Code

Update latest commits #34

Code

Merged

datquocnguyen merged 92 commits intodatquocnguyen:fast_tokenizers_BARTpho_PhoBERT_BERTweetdatquocnguyen/transformers:fast_tokenizers_BARTpho_PhoBERT_BERTweetfrom huggingface:mainhuggingface/transformers:mainCopy head branch name to clipboardSep 6, 2022

+11,798-934

Commits

Commits on Aug 23, 2022

• Unpin detectron2 (#18727)

  Show description for 84beb8a
  ydshieh

  authored
  84beb8a

  Copy full SHA for 84beb8a

• Removing warning of model type for microsoft/tapex-base-finetuned-wtq (#18711)

  Show description for 891704b
  Narsil

  authored
  891704b

  Copy full SHA for 891704b

• improve add_tokens docstring (#18687)

  Show description for 4386980
  SaulLu

  authored
  4386980

  Copy full SHA for 4386980

• CLI: Don't check the model head when there is no model head (#18733)

  gante

  authored
  6faf283

  Copy full SHA for 6faf283

Commits on Aug 24, 2022

• Update perf_infer_gpu_many.mdx (#18744)

  mishig25

  authored
  c12dbdc

  Copy full SHA for c12dbdc

• Add minor doc-string change to include hp_name param in hyperparameter_search (#18700)

  Show description for a442884
  constantin-huetterer

  authored
  a442884

  Copy full SHA for a442884

• fix pipeline_tutorial.mdx doctest (#18717)

  Show description for cecf9f9
  ydshieh

  authored
  cecf9f9

  Copy full SHA for cecf9f9

• Add TF implementation of XGLMModel (#16543)

  Show description for c72d7d9
  

  6 peopleauthored
  c72d7d9

  Copy full SHA for c72d7d9

• Bump nbconvert in /examples/research_projects/visual_bert (#18741)

  Show description for 5b24949
  dependabot[bot]

  authored
  5b24949

  Copy full SHA for 5b24949

• Bump nbconvert from 6.3.0 to 6.5.1 in /examples/research_projects/lxmert (#18742)

  Show description for e49c71f
  dependabot[bot]

  authored
  e49c71f

  Copy full SHA for e49c71f

• fixed docstring typos (#18739)

  Show description for dcff504
  

  JadeKim042386

  and

  김주영

  authored
  dcff504

  Copy full SHA for dcff504

• add warning to let the user know that the __call__ method is faster than encode + pad for a fast tokenizer (#18693)

  Show description for 6667b0d
  SaulLu

  authored
  6667b0d

  Copy full SHA for 6667b0d

• examples/run_summarization_no_trainer: fixed incorrect param to hasattr (#18720)

  Show description for c55d6e4
  rahular

  authored
  c55d6e4

  Copy full SHA for c55d6e4

Commits on Aug 25, 2022

• Add ONNX support for Longformer (#17176)

  Show description for 3223d49
  deutschmn

  authored
  3223d49

  Copy full SHA for 3223d49

• Determine framework automatically before ONNX export (#18615)

  Show description for fbf382c
  

  rachthree

  and

  lewtun

  authored
  fbf382c

  Copy full SHA for fbf382c

• streamlining 'checkpointing_steps' parsing (#18755)

  rahular

  authored
  e944244

  Copy full SHA for e944244

• CLI: Improved error control and updated hub requirement (#18752)

  gante

  authored
  06a6a4b

  Copy full SHA for 06a6a4b

Commits on Aug 26, 2022

• [VisionEncoderDecoder] Add gradient checkpointing (#18697)

  Show description for 8869bf4
  patrickvonplaten

  authored
  8869bf4

  Copy full SHA for 8869bf4

• [Wav2vec2 + LM Test] Improve wav2vec2 with lm tests and make torch version dependent for now (#18749)

  Show description for 62ceb4d
  patrickvonplaten

  authored
  62ceb4d

  Copy full SHA for 62ceb4d

• Fix incomplete outputs of FlaxBert (#18772)

  Show description for 21f6f58
  duongna21

  authored
  21f6f58

  Copy full SHA for 21f6f58

Commits on Aug 29, 2022

• Fix broken link DeepSpeed documentation link (#18783)

