huggingface/transformers

Pull Requests Commits

fix: Removed a wrong key-word argument in `sigmoid_focal_loss()` function call (#31951)

Sai-Suraj-27 committed 1 year ago

Verified 454bc14d

Whisper: move to tensor cpu before converting to np array at decode time (#31954)

gante committed 1 year ago

Verified a5c642fe

Generate: v4.42 deprecations 🧹🧹 (#31956)

gante committed 1 year ago

Verified df1c248a

Generate: remove deprecated code due to `Cache` and `cache_position` being default (#31898)

gante committed 1 year ago

Verified 739a6316

Fix `GenerationMixin.generate` compatibility with pytorch profiler (#31935)

fxmarty committed 1 year ago

Verified 8480fda6

fix prompt strip to support tensors and np arrays (#27818)

AvivSham committed 1 year ago

Verified 7f79a973

Docker: TF pin on the consistency job (#31928)

gante committed 1 year ago

Verified d1a1bcf5

[Bug Fix] fix qa pipeline tensor to numpy (#31585)

jiqing-feng committed 1 year ago

Verified aec1ca3a

Adding hiera (#30356)

Namangarg110 committed 1 year ago

Verified c1e139c2

Allow `Trainer.get_optimizer_cls_and_kwargs` to be overridden (#31875)

apoorvkh committed 1 year ago

Verified 574e68d5

🚨 fix(SigLip): remove spurious exclusion of first vision output token (#30952)

transmissions11 committed 1 year ago

Verified 52585019

Generate: fix `SlidingWindowCache.reset()` (#31917)

gante committed 1 year ago

Verified 6a05f68f

Refactor flash attention implementation in transformers (#31446)

ArthurZucker committed 1 year ago

Verified e3143952

Fix fx tests with inputs_embeds (#31862)

fxmarty committed 1 year ago

Verified ad4ef3a2

Add warning message for beta and gamma parameters (#31654)

OmarManzoor committed 1 year ago

Verified 1499a550

add gather_use_object arguments II (#31799)

SangbumChoi committed 1 year ago

Verified 23d6d0cc

fix: Fixed the `1st argument` name in classmethods (#31907)

Sai-Suraj-27 committed 1 year ago

Verified 2e48b3e8

Fix missing methods for Fuyu (#31880)

Isotr0py committed 1 year ago

Verified 48c20700

[`Gemma2`] Support FA2 softcapping (#31887)

ArthurZucker committed 1 year ago

Verified f4ec7a28

[`ConvertSlow`] make sure the order is preserved for addedtokens (#31902)

ArthurZucker committed 1 year ago

Verified f67e0f7f

Processor accepts any kwargs (#31889)

zucchini-nlp committed 1 year ago

Verified 14d3b3f0

Fixes to alternating SWA layers in Gemma2 (#31775)

turboderp committed 1 year ago

Verified a695c186

InstructBlipVideo: Update docstring (#31886)

zucchini-nlp committed 1 year ago

Verified d625294d

Add a condition for nested_detach (#31855)

haikuoxin committed 1 year ago

Verified c54af4c7

Modify `warnings` in a `with` block to avoid flaky tests (#31893)

ydshieh committed 1 year ago

Verified 080e14b2

[RT-DETR] Add resources (#31815)

NielsRogge committed 1 year ago

Verified ec03d97b

Push sharded checkpoint to hub when `push_to_hub=True` in `TrainingArguments` (#31808)

SunMarc committed 1 year ago

Verified 8df28bb3

fix: Removed `duplicate` field definitions in some classes (#31888)

Sai-Suraj-27 committed 1 year ago

Verified da79b180

Fix failed tests in #31851 (#31879)

ydshieh committed 1 year ago

Verified 9d98706b

Fix file type checks in data splits for contrastive training example script (#31720)

npyoung committed 1 year ago

Verified a0a3e2f4

Newer Older