Commits · 29b7b30eaae52d97d4391353c562d8008f036b87 · chenpangpang / transformers

18 Jun, 2019 29 commits
- 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
- spliting config and weight files for bert also · 382e2d1e
  thomwolf authored Jun 18, 2019
  
  382e2d1e
17 Jun, 2019 10 commits
- Merge pull request #694 from huggingface/release_0.6.3 · a6f25118
  Thomas Wolf authored Jun 17, 2019
```
Release 0.6.3
```
  a6f25118
- updating hub · 4447f270
  thomwolf authored Jun 17, 2019
  
  4447f270
- updating head masking, readme and docstrings · 33d3db5c
  thomwolf authored Jun 17, 2019
  
  33d3db5c
- output all hidden layers states in GPT/GPT-2 · 965f172d
  thomwolf authored Jun 17, 2019
  
  965f172d
- add head masking and pruning to openai GPT · f12007e4
  thomwolf authored Jun 17, 2019
  
  f12007e4
- add head masking and pruning to gpt-2 · b860e47c
  thomwolf authored Jun 17, 2019
  
  b860e47c
- adding head pruning and tests · 7220d47a
  thomwolf authored Jun 17, 2019
  
  7220d47a
- better error messages · 8415a38b
  thomwolf authored Jun 17, 2019
  
  8415a38b
- add head masking tests · 96c4d3d9
  thomwolf authored Jun 17, 2019
  
  96c4d3d9
- adding bert whole words, bertgerman and gpt-2 medium models, head masking · 34858ae1
  thomwolf authored Jun 17, 2019
  
  34858ae1
15 Jun, 2019 1 commit
- Merge pull request #690 from shashwath94/projadpsftmax_fix · 80684f6f
  Thomas Wolf authored Jun 15, 2019
```
Transformer XL ProjectedAdaptiveLogSoftmax output fix
```
  80684f6f