huggingface/transformers

Pull Requests Commits

use channel name directly instead of ID

ydshieh committed 1 year ago

b5b50db3

use channel name directly instead of ID

ydshieh committed 1 year ago

c53d4002

use channel name directly instead of ID

ydshieh committed 1 year ago

bae019e9

ydshieh committed 1 year ago

a20fd8b4

ydshieh committed 1 year ago

9ff0ebab

Fix 29807 sinusoidal positional encodings in Flaubert, Informer and XLM (#29904)

hovnatan committed 1 year ago

Verified 416711c3

[`generate`] fix breaking change for patch (#29976)

ArthurZucker committed 1 year ago

Verified 83b26dd7

[docs] Big model loading (#29920)

stevhliu committed 1 year ago

Verified 096f3046

Generate: move misplaced test (#29902)

gante committed 1 year ago

Verified c9f6e5e3

[tests] fix the wrong output in `ImageToTextPipelineTests.test_conditional_generation_llava` (#29975)

faaany committed 1 year ago

Verified e4f5b57a

Fix copies main ci (#29979)

ArthurZucker committed 1 year ago

Verified fa2c49b0

Fix FA2 tests (#29909)

ylacombe committed 1 year ago

Verified 569f6c7d

Rework tests to compare trainer checkpoint args (#29883)

muellerzr committed 1 year ago

Verified 3b8e2932

[`BC`] Fix BC for AWQ quant (#29965)

TechxGenus committed 1 year ago

Verified 6e584070

Update model card and link of blog post. (#29928)

bozheng-hit committed 1 year ago

Verified 46d63681

Reset alarm signal when the function is ended (#29706)

coldnight committed 1 year ago

Verified f6701bc6

fix: get mlflow version from mlflow-skinny (#29918)

Alexander Jipa committed 1 year ago

Verified e644b600

Add warning message for `run_qa.py` (#29867)

jla524 committed 1 year ago

Verified 156d30da

Fix rope theta for OpenLlama (#29893)

jla524 committed 1 year ago

Verified 6fd93fe9

Super tiny fix 12 typos about "with with" (#29926)

fzyzcjy committed 1 year ago

Verified 5ad7f170

Mark `test_eager_matches_sdpa_generate` flaky for some models (#29479)

ydshieh committed 1 year ago

Verified 43d17c18

Update installs in image classification doc (#29947)

MariaHei committed 1 year ago

Verified ba56ed08

[`LlamaSlowConverter`] Slow to Fast better support (#29797)

ArthurZucker committed 1 year ago

Verified 536ea2ac

Fix doc issue #29758 in DebertaV2Config class (#29842)

vinayakkgarg committed 1 year ago

Verified e2036468

[`BC`] Fix BC for other libraries (#29934)

ArthurZucker committed 1 year ago

Verified 2bbbf1be

Allow GradientAccumulationPlugin to be configured from AcceleratorConfig (#29589)

fabianlim committed 1 year ago

Verified 4df5b9b4

[ `TokenizationLlama`] fix the way we convert tokens to strings to keep leading spaces 🚨 breaking fix (#29453)

ArthurZucker committed 1 year ago

Verified a2a7f716

[`Mamba`] from pretrained issue with `self.embeddings` (#29851)

ArthurZucker committed 1 year ago

Verified e677479c

RoPE models: add numerical sanity-check test for RoPE scaling (#29808)

gante committed 1 year ago

Verified 441de62f

add functions to inspect model and optimizer status to trainer.py (#29838)

CKeibel committed 1 year ago

Verified aac7099c

Older