  Show description for f2fbe44
  

  philschmid

  and

  stas00

  authored
  f2fbe44

  Copy full SHA for f2fbe44

• fix missing block when there is no failure (#18775)

  Show description for 5f06a09
  ydshieh

  authored
  5f06a09

  Copy full SHA for 5f06a09

• fix a possible typo in auto feature extraction (#18779)

  fcakyon

  authored
  b10a3b3

  Copy full SHA for b10a3b3

• Fix memory leak issue in torch_fx tests (#18547)

  Show description for 8b67f20
  

  ydshieh

  and

  LysandreJik

  authored
  8b67f20

  Copy full SHA for 8b67f20

• Fix mock in test_cached_files_are_used_when_internet_is_down (#18804)

  Wauplin

  authored
  169b8cd

  Copy full SHA for 169b8cd

• Add SegFormer and ViLT links (#18808)

  Show description for f1fd460
  

  NielsRogge

  and

  Niels Rogge

  authored
  f1fd460

  Copy full SHA for f1fd460

• send model to the correct device (#18800)

  Show description for da5bb29
  ydshieh

  authored
  da5bb29

  Copy full SHA for da5bb29

• Revert to and safely handle flag in owlvit config (#18750)

  amyeroberts

  authored
  8c4a114

  Copy full SHA for 8c4a114

Commits on Aug 30, 2022

• Add docstring for BartForCausalLM (#18795)

  Show description for da02b40
  ekagra-ranjan

  authored
  da02b40

  Copy full SHA for da02b40

• up (#18805)

  qqaatw

  authored
  5c70217

  Copy full SHA for 5c70217

• [Swin, Swinv2] Fix attn_mask dtype (#18803)

  Show description for 7320d95
  

  NielsRogge

  and

  Niels Rogge

  authored
  7320d95

  Copy full SHA for 7320d95

• [LayoutLMv3] Add TensorFlow implementation (#18678)

  Show description for de8548e
  

  5 peopleauthored
  de8548e

  Copy full SHA for de8548e

• Run tests if skip condition not met (#18764)

  Show description for ef91a2d
  amyeroberts

  authored
  ef91a2d

  Copy full SHA for ef91a2d

• Remove ViltForQuestionAnswering from check_repo (#18762)

  Show description for b83796d
  

  NielsRogge

  and

  Niels Rogge

  authored
  b83796d

  Copy full SHA for b83796d

• Adds OWLViT to models exportable with ONNX (#18588)

  Show description for 46d0e26
  unography

  authored
  46d0e26

  Copy full SHA for 46d0e26

• Adds GroupViT to models exportable with ONNX (#18628)

  Show description for 220da3b
  unography

  authored
  220da3b

  Copy full SHA for 220da3b

• LayoutXLMProcessor: ensure 1-to-1 mapping between samples and images, and add test for it (#18774)

  anthony2261

  authored
  a98f6a1

  Copy full SHA for a98f6a1

• Added Docstrings for Deberta and DebertaV2 [PyTorch] (#18610)

  Show description for 5727dfc
  Tegzes

  authored
  5727dfc

  Copy full SHA for 5727dfc

• Improving the documentation for "word", within the pipeline. (#18763)

  Show description for 73c6273
  Narsil

  authored
  73c6273

  Copy full SHA for 73c6273

• Disable nightly CI temporarily (#18820)

  Show description for e88e9ff
  ydshieh

  authored
  e88e9ff

  Copy full SHA for e88e9ff

Commits on Aug 31, 2022

• Add LayoutLMForQuestionAnswering model (#18407)

  Show description for 5c4c869
  ankrgyl

  authored
  5c4c869

  Copy full SHA for 5c4c869

• Pin max tf version (#18818)

  gante

  authored
  fea4636

  Copy full SHA for fea4636

• Fix cost condition in DetrHungarianMatcher and YolosHungarianMatcher to allow zero-cost (#18647)

  Show description for c3be98e
  kongzii

  authored
  c3be98e

  Copy full SHA for c3be98e

• oob performance improvement for cpu DDP (#18595)

