NielsRogge/transformers

Pull Requests Commits

fix a hidden bug of `GenerationConfig`, now the `generation_config.json` can be loaded successfully (#28604)

ParadoxZW committed 2 years ago

Verified 39c3c0a7

Remove deprecated eager_serving fn (#28665)

Rocketknight1 committed 2 years ago

Verified ebc8f47b

Support single token decode for `CodeGenTokenizer` (#28628)

cmathw committed 2 years ago

Verified 9a4521dd

add dataloader prefetch factor in training args and trainer (#28498)

qmeeus committed 2 years ago

Verified 5b5e71dc

Fix windows err with checkpoint race conditions (#28637)

muellerzr committed 2 years ago

Verified 582d104b

`tensor_size` - fix copy/paste error msg typo (#28660)

scruel committed 2 years ago

Verified c475eca9

Enable instantiating model with pretrained backbone weights (#28214)

amyeroberts committed 2 years ago

Verified 27c79a0f

Enable safetensors conversion from PyTorch to other frameworks without the torch requirement (#27599)

LysandreJik committed 2 years ago

Verified 008a6a22

integrations: fix DVCLiveCallback model logging (#28653)

Dave Berenbaum committed 2 years ago

Verified 03986609

get default device through `PartialState().default_device` as it has been officially released (#27256)

ji-huazhong committed 2 years ago

Verified 1fc12960

Fix phi model doc checkpoint (#28581)

amyeroberts committed 2 years ago

Verified e547458c

[`SigLIP`] Only import tokenizer if sentencepiece available (#28636)

amyeroberts committed 2 years ago

Verified 590be773

Update image_processing_deformable_detr.py (#28561)

sounakdey committed 2 years ago

Verified a35ea570

[`GPTNeoX`] Fix GPTNeoX + Flash Attention 2 issue (#28645)

younesbelkada committed 2 years ago

Verified e201864b

[`Llava`] Update convert_llava_weights_to_hf.py script (#28617)

isaac-vidas committed 2 years ago

Verified dafd5951

Fix lr_scheduler in no_trainer training scripts (#27872)

bofenghuang committed 2 years ago

Verified deb2b590

Add config tip to custom model docs (#28601)

Rocketknight1 committed 2 years ago

Verified 692c3c6b

Avoid root logger's level being changed (#28638)

ydshieh committed 2 years ago

Verified d336c56d

Add missing key to TFLayoutLM signature (#28640)

Rocketknight1 committed 2 years ago

Verified bf674153

Fix id2label assignment in run_classification.py (#28590)

jheitmann committed 2 years ago

Verified f0acf7b6

[`GPTNeoX`] Fix BC issue with 4.36 (#28602)

ArthurZucker committed 2 years ago

Verified 83f9196c

Fix auxiliary loss related code in transformers (#28406)

SangbumChoi committed 2 years ago

Verified 3f69f415

RWKV: raise informative exception when attempting to manipulate `past_key_values` (#28600)

gante committed 2 years ago

Verified 948ffff4

Fix `_speculative_sampling` implementation (#28508)

ofirzaf committed 2 years ago

Verified 9efec114

Allow add_tokens for ESM (#28535)

Rocketknight1 committed 2 years ago

Verified d1578159

[`Llava`] Fix convert_llava_weights_to_hf.py script (#28570)

isaac-vidas committed 2 years ago

Verified 5b7f4bc6

[SigLIP] Don't pad by default (#28578)

NielsRogge committed 2 years ago

Verified faf03541

Fix wrong xpu device in DistributedType.MULTI_XPU mode (#28386)

faaany committed 2 years ago

Verified 8db64367

[Whisper] Finalize batched SOTA long-form generation (#27658)

patrickvonplaten committed 2 years ago

Verified 690fe73f

feat: Sequential beam search (#26304)

Saibo-creator committed 2 years ago

Verified d4fc1eb4

Newer Older