huggingface/transformers

Pull Requests Commits

resolving zero3 init when using accelerate config with Trainer (#25227)

pacman100 committed 2 years ago

Verified 904e7e0f

Add `token` arugment in example scripts (#25172)

ydshieh committed 2 years ago

Verified 149cb0cc

add pathname and line number to logging formatter in debug mode (#25203)

ranchlai committed 2 years ago

Verified c6a8768d

fix get_keys_to_not_convert() to return correct modules for full precision inference (#25105)

ranchlai committed 2 years ago

Verified 2230d149

Fix set of model parallel in the Trainer when no GPUs are available (#25239)

sgugger committed 2 years ago

Verified f6f567d0

Move rescale dtype recasting to match torchvision ToTensor (#25229)

amyeroberts committed 2 years ago

Verified d27e4c18

[`Detr`] Fix detr BatchNorm replacement issue (#25230)

younesbelkada committed 2 years ago

Verified 3170af71

[`MPT`] Add `require_bitsandbytes` on MPT integration tests (#25201)

younesbelkada committed 2 years ago

Verified 05ebb026

[`Docs`/`quantization`] Clearer explanation on how things works under the hood. + remove outdated info (#25216)

younesbelkada committed 2 years ago

Verified 972fdcc7

[`Pix2Struct`] Fix pix2struct cross attention (#25200)

younesbelkada committed 2 years ago

Verified 77c3973e

make build_mpt_alibi_tensor a method of MptModel so that deepspeed co… (#25193)

sywangyi committed 2 years ago

Verified 4033ea71

Fix docker image build failure (#25214)

ydshieh committed 2 years ago

Verified 0fd8d2aa

Update tiny model info. and pipeline testing (#25213)

ydshieh committed 2 years ago

Verified 1b4f6199

[`pipeline`] revisit device check for pipeline (#25207)

younesbelkada committed 2 years ago

Verified e0c50b27

[quantization.md] fix (#25190)

stas00 committed 2 years ago

Verified 52206066

Fix `all_model_classes` in `FlaxBloomGenerationTest` (#25211)

ydshieh committed 2 years ago

Verified 9ca3aa01

[`PreTrainedModel`] Wrap `cuda` and `to` method correctly (#25206)

younesbelkada committed 2 years ago

Verified 59dcea3f

Better error message in `_prepare_output_docstrings` (#25202)

ydshieh committed 2 years ago

Verified 67b85f24

Musicgen: CFG is manually added (#25173)

gante committed 2 years ago

Verified 4a564490

🚨🚨🚨 Fix rescale ViVit Efficientnet (#25174)

amyeroberts committed 2 years ago

Verified 05cda5df

[MusicGen] Fix integration tests (#25169)

sanchit-gandhi committed 2 years ago

Verified 03f98f96

Fix beam search to sample at least 1 non eos token (#25103) (#25115)

yonigottesman committed 2 years ago

Verified c90e14fb

🌐 [i18n-KO] Translated `transformers_agents.md` to Korean (#24881)

sim-so committed 2 years ago

Verified 31f137c0

[`InstructBlip`] Fix instructblip slow test (#25171)

younesbelkada committed 2 years ago

Verified dd9d45b6

[`Mpt`] Fix mpt slow test (#25170)

younesbelkada committed 2 years ago

Verified add0895d

Update `use_auth_token` -> `token` in example scripts (#25167)

ydshieh committed 2 years ago

Verified d53b8ad7

added compiled model support for inference (#25124)

Alexander Markov committed 2 years ago

Verified 3cbc560d

make run_generation more generic for other devices (#25133)

ji-huazhong committed 2 years ago

Verified afa96fff

Represent query_length in a different way to solve jit issue (#25164)

jiqing-feng committed 2 years ago

Verified d23d2c27

override .cuda() to check if model is already quantized (#25166)

ranchlai committed 2 years ago

Verified 2a787201

Newer Older