- 05 May, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 03 May, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 29 Apr, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 28 Apr, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 27 Apr, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 04 Apr, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 29 Mar, 2021 1 commit
-
-
TiagoMAntunes authored
-
- 28 Mar, 2021 2 commits
-
-
TiagoMAntunes authored
-
TiagoMAntunes authored
-
- 25 Mar, 2021 6 commits
-
-
TiagoMAntunes authored
-
Rick Ho authored
-
Jiezhong Qiu authored
citation from arxiv
-
Rick Ho authored
-
Jiezhong Qiu authored
fix tests after updating megatron
-
Rick Ho authored
-
- 23 Mar, 2021 4 commits
-
-
TiagoMAntunes authored
-
Rick Ho authored
-
TiagoMAntunes authored
Bias now being calculated directly in MOELinear layer. Added corresponding CUDA changes. Updated forward and backward functions of MOELinear
-
TiagoMAntunes authored
-
- 22 Mar, 2021 21 commits
-
-
Rick Ho authored
Split megatron python module
-
Rick Ho authored
-
Rick Ho authored
-
Rick Ho authored
-
Rick Ho authored
-
Sengxian authored
-
Rick Ho authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Jiezhong Qiu authored
-
Rick Ho authored
Add balance strategy
-
Sengxian authored
-