  Show description for cdde85a
  sywangyi

  authored
  cdde85a

  Copy full SHA for cdde85a

• Warn on TPUs when the custom optimizer and model device are not the same (#18668)

  Show description for e491021
  muellerzr

  authored
  e491021

  Copy full SHA for e491021

• Update location identification (#18834)

  LysandreJik

  authored
  ee40702

  Copy full SHA for ee40702

• fix bug: register_for_auto_class should be defined on TFPreTrainedModel instead of TFSequenceSummary (#18607)

  azonti

  authored
  811c4c9

  Copy full SHA for 811c4c9

• [DETR] Add num_channels attribute (#18714)

  Show description for 3b6943e
  

  NielsRogge

  and

  Niels Rogge

  authored
  3b6943e

  Copy full SHA for 3b6943e

• Pin ffspec (#18837)

  Show description for 74690b6
  sgugger

  authored
  74690b6

  Copy full SHA for 74690b6

• Improve GPT2 doc (#18787)

  Show description for f210e2a
  ekagra-ranjan

  authored
  f210e2a

  Copy full SHA for f210e2a

• Add an option to HfArgumentParser.parse_{dict,json_file} to raise an Exception when there extra keys (#18692)

  Show description for 86387fe
  FelixSchneiderZoom

  authored
  86387fe

  Copy full SHA for 86387fe

• Improve Text Generation doc (#18788)

  Show description for 89514f0
  

  ekagra-ranjan

  and

  sgugger

  authored
  89514f0

  Copy full SHA for 89514f0

• Add SegFormer ONNX support (#18006)

  Show description for 7e7f743
  

  NielsRogge

  and

  Niels Rogge

  authored
  7e7f743

  Copy full SHA for 7e7f743

• Add security warning about the from_pretrained() method (#18801)

  Show description for 80367cd
  

  lewtun

  and

  julien-c

  authored
  80367cd

  Copy full SHA for 80367cd

Commits on Sep 1, 2022

• Owlvit memory leak fix (#18734)

  Show description for 5d81a56
  alaradirik

  authored
  5d81a56

  Copy full SHA for 5d81a56

• Create pipeline_tutorial.mdx german docs (#18625)

  Show description for 359f7b4
  flozi00

  authored
  359f7b4

  Copy full SHA for 359f7b4

• Unpin fsspec (#18846)

  albertvillanova

  authored
  a26c752

  Copy full SHA for a26c752

• Delete state_dict to release memory as early as possible (#18832)

  Show description for 563a8d5
  

  ydshieh

  and

  sgugger

  authored
  563a8d5

  Copy full SHA for 563a8d5

• Generate: smaller TF serving test (#18840)

  gante

  authored
  6e01663

  Copy full SHA for 6e01663

• add a script to get time info. from GA workflow jobs (#18822)

  Show description for e7da38f
  ydshieh

  authored
  e7da38f

  Copy full SHA for e7da38f

• Pin rouge_score (#18247)

  Show description for fafbb57
  albertvillanova

  authored
  fafbb57

  Copy full SHA for fafbb57

• Minor typo in prose of model outputs documentation. (#18848)

  pcuenca

  authored
  f719c03

  Copy full SHA for f719c03

• reflect max_new_tokens in Seq2SeqTrainer (#18786)

  Show description for ab663b2
  kumapo

  authored
  ab663b2

  Copy full SHA for ab663b2

• Adds timeout argument to training_args to avoid socket timeouts in DDP (#18562)

  Show description for fe58929
  gugarosa

  authored
  fe58929

  Copy full SHA for fe58929

• TensorFlow MobileViT (#18555)

  Show description for 954e18a
  

  3 peopleauthored
  954e18a

  Copy full SHA for 954e18a

• Cache results of is_torch_tpu_available() (#18777)

  Show description for 1c381f3
  comaniac

  authored
  1c381f3

  Copy full SHA for 1c381f3

• Tie weights after preparing the model in run_clm (#18855)

  sgugger

  authored
  c61f116

  Copy full SHA for c61f116

• Add Image To Text Generation pipeline (#18821)

