huggingface/transformers

Pull Requests Commits

ArthurZucker committed 188 days ago

45401fbe

simplify and refactor the modular

ArthurZucker committed 188 days ago

b7b57744

ArthurZucker committed 188 days ago

5e31c070

ArthurZucker committed 188 days ago

7bf3f975

revert wrong merge

ArthurZucker committed 188 days ago

ec8e4993

ArthurZucker committed 188 days ago

940b6026

ArthurZucker committed 188 days ago

638ba2f3

[`CI`] Fix partially red CI (#39448)

vasqu committed 188 days ago

Verified 26be7f71

Fixes #39204: add fallback if get_base_model missing (#39226)

sebastianvlad1 committed 188 days ago

Verified 0a887519

make the loss context manager easier to extend (#39321)

winglian committed 188 days ago

Verified ba506f87

Remove something that should have never been there (#38254)

ArthurZucker committed 188 days ago

Verified 9f1ac6f1

Fix processor tests (#39450)

zucchini-nlp committed 188 days ago

Verified a7ca5b5d

[Bugfix] [Quantization] Remove unused init arg (#39324)

kylesayrs committed 188 days ago

Verified 71818f57

Better typing for model.config (#39132)

qubvel committed 188 days ago

Verified cc24b037

Fix typo in generation configuration for Janus model weight conversion (#39432)

thisisiron committed 188 days ago

Verified 4b258454

Responses API in `transformers serve` (#39155)

LysandreJik committed 188 days ago

Verified de5ca373

[cache] make all classes cache compatible finally (#38635)

zucchini-nlp committed 188 days ago

Verified c8524aeb

docs: add missing numpy import to minimal example (#39444)

IliasAarab committed 188 days ago

Verified 6cb43def

Remove runtime conditions for type checking (#37340)

cyyever committed 188 days ago

Verified 61163099

Add StableAdamW Optimizer (#39446)

SunMarc committed 188 days ago

Verified bfc9ddf5

add test scanner (#39419)

molbap committed 188 days ago

Verified b9ee5282

Fix missing definition of diff_file_url in notification service (#39445)

ahadnagy committed 188 days ago

Verified 79941c61

Add cosine_with_min_lr_schedule_with_warmup_lr_rate scheduler in Trainer (#31870)

richardodliu committed 188 days ago

Verified e048d48b

Change log level from warning to info for scheduled request logging in `ContinuousBatchProcessor` (#39372)

qgallouedec committed 188 days ago

Verified 0cf08e90

Defaults to adamw_torch_fused for Pytorch>=2.8 (#37358)

cyyever committed 188 days ago

Verified ae4e306a

Fix L270 - hasattr("moe_args") returning False error (#38715)

wjdghks950 committed 188 days ago

Verified 4524a68c

[chat template] add a testcase for kwargs (#39415)

zucchini-nlp committed 188 days ago

Verified d33a1c38

Fixed a bug calculating cross entropy loss in `JetMoeForCausalLM` (#37830)

Phoenix-Shen committed 188 days ago

Verified 99c97633

Remove double soft-max in load-balancing loss. Fixes #39055 . (#39056)

rudolfwilliam committed 188 days ago

Verified 667ad023

[Core] [Offloading] Fix saving offloaded submodules (#39280)

kylesayrs committed 188 days ago

Verified 31d81943

Older