huggingface/transformers

Pull Requests Commits

[Makefile] Exclude research projects from format (#28551)

patrickvonplaten committed 2 years ago

Verified 09eb11a1

Config: warning when saving generation kwargs in the model config (#28514)

gante committed 2 years ago

Verified f4f57f9d

Add is_model_supported for fx (#28521)

inisis committed 2 years ago

Verified 7142bdfa

Clearer error for SDPA when explicitely requested (#28006)

fxmarty committed 2 years ago

Verified 02f8738e

[`SpeechT5Tokenization`] Add copied from and fix the `convert_tokens_to_string` to match the fast decoding scheme (#28522)

ArthurZucker committed 2 years ago

Verified fe23256b

[`TokenizationRoformerFast`] Fix the save and loading (#28527)

ArthurZucker committed 2 years ago

Verified 96d08831

[ `TokenizationUtils`] Fix `add_special_tokens` when the token is already there (#28520)

ArthurZucker committed 2 years ago

Verified 716df5fb

Fix/speecht5 bug (#28481)

NimaYaqmuri committed 2 years ago

Verified 07ae53e6

Fix mismatching loading in from_pretrained with/without accelerate (#28414)

fxmarty committed 2 years ago

Verified 66db33dd

Improving Training Performance and Scalability Documentation (#28497)

HamzaFB committed 2 years ago

Verified 002566f3

Remove `task` arg in `load_dataset` in image-classification example (#28408)

regisss committed 2 years ago

Verified 0cdcd7a2

SiLU activation wrapper for safe importing (#28509)

amyeroberts committed 2 years ago

Verified edb17023

improve dev setup comments and hints (#28495)

4imothy committed 2 years ago

Verified ff86bc36

fix: sampling in flax keeps EOS (#28378)

borisdayma committed 2 years ago

Verified 735968b6

Generate: consolidate output classes (#28494)

gante committed 2 years ago

Verified 7e0ddf89

Add a use_safetensors arg to TFPreTrainedModel.from_pretrained() (#28511)

Rocketknight1 committed 2 years ago

Verified 72db39c0

Fixed minor typos (#28489)

rishit5 committed 2 years ago

Verified 78d767e3

[GPTQ] Fix test (#28018)

SunMarc committed 2 years ago

Verified 7c8dd88d

Tokenizer kwargs in textgeneration pipe (#28362)

farzadsbakhtiar committed 2 years ago

Verified 366c0327

Add the XPU device check for pipeline mode (#28326)

yuanwu2017 committed 2 years ago

Verified a573ac74

[`core`/ FEAT] Add the possibility to push custom tags using `PreTrainedModel` itself (#28405)

younesbelkada committed 2 years ago

Verified 1b9a2e4c

Don't set `finetuned_from` if it is a local path (#28482)

ydshieh committed 2 years ago

Verified 64bdbd88

[`chore`] Update warning text, a word was missing (#28017)

tomaarsen committed 2 years ago

Verified 881e966a

Fix paths to AI Sweden Models reference and model loading (#28423)

JuanFKurucz committed 2 years ago

Verified 121641ca

Generate: fix candidate device placement (#28493)

gante committed 2 years ago

Verified bc72b4e2

Adding Prompt lookup decoding (#27775)

apoorvumang committed 2 years ago

Verified e304f976

Change progress logging to once across all nodes (#28373)

siddartha-RE committed 2 years ago

Verified 29a2b142

Fix docstrings and update docstring checker error message (#28460)

Rocketknight1 committed 2 years ago

Verified 2382706a

TF: purge `TFTrainer` (#28483)

gante committed 2 years ago

Verified 4fb3d3a0

Generate: refuse to save bad generation config files (#28477)

gante committed 2 years ago

Verified afc45b13

Newer Older