  Show description for ddb69e5
  OlivierDehaene

  authored
  ddb69e5

  Copy full SHA for ddb69e5

• Pin revision for LayoutLMForQuestionAnswering and TFLayoutLMForQuestionAnswering tests (#18854)

  Show description for 23fab60
  ankrgyl

  authored
  23fab60

  Copy full SHA for 23fab60

• Split docs on modality (#18205)

  Show description for 142e12a
  stevhliu

  authored
  142e12a

  Copy full SHA for 142e12a

Commits on Sep 2, 2022

• if learning rate is a tensor, get item (float) (#18861)

  kmckiern

  authored
  9b3eb81

  Copy full SHA for 9b3eb81

• Fix naming issue with ImageToText pipeline (#18864)

  Show description for 129d732
  OlivierDehaene

  authored
  129d732

  Copy full SHA for 129d732

• [LayoutLM] Add clarification to docs (#18716)

  Show description for c60dd98
  

  NielsRogge

  and

  Niels Rogge

  authored
  c60dd98

  Copy full SHA for c60dd98

• Add OWL-ViT to the appropriate section (#18867)

  Show description for 17981fa
  

  NielsRogge

  and

  Niels Rogge

  authored
  17981fa

  Copy full SHA for 17981fa

• Clean up utils.hub using the latest from hf_hub (#18857)

  Show description for 38c3cd5
  sgugger

  authored
  38c3cd5

  Copy full SHA for 38c3cd5

• pin Slack SDK to 3.18.1 to avoid failing issue (#18869)

  Show description for 0ab465a
  ydshieh

  authored
  0ab465a

  Copy full SHA for 0ab465a

• Fix number of examples for iterable datasets in multiprocessing (#18856)

  Show description for 9e346f7
  sgugger

  authored
  9e346f7

  Copy full SHA for 9e346f7

• postpone bnb load until it's needed (#18859)

  stas00

  authored
  c5be7ca

  Copy full SHA for c5be7ca

• Generate: validate model_kwargs on TF (and catch typos in generate arguments) (#18651)

  gante

  authored
  9196f48

  Copy full SHA for 9196f48

• A script to download artifacts and perform CI error statistics (#18865)

  Show description for 4e29b3f
  ydshieh

  authored
  4e29b3f

  Copy full SHA for 4e29b3f

• Remove cached torch_extensions on CI runners (#18868)

  Show description for ecdf9b0
  ydshieh

  authored
  ecdf9b0

  Copy full SHA for ecdf9b0

• PEGASUS-X (#18551)

  Show description for 53e33e6
  zphang

  authored
  53e33e6

  Copy full SHA for 53e33e6

• Update docs landing page (#18590)

  Show description for bf9d506
  stevhliu

  authored
  bf9d506

  Copy full SHA for bf9d506

• Finetune guide for semantic segmentation (#18640)

  Show description for ae32f3a
  

  stevhliu

  and

  Steven

  authored
  ae32f3a

  Copy full SHA for ae32f3a

• Add Trainer to quicktour (#18723)

  Show description for 65fb71b
  stevhliu

  authored
  65fb71b

  Copy full SHA for 65fb71b

Commits on Sep 5, 2022

• TF: TFMarianMTModel final logits bias as a layer (#18833)

  Show description for 7f27e00
  gante

  authored
  7f27e00

  Copy full SHA for 7f27e00

• Mention TF and Flax checkpoints (#18894)

  LysandreJik

  authored
  591cfc6

  Copy full SHA for 591cfc6

• Correct naming pegasus x (#18896)

  Show description for badb9d2
  patrickvonplaten

  authored
  badb9d2

  Copy full SHA for badb9d2

• Update perf_train_gpu_one.mdx (#18442)

  thepurpleowl

  authored
  17c634f

  Copy full SHA for 17c634f

• Add type hints to XLM-Roberta-XL models (#18475)

  Show description for cfd623a
  asofiaoliveira

  authored
  cfd623a

  Copy full SHA for cfd623a

• Update Chinese documentation (#18893)

  Show description for c6d3dab
  zkep

  authored
  c6d3dab

  Copy full SHA for c6d3dab

• Generate: get the correct beam index on eos token (#18851)

  gante

  authored
  d4dbd7c

  Copy full SHA for d4dbd7c