"sgl-kernel/csrc/moe/moe_align_kernel.cu" did not exist on "00fa7d0417bf8d49b332c0cfc9a2647d6a9307fe"
- 05 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 03 Jul, 2019 1 commit
-
-
thomwolf authored
-
- 25 Jun, 2019 1 commit
-
-
thomwolf authored
-
- 11 Jun, 2019 1 commit
-
-
Meet Pragnesh Shah 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
-
- 02 May, 2019 1 commit
-
-
MottoX authored
-
- 29 Apr, 2019 1 commit
-
-
Mathieu Prouveur authored
-
- 24 Apr, 2019 1 commit
-
-
Mathieu Prouveur authored
-
- 23 Apr, 2019 1 commit
-
-
thomwolf authored
-
- 15 Apr, 2019 2 commits
- 21 Mar, 2019 1 commit
-
-
Yuqiang Xie authored
Maybe better.
-
- 15 Mar, 2019 1 commit
-
-
tseretelitornike authored
-
- 06 Mar, 2019 2 commits
- 11 Feb, 2019 1 commit
-
-
thomwolf authored
-
- 05 Feb, 2019 2 commits
- 30 Jan, 2019 1 commit
-
-
Matej Svejda authored
-
- 27 Jan, 2019 1 commit
-
-
Matej Svejda authored
-
- 07 Jan, 2019 1 commit
-
-
thomwolf authored
-
- 02 Jan, 2019 1 commit
-
-
Grégory Châtel authored
-
- 14 Dec, 2018 1 commit
-
-
thomwolf authored
-
- 13 Dec, 2018 3 commits
- 10 Dec, 2018 1 commit
-
-
Grégory Châtel authored
-
- 06 Dec, 2018 7 commits
-
-
Grégory Châtel authored
-
Grégory Châtel authored
-
Grégory Châtel authored
-
Grégory Châtel authored
-
Grégory Châtel authored
-
Grégory Châtel authored
-
Grégory Châtel authored
-
