huggingface/transformers

Pull Requests Commits

Revert deletion of self-push-amd.yml for now

ivarflakstad committed 1 year ago

6054220b

Merge branch 'main' into secure-amd-ci

ivarflakstad committed 1 year ago

Verified 2a938385

Adaptive dynamic number of speculative tokens (#34156)

jmamou committed 1 year ago

Verified e27465c8

Fix flaky Hub CI (`test_trainer.py`) (#35062)

ydshieh committed 1 year ago

Verified b0a51e5c

[`trainer`] fix the GA `model_accepts_loss_kwargs` (#34915)

ArthurZucker committed 1 year ago

Verified a928d9c1

BLIP: this is correct now (#35081)

zucchini-nlp committed 1 year ago

Verified e682c17e

Add I-JEPA (#33125)

jmtzt committed 1 year ago

Verified 50189e36

Deprecate quanto and switch to optimum-quanto (#35001)

MekkCyber committed 1 year ago

Verified 95a855e2

Fix `tie_word_embeddings` handling for GGUF models (#35085)

Isotr0py committed 1 year ago

Verified 482cb28a

Update Mistral conversion script (#34829)

Cyrilvallez committed 1 year ago

Verified 35447054

[`tokenizers`] bump to 0.21 (#34972)

ArthurZucker committed 1 year ago

Verified 93f87d3c

[Whisper] Fix whisper tokenizer (#34537)

eustlb committed 1 year ago

Verified 54aae121

Informative (#35059)

ydshieh committed 1 year ago

Verified beb2c66e

[docs] Increase visibility of torch_dtype="auto" (#35067)

stevhliu committed 1 year ago

Verified 1ed1de2f

[docs] add a comment that offloading requires CUDA GPU (#35055)

faaany committed 1 year ago

Verified baa3b221

Support for easier multimodal use of modular (#35056)

Cyrilvallez committed 1 year ago

Verified 1da1e0d7

[`GPTNeoX`] Flex Attention + Refactor (#34896)

vasqu committed 1 year ago

Verified 46df8599

Add Pytorch Tensor Parallel support for Qwen2, Qwen2Moe, Starcoder2 (#35007)

VladOS95-cyber committed 1 year ago

Verified accb7204

Use hf-workflows for both push and scheduled AMD CI

ivarflakstad committed 1 year ago

32b37186

Fix `pad_token_tensor` is None in warning (#34005)

tshu-w committed 1 year ago

Verified c7a109ec

[docs] use device-agnostic API instead of hard-coded cuda (#35048)

faaany committed 1 year ago

Verified 329f5dbf

[docs] use device-agnostic instead of `cuda` (#35047)

faaany committed 1 year ago

Verified b8cdc262

Translate community.md into Chinese (#35013)

wwwbai committed 1 year ago

Verified 346597b6

[docs] fix example code bug (#35054)

faaany committed 1 year ago

Verified 3deaa817

fix speecht5 failure issue in test_peft_gradient_checkpointing_enable… (#34454)

sywangyi committed 1 year ago

Verified 125de416

Use AMD CI workflow defined in hf-workflows

ivarflakstad committed 1 year ago

d38ed44e

Fix `BertGeneration` (#35043)

ydshieh committed 1 year ago

Verified 7a7f2769

Add token cost + runtime monitoring to Agent and HfEngine children (#34548)

aymeric-roucher committed 1 year ago

Verified 901f5045

Automatic compilation in generate: do not rely on inner function (#34923)

Cyrilvallez committed 1 year ago

Verified ee37bf0d

Translate bertlogy.md into Chinese (#34908)

wwwbai committed 1 year ago

Verified f9c7e602

Older