- 19 Aug, 2019 1 commit
-
-
Peng Qi authored
-
- 18 Aug, 2019 1 commit
-
-
Chi-Liang Liu authored
-
- 05 Aug, 2019 2 commits
- 26 Jul, 2019 1 commit
-
-
zijunsun authored
-
- 24 Jul, 2019 1 commit
-
-
Chi-Liang Liu authored
-
- 23 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 21 Jul, 2019 1 commit
-
-
Peiqin Lin authored
-
- 17 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 15 Jul, 2019 2 commits
- 14 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 12 Jul, 2019 2 commits
- 10 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 05 Jul, 2019 2 commits
- 03 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 27 Jun, 2019 1 commit
-
-
thomwolf authored
-
- 26 Jun, 2019 1 commit
-
-
thomwolf authored
-
- 24 Jun, 2019 2 commits
- 19 Jun, 2019 1 commit
-
-
thomwolf authored
-
- 18 Jun, 2019 13 commits
- 22 May, 2019 1 commit
-
-
tguens authored
Indentation change so that the output "nbest_predictions.json" is not empty.
-
- 13 May, 2019 1 commit
-
-
samuelbroscheit authored
-
- 11 May, 2019 2 commits
-
-
samuel.broscheit authored
-
https://github.com/huggingface/pytorch-pretrained-BERT/issues/556samuel.broscheit authored
Reason for issue was that optimzation steps where computed from example size, which is different from actual size of dataloader when an example is chunked into multiple instances. Solution in this pull request is to compute num_optimization_steps directly from len(data_loader).
-
