Commits · 585f177b266e6233f05b0e3a72eacb351213ab11 · OpenDAS / fairscale

"vscode:/vscode.git/clone" did not exist on "1ac38ec89c5899f44f84e44ee461c714544c6af0"

14 Aug, 2020 1 commit

[fix] Properly restore a sharded optim state (#39) · 585f177b

Benjamin Lefaudeux authored Aug 14, 2020



* hotfix a half-cooked optimizer state restoration, the global shared state also needs to be restored

* [cleanup] get 100% coverage on oss.py (#38)
authored-by: Mandeep Singh Baines <msb@fb.com>

* better unit testing, check that the .param_groups attribute is properly in sync with the loaded state
Co-authored-by: msbaines <35972327+msbaines@users.noreply.github.com>

585f177b

13 Aug, 2020 1 commit

Aligning OSS state dict with... · 57079b08

Benjamin Lefaudeux authored Aug 12, 2020

Aligning OSS state dict with `https://pytorch.org/docs/stable/_modules/torch/optim/optimizer.html#Optimizer` (#31)

57079b08

08 Aug, 2020 1 commit
- [fix] fix test_oss.py when host have 2 GPUs (#26) · d9e6ceaa
  Min Xu authored Aug 07, 2020
```
Co-authored-by: Min Xu <m1n@fb.com>
```
  d9e6ceaa
06 Aug, 2020 1 commit
- [feat] add ddp that works with oss with reduce() not all_reduce() (#19) · 525e709b
  Min Xu authored Aug 06, 2020
```
Co-authored-by: Min Xu <m1n@fb.com>
```
  525e709b
31 Jul, 2020 3 commits
- [feat] Implement OSS save and load of the sharded state from a single replica (#16) · 8e363567
  Benjamin Lefaudeux authored Jul 31, 2020
  
  8e363567
- [feat] add FusedAdam (#10) · bfba68d8
  Jun Ru Anderson authored Jul 30, 2020
```
Add FusedAdam, update benchmark and add tests.
Co-authored-by: Jun Ru Anderson <andersonic@fb.com>
```
  bfba68d8
- [feat] Model parallel (#3) · 30f5009a
  Tom Birch authored Jul 22, 2020
  
  30f5009a
08 Jul, 2020 1 commit
- Initial commit · 0cd65242
  Mandeep Singh Baines authored Jul 07, 2020
  
  0cd65242