Commits · 5942af978a8a8ff706a302b1ba2d9ef3ce144444 · OpenDAS / Megatron-LM · GitLab

"test/srt/test_moe_deepep.py" did not exist on "8207637029082563cab74951fe8d5f86b574b85e"

24 Sep, 2022 1 commit
- Alias core.parallel_state as mpu and use it throughout code. RIP mpu. · 5942af97
  Jared Casper authored Sep 23, 2022
  
  5942af97
12 Jul, 2022 1 commit
- Add error checking by wrapping all weights and biases into named messages with named tensors. · 748a9a7d
  Jared Casper authored Jul 11, 2022
  
  748a9a7d
08 Mar, 2022 1 commit
- Transfer consumed train and valid samples to converted checkpoint. · 58cca6e9
  Jared Casper authored Mar 07, 2022
  
  58cca6e9
19 Feb, 2022 1 commit
- Making loading arguments from checkpoint cleaner and available more broadly. · 942c402d
  Jared Casper authored Feb 18, 2022
  
  942c402d
18 Feb, 2022 1 commit
- Better handling of padding in embedding table. · ec561daa
  Jared Casper authored Feb 18, 2022
  
  ec561daa
08 Feb, 2022 1 commit
- Don't require an even divide of layers in target model. · 96d19aae
  Jared Casper authored Feb 07, 2022
  
  96d19aae
02 Nov, 2021 1 commit
- Update to work with latest main branch. · 3b9dc880
  Jared Casper authored Nov 01, 2021
  
  3b9dc880
02 Sep, 2021 1 commit
- Adding checkpoint_util and associted loader and saver. · 03d09af0
  Jared Casper authored Jun 30, 2021
  
  03d09af0