Commits · 3ff2ec5eb38edab190c4bad278e9a2cc76f4ef7f · chenpangpang / transformers

"vscode:/vscode.git/clone" did not exist on "49759c0cda29ab614b81e0869972c99f2edba7aa"

02 Nov, 2018 25 commits
- Move command-line argparse arguments into main() function · 3ff2ec5e
  Tim Rault authored Nov 02, 2018
  
  3ff2ec5e
- adding jupyter, updating extract features adding simple test file · c9690e57
  thomwolf authored Nov 02, 2018
  
  c9690e57
- Small update Readme · 844b2f0e
  VictorSanh authored Nov 02, 2018
  
  844b2f0e
- 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
- Update README · 72d69a4e
  VictorSanh authored Nov 02, 2018
  
  72d69a4e
- Fix small bug in `run_squad_pytorch.py` · 62ac7e9a
  VictorSanh authored Nov 02, 2018
  
  62ac7e9a
- Quick fix metrics evaluation on run_classif_pytorch · 98b9771d
  VictorSanh authored Nov 02, 2018
  
  98b9771d
- Begin Updating the README.md · bf65d4db
  VictorSanh authored Nov 02, 2018
  
  bf65d4db
- Quick fix on eval accuracy · b54de837
  VictorSanh authored Nov 02, 2018
  
  b54de837
- Fix cuda compability for evaluation · 1d53f9cb
  VictorSanh authored Nov 02, 2018
  
  1d53f9cb
- FIX small bugs in `run_classifier_pytorch.py` · 1d8511f8
  VictorSanh authored Nov 02, 2018
  
  1d8511f8
- FIX small bugs in `run_classifier_pytorch.py` · 936eb4c3
  VictorSanh authored Nov 02, 2018
  
  936eb4c3
- Small fix to ensure Python3 compativility. · cc228089
  VictorSanh authored Nov 02, 2018
  
  cc228089
- 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
- implemented BertForQuestionAnswering · c0065af6
  thomwolf authored Nov 02, 2018
  
  c0065af6
- update name · 5383fca4
  thomwolf authored Nov 02, 2018
  
  5383fca4
- various fixes · dee09a40
  thomwolf authored Nov 02, 2018
  
  dee09a40
- small tweaks · 2c731fd1
  thomwolf authored Nov 02, 2018
  
  2c731fd1
- model training loop working – still have to check that everything is exactly same · 9343a231
  thomwolf authored Nov 02, 2018
  
  9343a231
01 Nov, 2018 15 commits
- run_classifier WIP + added classifier head and initialization to the model · f690f0e1
  thomwolf authored Nov 02, 2018
  
  f690f0e1
- run_classifier WIP · 4a0b59e9
  thomwolf authored Nov 01, 2018
  
  4a0b59e9
- Fix oubli · 7af7f817
  VictorSanh authored Nov 01, 2018
  
  7af7f817
- optimization_pytorch from OpenAI with adapted HP - check if there is weights decay on bias · e1bb7904
  thomwolf authored Nov 01, 2018
  
  e1bb7904
- typos in `input_fn_builder` · d3a8df6b
  VictorSanh authored Nov 01, 2018
  
  d3a8df6b
- wip · 836faed9
  VictorSanh authored Nov 01, 2018
  
  836faed9
- probably ok weights convertion script · 960ef4df
  thomwolf authored Nov 01, 2018
  
  960ef4df
- convertion script WIP · ab0e8932
  thomwolf authored Nov 01, 2018
  
  ab0e8932
- Merge remote-tracking branch 'origin/master' · 5581edb4
  Tim Rault authored Nov 01, 2018
  
  5581edb4
- Remove TensorFlow dependence in `tokenization_test_pytorch.py` · cdd39d21
  Tim Rault authored Nov 01, 2018
  
  cdd39d21
- added conversion script · c5d532e5
  thomwolf authored Nov 01, 2018
  
  c5d532e5
- WIP · 90d360a7
  VictorSanh authored Nov 01, 2018
  
  90d360a7
- Remove TensorFlow dependence in `tokenization_pytorch.py` · f8e7c95d
  Tim Rault authored Nov 01, 2018
  
  f8e7c95d
- working on model · 71557b16
  thomwolf authored Nov 01, 2018
  
  71557b16
- Convert flags to argparse in 'create_pretraining_data_pytorch.py' · 8627a675
  Tim Rault authored Nov 01, 2018
  
  8627a675