Commits · 7f00a36e27ce89dea736a2bf52cc3073cf7911ec · chenpangpang / transformers

19 Jun, 2019 9 commits
- pruning should keep on device · 7f00a36e
  thomwolf authored Jun 19, 2019
  
  7f00a36e
- update head pruning · e4b46d86
  thomwolf authored Jun 19, 2019
  
  e4b46d86
- debugger · 0f40e8d6
  thomwolf authored Jun 19, 2019
  
  0f40e8d6
- more logging · 0e1e8128
  thomwolf authored Jun 19, 2019
  
  0e1e8128
- cuda again · 909d4f1a
  thomwolf authored Jun 19, 2019
  
  909d4f1a
- fix cuda · 14f0e8e5
  thomwolf authored Jun 19, 2019
  
  14f0e8e5
- pruning in bertology · 34d706a0
  thomwolf authored Jun 19, 2019
  
  34d706a0
- updating examples · dc8e0019
  thomwolf authored Jun 19, 2019
  
  dc8e0019
- small fix and updates to readme · 68ab9599
  thomwolf authored Jun 19, 2019
  
  68ab9599
18 Jun, 2019 31 commits
- update barrier · f7e2ac01
  thomwolf authored Jun 18, 2019
  
  f7e2ac01
- test barrier in distrib training · 4d8c4337
  thomwolf authored Jun 18, 2019
  
  4d8c4337
- updating run_classif · 33599556
  thomwolf authored Jun 18, 2019
  
  33599556
- updating evaluation on a single gpu · 29b7b30e
  thomwolf authored Jun 18, 2019
  
  29b7b30e
- overwrite_output_dir · 7d2001aa
  thomwolf authored Jun 18, 2019
  
  7d2001aa
- fixing · 16a1f338
  thomwolf authored Jun 18, 2019
  
  16a1f338
- no numpy · 92e0ad5a
  thomwolf authored Jun 18, 2019
  
  92e0ad5a
- hop · 4e6edc32
  thomwolf authored Jun 18, 2019
  
  4e6edc32
- fixing again · f55b60b9
  thomwolf authored Jun 18, 2019
  
  f55b60b9
- quick fix · 8bd91182
  thomwolf authored Jun 18, 2019
  
  8bd91182
- fix out_label_ids · 3e847449
  thomwolf authored Jun 18, 2019
  
  3e847449
- fix paths · aad3a54e
  thomwolf authored Jun 18, 2019
  
  aad3a54e
- updating classification example · 40dbda68
  thomwolf authored Jun 18, 2019
  
  40dbda68
- update run_classifier for distributed eval · 7388c83b
  thomwolf authored Jun 18, 2019
  
  7388c83b
- fix pickle · 97277232
  thomwolf authored Jun 18, 2019
  
  97277232
- fix pickles · 9710b68d
  thomwolf authored Jun 18, 2019
  
  9710b68d
- cache in run_classifier + various fixes to the examples · 15ebd67d
  thomwolf authored Jun 18, 2019
  
  15ebd67d
- fix · e6e5f192
  thomwolf authored Jun 18, 2019
  
  e6e5f192
- distributed traing t_total · a432b3d4
  thomwolf authored Jun 18, 2019
  
  a432b3d4
- split squad example in two · c5407f34
  thomwolf authored Jun 18, 2019
  
  c5407f34
- only on main process · 335f57ba
  thomwolf authored Jun 18, 2019
  
  335f57ba
- add tensorboard to run_squad · 326944d6
  thomwolf authored Jun 18, 2019
  
  326944d6
- set find_unused_parameters=True in DDP · d82e5dee
  thomwolf authored Jun 18, 2019
  
  d82e5dee
- DDP update · a59abedf
  thomwolf authored Jun 18, 2019
  
  a59abedf
- switch to pytorch DistributedDataParallel · 2ef5e0de
  thomwolf authored Jun 18, 2019
  
  2ef5e0de
- oups · 9ce37af9
  thomwolf authored Jun 18, 2019
  
  9ce37af9
- no need to duplicate models anymore · a40955f0
  thomwolf authored Jun 18, 2019
  
  a40955f0
- Merge pull request #696 from huggingface/split_config_weights · 3763f894
  Thomas Wolf authored Jun 18, 2019
```
Split config weights
```
  3763f894
- explanation on the current location of the caching folder · f9647530
  thomwolf authored Jun 18, 2019
  
  f9647530
- updating weights loading · 868de8d1
  thomwolf authored Jun 18, 2019
  
  868de8d1
- better error message · 64e0adda
  thomwolf authored Jun 18, 2019
  
  64e0adda