Commits · cd65c41a83da9121f8260490584d9a21b47cc4d8 · chenpangpang / transformers

30 Aug, 2019 5 commits
- added test and debug tokenizer configuration serialization · 69da972a
  thomwolf authored Aug 30, 2019
  
  69da972a
- saving and reloading tokenizer configurations · 88111de0
  thomwolf authored Aug 30, 2019
  
  88111de0
- adding xlm 17 and 100 models and config on aws · 3871b8a1
  thomwolf authored Aug 30, 2019
  
  3871b8a1
- adding 17 and 100 xlm models · 8678ff8d
  thomwolf authored Aug 30, 2019
  
  8678ff8d
- Added option to setup pretrained tokenizer arguments · 82462c5c
  thomwolf authored Aug 30, 2019
  
  82462c5c
28 Aug, 2019 20 commits
- fix #1131 · 0a74c88a
  thomwolf authored Aug 28, 2019
  
  0a74c88a
- distilbert: fix number of hidden_size · ed2ab1c2
  Stefan Schweter authored Aug 28, 2019
  
  ed2ab1c2
- update again · e7706f51
  thomwolf authored Aug 28, 2019
  
  e7706f51
- update credits · b5eb283a
  thomwolf authored Aug 28, 2019
  
  b5eb283a
- Removed typings for Python 2 · f753d4e3
  LysandreJik authored Aug 28, 2019
  
  f753d4e3
- Updated article link · 75bc2a03
  LysandreJik authored Aug 28, 2019
  
  75bc2a03
- Documentation additions · 1dc43e56
  LysandreJik authored Aug 28, 2019
  
  1dc43e56
- dilbert -> distilbert · 912a377e
  thomwolf authored Aug 28, 2019
  
  912a377e
- fixing model to add torchscript, embedding resizing, head pruning and masking + tests · c9bce181
  thomwolf authored Aug 28, 2019
  
  c9bce181
- add dilbert tokenizer and tests · 62df4ba5
  thomwolf authored Aug 28, 2019
  
  62df4ba5
- fix typo · a5fe1668
  VictorSanh authored Aug 28, 2019
  
  a5fe1668
- add DilBert model for squad · 19b7c9b0
  VictorSanh authored Aug 28, 2019
  
  19b7c9b0
- Dilbert tests from CommonTests · c513415b
  LysandreJik authored Aug 27, 2019
  
  c513415b
- GilBert added to AutoModels · 778a263f
  LysandreJik authored Aug 27, 2019
  
  778a263f
- fix: add qa_dropout and seq_classif_dropout · 74d78bee
  VictorSanh authored Aug 28, 2019
  
  74d78bee
- fix small typo · 7f5d8534
  VictorSanh authored Aug 28, 2019
  
  7f5d8534
- add s3 links for dilbert (+fix small typo) · 906581ae
  VictorSanh authored Aug 28, 2019
  
  906581ae
- fix bugs · 5d29f8e9
  VictorSanh authored Aug 28, 2019
  
  5d29f8e9
- fix bugs · a8ad8304
  VictorSanh authored Aug 28, 2019
  
  a8ad8304
- Match order of casing in OSS XLM; Improve document; Clean up dependency · ca4baf8c
  Shijie Wu authored Aug 27, 2019
  
  ca4baf8c
27 Aug, 2019 7 commits
- fix bugs · 60c984da
  VictorSanh authored Aug 27, 2019
  
  60c984da
- wip wouf · 42968138
  VictorSanh authored Aug 27, 2019
  
  42968138
- wip · 1d232400
  VictorSanh authored Aug 27, 2019
  
  1d232400
- add kwargs to base encode function · a175a9dc
  thomwolf authored Aug 27, 2019
  
  a175a9dc
- Change attention mask dtype to be bool. Fix #1119 · 53282b5b
  Nikolay Korolev authored Aug 27, 2019
  
  53282b5b
- Fix documentation #1117 · 26bda772
  Nikolay Korolev authored Aug 27, 2019
```
Rename parameter in documentation + Delete its second occurrence.
```
  26bda772
- Delete nonexistent parameter from documentation · c8933bb2
  Nikolay Korolev authored Aug 27, 2019
```
Changed documentation of GPT2Model, GPT2LMHeadModel and GPT2DoubleHeadsModel
```
  c8933bb2
26 Aug, 2019 2 commits
- fix #1102 · e08c01aa
  LysandreJik authored Aug 26, 2019
  
  e08c01aa
- Generic encoding implementation. · 529a16de
  LysandreJik authored Aug 26, 2019
  
  529a16de
24 Aug, 2019 2 commits
- Add use_lang_emb to config · f1b01874
  Shijie Wu authored Aug 23, 2019
  
  f1b01874
- Add custom tokenizer for zh and ja · e85123d3
  Shijie Wu authored Aug 23, 2019
  
  e85123d3
23 Aug, 2019 4 commits
- max_len_single_sentence & max_len_sentences_pair as attributes so they can be modified · 3bcbebd4
  thomwolf authored Aug 23, 2019
  
  3bcbebd4
- Tokenization behave the same as original XLM proprocessing for most languages... · 436ce072
  Shijie Wu authored Aug 23, 2019
```
Tokenization behave the same as original XLM proprocessing for most languages except zh, ja and th; Change API to allow specifying language in `tokenize`
```
  436ce072
- adding max_lengths for single sentences and sentences pairs · 47d68534
  thomwolf authored Aug 23, 2019
  
  47d68534
- change layernorm code to pytorch's native layer norm · e13465fb
  David Pollack authored Aug 23, 2019
  
  e13465fb