huggingface/transformers

Pull Requests Commits

Fix broken link on page (#28451)

keenranger committed 2 years ago

Verified 995a7ce9

Fix docstring checker issues with PIL enums (#28450)

Rocketknight1 committed 2 years ago

Verified 14345135

jiqing-feng committed 2 years ago

Verified 19e83d17

Byebye torch 1.10 (#28207)

ydshieh committed 2 years ago

Verified 59cd9de3

Fix load balancing loss func for mixtral (#28256)

liangxuZhang committed 2 years ago

Verified e768616a

Correctly resolve trust_remote_code=None for AutoTokenizer (#28419)

Rocketknight1 committed 2 years ago

Verified 5d4d62d0

[Phi] Extend implementation to use GQA/MQA. (#28163)

gugarosa committed 2 years ago

Verified 55090585

Optionally preprocess segmentation maps for MobileViT (#28420)

harisankar95 committed 2 years ago

Verified d5606378

Set `cache_dir` for `evaluate.load()` in example scripts (#28422)

aphedges committed 2 years ago

Verified 95091e15

Fix docker file (#28452)

ydshieh committed 2 years ago

Verified 5fd5ef76

Use python 3.10 for docbuild (#28399)

ydshieh committed 2 years ago

Verified d019acb8

Optimize the speed of the truncate_sequences function. (#28263)

ikkvix committed 2 years ago

Verified 2a85345a

Enable multi-label image classification in pipeline (#28433)

amyeroberts committed 2 years ago

Verified 66964c00

Assitant model may on a different device (#27995)

jiqing-feng committed 2 years ago

Verified 8205b264

[Whisper] Fix slow test (#28407)

patrickvonplaten committed 2 years ago

Verified cbbe3074

[docstring] Fix docstring for ErnieConfig, ErnieMConfig (#27029)

Sparty committed 2 years ago

Verified 6c78bbcb

Fix load correct tokenizer in Mixtral model documentation (#28437)

JuanFKurucz committed 2 years ago

Verified 3724156b

Fix for checkpoint rename race condition (#28364)

tblattner committed 2 years ago

Verified cef2e40e

update docs to add the `phi-2` example (#28392)

susnato committed 2 years ago

Verified fff8ca8e

CI: limit natten version (#28432)

gante committed 2 years ago

Verified ee2482b6

Fix number of models in README.md (#28430)

bayllama committed 2 years ago

Verified ffd37103

Support `DeepSpeed` when using auto find batch size (#28088)

muellerzr committed 2 years ago

Verified 6015d0ad

Skip now failing test in the Trainer tests (#28421)

muellerzr committed 2 years ago

Verified a777f525

[BUG] BarkEosPrioritizerLogitsProcessor eos_token_id use list, tensor size mismatch (#28201)

inkinworld committed 2 years ago

Verified 4df1d696

Bump fonttools from 4.31.1 to 4.43.0 in /examples/research_projects/decision_transformer (#28417)

dependabot[bot] committed 2 years ago

Verified 932ad8af

Use mmap option to load_state_dict (#28331)

Weiming Zhao committed 2 years ago

Verified 701298d2

Fix `_merge_input_ids_with_image_features` for llava model (#28333)

VictorSanh committed 2 years ago

Verified 0f2f0c63

Fix initialization for missing parameters in `from_pretrained` under ZeRO-3 (#28245)

XuehaiPan committed 2 years ago

Verified 976189a6

fix auxiliary loss training in DetrSegmentation (#28354)

SangbumChoi committed 2 years ago

Verified 357971ec

[SDPA] Make sure attn mask creation is always done on CPU (#28400)

patrickvonplaten committed 2 years ago

Verified 8604dd30

Newer Older