huggingface/transformers

Pull Requests Commits

fix pixtral tests

ArthurZucker committed 1 year ago

3c0cb388

Fix-red-ci (#34230)

ArthurZucker committed 1 year ago

Verified b54109c7

Enable users to use their own loss functions + deal with prefetching for grad accum (#34198)

muellerzr committed 1 year ago

Verified 6ba31a8a

Support Llama 3.2 conversion (text models) (#33778)

pcuenca committed 1 year ago

Verified 7a06d07e

Fix Gradient Accumulation issue (#34191)

ArthurZucker committed 1 year ago

Verified c1c7e896

Generate: visit non-llm `prepare_inputs_for_generation` (#34199)

gante committed 1 year ago

Verified f51ac9e0

Fix bus error when using GPT2 on M1 macs (#34031)

chanind committed 1 year ago

Verified 1d2c29f0

Llama3 and Llama2 are ExecuTorch compatible (#34101)

guangy10 committed 1 year ago

Verified 9470c000

removes decord (#33987)

vrnvu committed 1 year ago

Verified 7f508850

Fix for tokenizer.apply_chat_template with continue_final_message=True (#34214)

schoennenbeck committed 1 year ago

Verified f2846ad2

fix(Wav2Vec2ForCTC): torch export (#34023)

chrsmcgrr committed 1 year ago

Verified b57c7bce

Ping team members for new failed tests in daily CI (#34171)

ydshieh committed 1 year ago

Verified fce1fcfe

Fix warning message for fp32_cpu_offloading in bitsandbytes configs (#34079)

amosyou committed 1 year ago

Verified aa3e35ac

Update `trainer._get_eval_sampler()` to support `group_by_length` arg (#33514)

larin92 committed 1 year ago

Verified 6d2b2033

Revert "Fix FSDP resume Initialization issue" (#34193)

SunMarc committed 1 year ago

Verified 3f06f95e

Avoid using torch's Tensor or PIL's Image in chat template utils if not available (#34165)

RezaRahemtola committed 1 year ago

Verified 3a10c619

Fix wrong name for llava onevision and qwen2_vl in tokenization auto (#34177)

yonigozlan committed 1 year ago

Verified bd5dc10f

Revert `accelerate` error caused by `46d09af` (#34197)

steveepreston committed 1 year ago

Verified cc7d8b87

[fix] fix token healing tests and usage errors (#33931)

alpertunga-bile committed 1 year ago

Verified 98bad9c6

Moshi integration (#33624)

ylacombe committed 1 year ago

Verified 9ba021ea

IDEFICS: support inputs embeds (#34043)

zucchini-nlp committed 1 year ago

Verified d087165d

🌐 [i18n-KO] Translated `blip-2.md` to Korean (#33516)

cjfghk5697 committed 1 year ago

Verified 9d6998c7

🌐 [i18n-KO] Translated `trainer_utils.md` to Korean (#33817)

yijun-lee committed 1 year ago

Verified 554ed5d1

🌐 [i18n-KO] Translated `gemma2.md` to Korean (#33937)

yijun-lee committed 1 year ago

Verified 8c33cf4e

🌐 [i18n-KO] Translated `vivit.md` to Korean (#33935)

mreraser committed 1 year ago

Verified 67acb0b1

[feat] LlavaNext add feature size check to avoid CUDA Runtime Error (#33608)

laurentd-lunit committed 1 year ago

Verified 0f49deac

Fix optuna ddp hp search (#34073)

SunMarc committed 1 year ago

Verified d00f1ca8

Add support for inheritance from class with different suffix in modular (#34077)

yonigozlan committed 1 year ago

Verified 65442718

Generate: move `logits` to same device as `input_ids` (#34076)

gante committed 1 year ago

Verified d314ce70

Fix default behaviour in TextClassificationPipeline for regression problem type (#34066)

subhalingamd committed 1 year ago

Verified 5ee9e786

Older