Commits · cba85a67b9c2897593321012f3c6a575545e49e2 · chenpangpang / transformers

"git@developer.sourcefind.cn:Fzc7075/nunchaku.git" did not exist on "67db241001c9f7030198c4b814d9b671075a72e3"

15 Nov, 2018 2 commits
- fix nan in optimizer_on_cpu · cba85a67
  thomwolf authored Nov 15, 2018
  
  cba85a67
- preparing for first release · 1de35b62
  thomwolf authored Nov 15, 2018
  
  1de35b62
13 Nov, 2018 11 commits
- Merge pull request #17 from lukovnikov/master · 8513741b
  Thomas Wolf authored Nov 13, 2018
```
activation function in BERTIntermediate
```
  8513741b
- Merge remote-tracking branch 'origin/master' · 470076e4
  lukovnikov authored Nov 13, 2018
  
  470076e4
- clean up pr · 9f3cd271
  lukovnikov authored Nov 13, 2018
  
  9f3cd271
- Delete __init__.py · 3d4c7a6f
  Denis authored Nov 13, 2018
  
  3d4c7a6f
- clean up pr · d64db6df
  lukovnikov authored Nov 13, 2018
  
  d64db6df
- clean up pr · 7ba83730
  lukovnikov authored Nov 13, 2018
  
  7ba83730
- clean up pr · fa0c5a2e
  lukovnikov authored Nov 13, 2018
  
  fa0c5a2e
- Merge remote-tracking branch 'upstream/master' · f4d79f44
  lukovnikov authored Nov 13, 2018
  
  f4d79f44
- Merge pull request #16 from donatasrep/master · 5cd8d7ad
  Thomas Wolf authored Nov 13, 2018
```
Excluding AdamWeightDecayOptimizer internal variables from restoring
```
  5cd8d7ad
- Excluding AdamWeightDecayOptimizer internal variables from restoring · 20d07b3a
  Donatas Repecka authored Nov 13, 2018
  
  20d07b3a
- added results for 16-bit fine-tuning in readme · 278fd28a
  Thomas Wolf authored Nov 13, 2018
  
  278fd28a
12 Nov, 2018 5 commits
- typo · d940eeda
  thomwolf authored Nov 12, 2018
  
  d940eeda
- cleaning up readme · 1cf0a16c
  thomwolf authored Nov 12, 2018
  
  1cf0a16c
- add fp16 training · 66b00908
  thomwolf authored Nov 12, 2018
  
  66b00908
- fix typo in readme · 5dfd1906
  Thomas Wolf authored Nov 12, 2018
  
  5dfd1906
- fix typo in readme bach examples · fa1aa81f
  Thomas Wolf authored Nov 12, 2018
  
  fa1aa81f
11 Nov, 2018 2 commits
- update to BERT-large results · 6d6b916f
  Thomas Wolf authored Nov 11, 2018
  
  6d6b916f
- Add results of fine-tuning BERT-large on GPUs · c4bfc646
  Thomas Wolf authored Nov 11, 2018
  
  c4bfc646
10 Nov, 2018 2 commits
- Merge pull request #2 from elyase/patch-1 · 48930a4c
  Thomas Wolf authored Nov 10, 2018
```
Port tokenization for the multilingual model
```
  48930a4c
- fixing learning rate schedule when using gradient_accumulation_steps · a81a1ef8
  thomwolf authored Nov 10, 2018
  
  a81a1ef8
09 Nov, 2018 6 commits
- adding optimize_on_cpu explanation in readme · ea85cca8
  thomwolf authored Nov 09, 2018
  
  ea85cca8
- option to perform optimization and keep the optimizer averages on CPU · 5f04aa00
  thomwolf authored Nov 09, 2018
  
  5f04aa00
- clean up optimizer from unused functions · 9e95cd8c
  thomwolf authored Nov 09, 2018
  
  9e95cd8c
- update code comment · 34a1a010
  thomwolf authored Nov 09, 2018
  
  34a1a010
- remove duplicate accumulate gradient step arguments · 34bdc8b5
  thomwolf authored Nov 09, 2018
  
  34bdc8b5
- update results for SQuAD · 0c24db9d
  Thomas Wolf authored Nov 09, 2018
  
  0c24db9d
08 Nov, 2018 2 commits
- update readme - fix SQuAD model on multi-GPU · 2c5d993b
  thomwolf authored Nov 08, 2018
  
  2c5d993b
- fixed small typos in the README.md (#8) · 4850ec58
  Gopal Krishna authored Nov 09, 2018
  
  4850ec58
07 Nov, 2018 8 commits
- updating pytest command · 3bfbc213
  Thomas Wolf authored Nov 08, 2018
  
  3bfbc213
- Updated MRPC results · 0ed76961
  Thomas Wolf authored Nov 08, 2018
  
  0ed76961
- typo fix in output tuple · 48d4a531
  thomwolf authored Nov 07, 2018
  
  48d4a531
- Removing note on run_squad.py example · d92a7f77
  Thomas Wolf authored Nov 07, 2018
  
  d92a7f77
- Merge pull request #7 from huggingface/develop · 5c0838d8
  Thomas Wolf authored Nov 07, 2018
```
Develop
```
  5c0838d8
- Merge branch 'master' into develop · efeb6b1a
  Thomas Wolf authored Nov 07, 2018
  
  efeb6b1a
- cleaning up - speeding up a bit multi-gpu · dbc318a4
  thomwolf authored Nov 07, 2018
  
  dbc318a4
- fixing pre-processing bug - averaging loss for gradient accumulation - no_grad on evaluation · 6bb7510a
  thomwolf authored Nov 07, 2018
  
  6bb7510a
06 Nov, 2018 2 commits
- moved bert to qelos-util · bd91ae65
  lukovnikov authored Nov 06, 2018
  
  bd91ae65
- bert weight loading from tf · 4e521884
  lukovnikov authored Nov 06, 2018
  
  4e521884