- 17 Apr, 2025 1 commit
-
-
wxj authored
-
- 19 Mar, 2025 2 commits
-
-
silencealiang authored
-
silencealiang authored
-
- 14 Mar, 2025 2 commits
-
-
wxj authored
-
silencealiang authored
-
- 17 Dec, 2024 2 commits
-
-
silencealiang authored
-
wxj authored
-
- 16 Dec, 2024 2 commits
- 09 Dec, 2024 4 commits
-
-
wxj authored
-
xingjinliang authored
-
silencealiang authored
-
xingjinliang authored
-
- 27 Mar, 2024 1 commit
-
-
liangjing authored
-
- 26 Apr, 2023 1 commit
-
-
zhuwenwen authored
-
- 09 Mar, 2023 1 commit
-
-
Maanu Grover authored
-
- 22 Feb, 2023 3 commits
-
-
Maanu Grover authored
-
Maanu Grover authored
-
Lawrence McAfee authored
-
- 17 Feb, 2023 1 commit
-
-
Maanu Grover authored
-
- 11 Jan, 2023 1 commit
-
-
huchen authored
-
- 09 Dec, 2022 1 commit
-
-
Tri Dao authored
-
- 24 Nov, 2022 1 commit
-
-
Boxin Wang authored
-
- 19 Jul, 2022 4 commits
-
-
Jared Casper authored
-
Lawrence McAfee authored
-
Lawrence McAfee authored
-
Lawrence McAfee authored
-
- 19 May, 2022 4 commits
-
-
Jared Casper authored
-
Jared Casper authored
-
Jared Casper authored
-
Jared Casper authored
-
- 11 Feb, 2022 1 commit
-
-
Jared Casper authored
-
- 27 Jan, 2022 2 commits
-
-
Stas Bekman authored
-
Stas Bekman authored
The paper has this info, so proposing to copy it next to the table. Otherwise it's hard to guess whether you used 40GB A100s or 80GB ones (and secondary, n_gpus per node). Thank you!
-
- 13 Dec, 2021 1 commit
-
-
Rajesh Koilpillai authored
-
- 30 Nov, 2021 2 commits
-
-
Jared Casper authored
-
Kamil Toraman authored
Remove duplicated bulletpoint
-
- 02 Nov, 2021 1 commit
-
-
James Reed authored
PP seems to have been added in https://github.com/NVIDIA/Megatron-LM/commit/46c74b4ca06a7794db1e2615544095535cdf12c2, so I think this clause is not accurate anymore
-
- 31 Oct, 2021 1 commit
-
-
Satpal Singh Rathore authored
-
- 18 Oct, 2021 1 commit
-
-
Jared Casper authored
-
