- 27 Dec, 2020 1 commit
-
-
mohammad authored
-
- 12 Nov, 2020 5 commits
-
-
Deepak Narayanan authored
Refactor code according to Jared's comments: move pipelining and non-pipelining training loops into separate methods Also, use mpu.get_*_model_parallel_size() instead of args.*_model_parallel_size
-
Deepak Narayanan authored
-
Deepak Narayanan authored
-
Deepak Narayanan authored
-
Deepak Narayanan authored
Also includes following changes for inter-layer model-parallel implementation: - Refactoring of model implementations - Training loop changes to support inter-layer communication using `ring_exchange` - New groups for inter-layer communication - Checkpoint changes - Command line arguments
-
- 23 Jul, 2020 1 commit
-
-
Neel Kant authored
-
- 07 Jul, 2020 1 commit
-
-
Neel Kant authored
-
- 29 Jun, 2020 1 commit
-
-
Neel Kant authored
-
- 24 Jun, 2020 1 commit
-
-
Neel Kant authored
-
- 22 Jun, 2020 1 commit
-
-
Neel Kant authored
-
- 10 Jun, 2020 1 commit
-
-
Neel Kant authored
-
- 09 Jun, 2020 1 commit
-
-
Neel Kant authored
-
- 05 Jun, 2020 3 commits
- 31 May, 2020 1 commit
-
-
Neel Kant authored
-
- 27 May, 2020 1 commit
-
-
Neel Kant authored
-
- 26 May, 2020 1 commit
-
-
Neel Kant authored
-
- 24 May, 2020 1 commit
-
-
Neel Kant authored
-
- 20 May, 2020 1 commit
-
-
Neel Kant authored
-
- 19 May, 2020 1 commit
-
-
Neel Kant authored
-
- 14 May, 2020 2 commits
- 12 May, 2020 1 commit
-
-
Neel Kant authored
-
- 07 May, 2020 2 commits
- 05 May, 2020 4 commits
- 03 May, 2020 1 commit
-
-
Neel Kant authored
-
- 29 Apr, 2020 1 commit
-
-
Neel Kant authored
-
- 24 Apr, 2020 3 commits
- 23 Apr, 2020 1 commit
-
-
Neel Kant authored
-
- 21 Apr, 2020 3 commits
