Commits · 3c24e4bef1d7babbd69ecc95a8974b5a1c7171f6 · chenpangpang / transformers

"git@developer.sourcefind.cn:sugon_wxj/megatron-lm.git" did not exist on "05ea0cca6b0de8caa340ea52b716a72f3bea4571"

03 Nov, 2018 8 commits
- Multi-Gpu loss - Cleaning · 3c24e4be
  VictorSanh authored Nov 03, 2018
  
  3c24e4be
- Fix import on initalization · cd09cd5b
  VictorSanh authored Nov 03, 2018
  
  cd09cd5b
- Add seed in initialization · a1af5247
  VictorSanh authored Nov 03, 2018
  
  a1af5247
- update optimizer run_squad · 25f73add
  thomwolf authored Nov 03, 2018
  
  25f73add
- update run_squad with tqdm · f514cbbf
  thomwolf authored Nov 03, 2018
  
  f514cbbf
- add model.zero_grad() · cb76c1dd
  thomwolf authored Nov 03, 2018
  
  cb76c1dd
- typo on tokenization · 8bd6b235
  VictorSanh authored Nov 03, 2018
  
  8bd6b235
- Create DataParallel model if several GPUs · 5f432480
  VictorSanh authored Nov 03, 2018
  
  5f432480
02 Nov, 2018 14 commits
- Fix bug writing predictions in run_squad_pytorch · 38f740a1
  VictorSanh authored Nov 02, 2018
  
  38f740a1
- Debug run_squad_pytorch · ee29871f
  VictorSanh authored Nov 02, 2018
  
  ee29871f
- Debug run_squad_pytorch · 101eabff
  VictorSanh authored Nov 02, 2018
  
  101eabff
- Print for debug run_squad · bb0a5103
  VictorSanh authored Nov 02, 2018
  
  bb0a5103
- Move command-line argparse arguments into main() function · 3ff2ec5e
  Tim Rault authored Nov 02, 2018
  
  3ff2ec5e
- Bug fix type=bool -> action='store_true' in argparse · 49006d06
  Tim Rault authored Nov 02, 2018
  
  49006d06
- Fix size compatibility for model.forward · beb59080
  VictorSanh authored Nov 02, 2018
```
Error was coming from "modeling_pytorch.py", line 484, in forward: start_loss = loss_fct(start_logits, start_positions) --> ValueError: Expected target size (12, 1), got torch.Size([12])
```
  beb59080
- FIX errors in loading eval Dataset in `run_squad_pytorch` · 8cbe7d6a
  VictorSanh authored Nov 02, 2018
  
  8cbe7d6a
- FIX errors in loading Dataset in `run_squad_pytorch` · 833c3a7a
  VictorSanh authored Nov 02, 2018
  
  833c3a7a
- Fix small bug in `run_squad_pytorch.py` · 62ac7e9a
  VictorSanh authored Nov 02, 2018
  
  62ac7e9a
- updated extract_features · ebfffa0a
  thomwolf authored Nov 02, 2018
  
  ebfffa0a
- conversion run_squad ok · 9af479b3
  thomwolf authored Nov 02, 2018
  
  9af479b3
- working on squad · 8e81e5e6
  thomwolf authored Nov 02, 2018
  
  8e81e5e6
- run_squad WIP · e61db0d1
  thomwolf authored Nov 02, 2018
  
  e61db0d1
01 Nov, 2018 2 commits
- Convert flags to argparse in 'run_squad_pytorch.py' · 371945cf
  Tim Rault authored Nov 01, 2018
  
  371945cf
- Convert indentation from 2 spaces to 4 spaces · 8163baab
  Tim Rault authored Nov 01, 2018
  
  8163baab
31 Oct, 2018 1 commit
- switch to full google code · 13ee61e4
  thomwolf authored Oct 31, 2018
  
  13ee61e4