huggingface/transformers

Pull Requests Commits

ArthurZucker committed 1 year ago

052e652d

FSDP grad accum fix (#34645)

winglian committed 1 year ago

e01a61ae

MPS: `isin_mps_friendly` can support 0D tensors (#34538)

gante committed 1 year ago

ccbd57a8

ArthurZucker committed 1 year ago

e66224b5

Update trainer for easier handling of accumulate, compile fixes, and proper reporting (#34511)

muellerzr committed 1 year ago

8c62a92b

enable average tokens across devices (#34373)

techkang committed 1 year ago

5b36cdab

fix pixtral processor (#34486)

molbap committed 1 year ago

f784d95c

VLMs: fix number of image tokens (#34332)

zucchini-nlp committed 1 year ago

7da0eefc

ArthurZucker committed 1 year ago

bc598c00

Fix regression loading dtype (#34409)

SunMarc committed 1 year ago

94ed13c1

LLaVA: latency issues (#34460)

zucchini-nlp committed 1 year ago

72c716de

Fix pix2struct (#34374)

IlyasMoutawwakil committed 1 year ago

97bb9299

Fix onnx non-expotable inplace aten op (#34376)

IlyasMoutawwakil committed 1 year ago

565f0e97

Fix `torch.fx` issue related to the new `loss_kwargs` keyword argument (#34380)

michaelbenayoun committed 1 year ago

dcfe3c7e

ArthurZucker committed 1 year ago

c2820c94

ArthurZucker committed 1 year ago

b2981611

Add SynthID (watermerking by Google DeepMind) (#34350)

gante committed 1 year ago

Verified b0f0c618

Fix red CI: benchmark script (#34351)

ArthurZucker committed 1 year ago

Verified e50bf61d

skip `test_pipeline_depth_estimation` temporarily (#34316)

ydshieh committed 1 year ago

Verified c42b3223

Enable Gradient Accumulation fix across all models + trainer fully in forward() (#34283)

muellerzr committed 1 year ago

Verified d9f73362

Support boolean tool args (#34208)

aymeric-roucher committed 1 year ago

Verified 1fb575fc

Added Deberta model type support (#34308)

FilipposVentirozos committed 1 year ago

Verified 343c8cb8

[docs] Fix Korean toctree (#34324)

stevhliu committed 1 year ago

Verified 5ba85de7

Example doc for token classification of Llama and Dependent/Copied Models (#34139)

h3110Fr13nd committed 1 year ago

Verified 049682a5

🌐 [i18n-KO] Translated `model_doc/bartpho.md` to Korean (#33981)

Jwaminju committed 1 year ago

Verified 644d5287

🌐 [i18n-KO] Translated `bert japanese.md` to Korean (#33890)

ahnjj committed 1 year ago

Verified b03dc0a8

🌐 [i18n-KO] Translated `executorch.md` to Korean (#33888)

ahnjj committed 1 year ago

Verified 4b14aa1b

[docs] fix typo (#34235)

faaany committed 1 year ago

Verified 688eeac8

fix error in _get_eval_sampler when group_by_length enabled (#34237)

akakakakakaa committed 1 year ago

Verified a65a6ce7

Fix continue_final_message for image-text-to-text chat templates (#34236)

yonigozlan committed 1 year ago

Verified e7c3fa7f

Older