- 10 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 09 Jul, 2019 5 commits
- 05 Jul, 2019 5 commits
- 03 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 27 Jun, 2019 1 commit
-
-
thomwolf authored
-
- 24 Jun, 2019 2 commits
- 19 Jun, 2019 3 commits
- 18 Jun, 2019 17 commits
-
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
thomwolf authored
-
- 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).
-
- 09 May, 2019 2 commits
-
-
burcturkoglu authored
-
burcturkoglu authored
-
