada committed 1 year ago

Verified 1aa7256f

Fix Docstring of BarkProcessor (#39546)

st81 committed 1 year ago

Verified a419a402

use the enable_gqa param in torch.nn.functional.scaled_dot_product_at… (#39412)

sywangyi committed 1 year ago

Verified 9323d087

Fix missing initializations for models created in 2023 (#39239)

bvantuan committed 1 year ago

Verified 6b3a1f2f

Raise `TypeError` instead of ValueError for invalid types (#38660)

Sai-Suraj-27 committed 1 year ago

Verified 970d9a75

Fix pylint warnings (#39477)

cyyever committed 1 year ago

Verified 822c5e45

Fix Qwen Omni integration test (#39553)

Cyrilvallez committed 1 year ago

Verified dc017cd7

🚨🚨🚨 [Trainer] Enable `average_tokens_across_devices` by default in `TrainingArguments` (#39395)

Krish0909 committed 1 year ago

Verified fdc0566e

Rename `_supports_flash_attn_2` in examples and tests (#39471)

zucchini-nlp committed 1 year ago

Verified 8c102e2e

Fix the check in flex test (#39548)

Cyrilvallez committed 1 year ago

Verified 3a152e3a

Fix bad tensor shape in failing Hubert test. (#39502)

ebezzam committed 1 year ago

Verified 78fb2d27

GLM-4 Update (#39393)

zRzRzRzRzRzRzR committed 1 year ago

Verified 39ba5f3c

[qwen2 vl] fix packing with all attentions (#39447)

zucchini-nlp committed 1 year ago

Verified 344012b3

[gemma3] support sequence classification task (#39465)

zucchini-nlp committed 1 year ago

Verified e42681b4

Fix placeholders replacement logic in auto_docstring (#39433)

yonigozlan committed 1 year ago

Verified 34133d0a

Update SAM/SAM HQ attention implementation + fix Cuda sync issues (#39386)

yonigozlan committed 1 year ago

Verified 433d2a23

Improve @auto_docstring doc and rename `args_doc.py` to `auto_docstring.py` (#39439)

yonigozlan committed 1 year ago

Verified 541bed22

Add fast image processor SAM (#39385)

yonigozlan committed 1 year ago

Verified de0dd313

Fix BatchEncoding.to() for nested elements (#38985)

eginhard committed 1 year ago

Verified 561a79a2

[gemma3] Fix do_convert_rgb in image processors. (#39438)

MohitIntel committed 1 year ago

Verified f4d07656

[chat template] return assistant mask in processors (#38545)

zucchini-nlp committed 1 year ago

Verified bcc00919

[dependencies] Update `datasets` pin (#39500)

gante committed 1 year ago

Verified 328ca9cf

Slack CI bot: set default result for non-existing artifacts (#39499)

ahadnagy committed 1 year ago

Verified fb583777

Newer Older