huggingface/transformers

Pull Requests Commits

fix tp plan for multi node runs

ArthurZucker committed 2 days ago

299c53c3

[Deespeed] fix WeightConverter.convert() use (#43926)

kashif committed 2 days ago

Verified 360d7262

Use torch.xlogy (#44006)

cyyever committed 2 days ago

Verified 8bb26307

Improve use of torch.is_autocast_enabled (#43930)

cyyever committed 2 days ago

Verified 3955dc76

update python requirement to 3.10+ to match codebase (#44009)

mariam851 committed 2 days ago

Verified 33d9b320

docs: fix typos across documentation files (#43993)

saurav0369 committed 2 days ago

Verified b1141006

Add Voxtral Realtime (#43769)

eustlb committed 2 days ago

Verified b768d8b1

create .git-blame-ignore-revs file (#43982)

SunMarc committed 2 days ago

Verified 737f1777

stable grouped_mm API (#43977)

IlyasMoutawwakil committed 2 days ago

Verified 8ead2903

Remove `other_workflow_run_ids` for `issue_comment` in `utils/notification_service.py` (#44036)

ydshieh committed 2 days ago

Verified d4bea33f

Improve new failing test analysis for PR comment CI (#44033)

ydshieh committed 2 days ago

Verified 0dda1ffb

Remove unused squeeze from VJEPA2 embeddings rotation (#43984)

materight committed 5 days ago

Verified c9ea365a

Enhance JSON schema generation to support instance, static, and class methods (#43968)

qgallouedec committed 5 days ago

Verified 153894c0

Fix TypeError in dot_natural_key when state_dict keys have mixed types at same position (#43966)

shtse8 committed 5 days ago

Verified 44412f6e

Reorder Trainer methods (#43914)

SunMarc committed 5 days ago

Verified 8008e6c8

JaredforReal committed 5 days ago

62454529

Fix sync gradient (#43919)

SunMarc committed 5 days ago

Verified 0dfdd624

Inclusion of process_group in the gather_full_tensor function in tensor_parallel.py (#43932)

quic-meetkuma committed 5 days ago

Verified f73a4db3

Improve memory allocator during loading (#43945)

Cyrilvallez committed 5 days ago

Verified 3e4b8d63

update somes audio models (#43865)

Deep-unlearning committed 5 days ago

Verified a780072a

fix(models): Preserve custom token IDs through DiaConfig save and load (#43928)

harshaljanjani committed 5 days ago

Verified 403990ca

[Doc tests] Fix bug (#43729)

NielsRogge committed 5 days ago

Verified 5405f80e

[kernels] fix kernel versions (#43955)

MekkCyber committed 5 days ago

Verified d0c054ba

Fix LASR test layerdrop issue (#43954)

Rocketknight1 committed 6 days ago

Verified 609e3d58

Fix typos in docstrings, comments, and error messages (#43949)

DimiChatzipavlis committed 6 days ago

Verified c8f112d4

Fix `get_number_of_image_tokens` (#43948)

zucchini-nlp committed 6 days ago

Verified 53a49c48

Fix qwen-vl position ids when generating several times (#43952)

zucchini-nlp committed 6 days ago

Verified 3b8f0948

Timm backbone saves and loads `out_features` (#43886)

zucchini-nlp committed 6 days ago

Verified 0aebdc5c

fix: Better weight decay exclusion in `run_*_no‑trainer.py` examples (#43947)

casinca committed 6 days ago

Verified 2a232cd9

Fix weight decay exclusions in `run_*_no‑trainer.py` examples (#42769)

casinca committed 6 days ago

Verified 2caa05de

Newer Older