Commits · a6981076eca5494b9d230f13217c14b93443888a · chenpangpang / transformers

24 Sep, 2019 6 commits
- various updates · a6981076
  thomwolf authored Sep 24, 2019
  
  a6981076
- Fixed DistilBERT tokenizer · 72402d1a
  LysandreJik authored Sep 24, 2019
  
  72402d1a
- create_mask_from_sequences -> create_token_type_ids_from_sequences · d340e232
  LysandreJik authored Sep 24, 2019
  
  d340e232
- `output_token_type` -> `token_type_ids` · c832f43a
  LysandreJik authored Sep 24, 2019
  
  c832f43a
- Updated tests · 0ea82b24
  LysandreJik authored Sep 24, 2019
  
  0ea82b24
- Updated DistilBERT · 9d44236f
  LysandreJik authored Sep 24, 2019
  
  9d44236f
19 Sep, 2019 17 commits
- Python 2 compatibility · ab984a8b
  LysandreJik authored Sep 19, 2019
  
  ab984a8b
- Tokenizer accepts token list as well as string · 3df208c9
  LysandreJik authored Sep 19, 2019
  
  3df208c9
- prepare_for_model and prepare_pair_for_model methods. Added an option to... · 66ea76b8
  LysandreJik authored Sep 19, 2019
```
prepare_for_model and prepare_pair_for_model methods. Added an option to select which sequence will be truncated.
```
  66ea76b8
- Stride + tests + small fixes · baa74326
  LysandreJik authored Sep 19, 2019
  
  baa74326
- Mask computing in standalone method. Tests. · c10c7d59
  LysandreJik authored Sep 19, 2019
  
  c10c7d59
- Sentence -> Sequence. Removed output_mask from the special token addition methods. · bf503158
  LysandreJik authored Sep 19, 2019
  
  bf503158
- Doc + remove artefacts · 8cba0572
  LysandreJik authored Sep 19, 2019
  
  8cba0572
- encode + encode_plus tests modified · 6393261e
  LysandreJik authored Sep 19, 2019
  
  6393261e
- Modified encode to return only lists. Added a more complete encode_plus method · dcc9bb32
  LysandreJik authored Sep 19, 2019
  
  dcc9bb32
- Max encoding length + corresponding tests · af23b626
  LysandreJik authored Sep 11, 2019
  
  af23b626
- Updated DistilBERT test to reflect the sequence encoding · c4d4f3ec
  LysandreJik authored Sep 11, 2019
  
  c4d4f3ec
- Number of added tokens calculator · d572d702
  LysandreJik authored Sep 11, 2019
  
  d572d702
- Changed warning to be more explicit · 2d8ec5a6
  LysandreJik authored Sep 06, 2019
```
Co-authored by: julien_c <chaumond@gmail.com>
```
  2d8ec5a6
- Distilbert sequence builder w/ mask · 92a9976e
  LysandreJik authored Sep 06, 2019
  
  92a9976e
- Updated the GLUE data processor. Corrections to RoBERTa and XLNet. · bac332fe
  LysandreJik authored Sep 02, 2019
  
  bac332fe
- Added binary masking tests · c3df2136
  LysandreJik authored Sep 02, 2019
  
  c3df2136
- Tokenizers' encode function can output binary masks · e391d473
  LysandreJik authored Sep 02, 2019
  
  e391d473
17 Sep, 2019 2 commits
- tiny fixes · 62760baf
  Julien Chaumond authored Sep 17, 2019
  
  62760baf
- fix #1223 · 45de034b
  thomwolf authored Sep 17, 2019
  
  45de034b
16 Sep, 2019 2 commits
- fixed init_weight · a9debaca
  erenup authored Sep 16, 2019
  
  a9debaca
- add doc string · 4812a5a7
  erenup authored Sep 16, 2019
  
  4812a5a7
11 Sep, 2019 1 commit
- fixed: hard coding for max and min number will out of range in fp16, which will cause nan. · 8bdee1cb
  Zili Wang authored Sep 11, 2019
  
  8bdee1cb
10 Sep, 2019 1 commit
- Fixing typo in gpt2 for doc site's class link · f2cf6ce4
  mattolson93 authored Sep 10, 2019
  
  f2cf6ce4
09 Sep, 2019 1 commit
- fix #1208 · 3401980f
  thomwolf authored Sep 09, 2019
  
  3401980f
05 Sep, 2019 4 commits
- adding test for common properties and cleaning up a bit base class · 5c6cac10
  thomwolf authored Sep 05, 2019
  
  5c6cac10
- Fix typo · d7379477
  maru0kun authored Sep 05, 2019
  
  d7379477
- also gathering file names in file_utils · 85df4f7c
  thomwolf authored Sep 05, 2019
  
  85df4f7c
- update conversion scripts · 121f88ca
  thomwolf authored Sep 05, 2019
  
  121f88ca
04 Sep, 2019 6 commits
- clean ups · d77abd4d
  thomwolf authored Sep 05, 2019
  
  d77abd4d
- split configuration and modeling files · 2a667b1e
  thomwolf authored Sep 05, 2019
  
  2a667b1e
- be sure we have uint8 · 0be6a2a6
  thomwolf authored Sep 04, 2019
  
  0be6a2a6
- WIP reordering · 7fba47b7
  thomwolf authored Sep 04, 2019
  
  7fba47b7
- WIP reodering arguments for torchscript and TF · e25cba78
  thomwolf authored Sep 04, 2019
  
  e25cba78
- Fixing this TransformerXL bool issue · 38b79b5a
  thomwolf authored Sep 04, 2019
  
  38b79b5a