Commits · 282c276e0920b85f0a2b4d348832f4dc83c1d69f · chenpangpang / transformers

"...models/git@developer.sourcefind.cn:Wenxuan/LightX2V.git" did not exist on "3fb9aa7e0d89ce29b6b39f38c609c8b4988921ca"

30 Aug, 2019 4 commits
- typos + file name coherence in distillation README · 282c276e
  VictorSanh authored Aug 30, 2019
  
  282c276e
- fix relative import bug cf Issue #1140 · 803c1cc4
  VictorSanh authored Aug 30, 2019
  
  803c1cc4
- remove reloading tokenizer in the training, adding it to the evaluation part · 39eb31e1
  Rabeeh KARIMI authored Aug 30, 2019
  
  39eb31e1
- updated tokenizer loading for addressing reproducibility issues · 350bb6bf
  Rabeeh KARIMI authored Aug 30, 2019
  
  350bb6bf
29 Aug, 2019 2 commits
- Closing bracket in DistilBERT's token count. · caf1d116
  LysandreJik authored Aug 29, 2019
  
  caf1d116
- Small modification of comment in the run_glue.py example · fe8fb10b
  Luis authored Aug 29, 2019
```
Add RoBERTa to the comment as it was not explicit that RoBERTa don't use token_type_ids.
```
  fe8fb10b
28 Aug, 2019 15 commits
- Changed learning rate for run_squad test · bf3dc778
  LysandreJik authored Aug 28, 2019
  
  bf3dc778
- swap order of optimizer.step() and scheduler.step() · 1d15a7f2
  Andreas Daiminger authored Aug 28, 2019
  
  1d15a7f2
- update credits · b5eb283a
  thomwolf authored Aug 28, 2019
  
  b5eb283a
- dilbert -> distilbert · 912a377e
  thomwolf authored Aug 28, 2019
  
  912a377e
- update readmes · 4ce5f36f
  thomwolf authored Aug 28, 2019
  
  4ce5f36f
- Write README for DilBERT · 93e82ab4
  VictorSanh authored Aug 28, 2019
  
  93e82ab4
- add licensing · fea921d3
  VictorSanh authored Aug 28, 2019
  
  fea921d3
- some fixes in `train.py` for loading previous checkpoint · da1e4e53
  VictorSanh authored Aug 28, 2019
  
  da1e4e53
- add `scripts/extract_for_distil.py` · 0d8f8848
  VictorSanh authored Aug 28, 2019
  
  0d8f8848
- add `scripts/token_counts.py` · 7f2c384c
  VictorSanh authored Aug 28, 2019
  
  7f2c384c
- add `scripts/binarized_data.py` · 4d16b279
  VictorSanh authored Aug 28, 2019
  
  4d16b279
- add `train.py` for distillation · b247b0d8
  VictorSanh authored Aug 28, 2019
  
  b247b0d8
- add requirements · 780f183e
  VictorSanh authored Aug 28, 2019
  
  780f183e
- add README · e424d2e4
  VictorSanh authored Aug 28, 2019
  
  e424d2e4
- add dataset. distiller, utils · 1ae81e4a
  VictorSanh authored Aug 28, 2019
  
  1ae81e4a
23 Aug, 2019 2 commits
- typo · 06510ccb
  thomwolf authored Aug 23, 2019
  
  06510ccb
- fixing tokenization and training · ab7bd5ef
  thomwolf authored Aug 23, 2019
  
  ab7bd5ef
22 Aug, 2019 2 commits
- fix for glue · 57272d5d
  VictorSanh authored Aug 22, 2019
  
  57272d5d
- fix for squad · b006a7a1
  VictorSanh authored Aug 22, 2019
  
  b006a7a1
20 Aug, 2019 7 commits
- Sequence special token handling for BERT and RoBERTa · 2d042274
  Lysandre authored Aug 20, 2019
  
  2d042274
- more fixes · 3bffd2e8
  Peng Qi authored Aug 20, 2019
  
  3bffd2e8
- various fix and clean up on run_lm_finetuning · a690edab
  thomwolf authored Aug 20, 2019
  
  a690edab
- swap optimizer.step and scheduler.step · d86b49ac
  Duzeyao authored Aug 20, 2019
  
  d86b49ac
- Revert "Update finetune_on_pregenerated.py" · 45ab8bf6
  Duzeyao authored Aug 20, 2019
```
This reverts commit a1359b97.
```
  45ab8bf6
- Update finetune_on_pregenerated.py · a1359b97
  Zeyao Du authored Aug 20, 2019
  
  a1359b97
- swap optimizer.step and scheduler.step · 28f7ca1f
  Zeyao Du authored Aug 20, 2019
  
  28f7ca1f
19 Aug, 2019 2 commits
- Fix #1015 · a368b877
  Peng Qi authored Aug 19, 2019
  
  a368b877
- Distributed training + tokenizer agnostic mask token · f94f1c60
  Lysandre authored Aug 19, 2019
  
  f94f1c60
18 Aug, 2019 4 commits
- don't save model without training · 40acf6b5
  Chi-Liang Liu authored Jul 30, 2019
  
  40acf6b5
- Fix: save model/model.module · 856a63da
  wangfei authored Aug 18, 2019
  
  856a63da
- Revert "Fix: save model/model.module" · 1ef41b83
  wangfei authored Aug 18, 2019
```
This reverts commit 00e9c4cc.
```
  1ef41b83
- Fix: save model/model.module · 00e9c4cc
  wangfei authored Aug 18, 2019
  
  00e9c4cc
16 Aug, 2019 2 commits
- Correct truncation for RoBERTa in 2-input GLUE · d8923270
  Jason Phang authored Aug 16, 2019
  
  d8923270
- Simplified data generator + better perplexity calculator · 5652f54a
  Lysandre authored Aug 16, 2019
```
GPT-2 now obtains ~20 perplexity on WikiText-2
```
  5652f54a