Commits · 0ec8addb201c22971e3608a948dbc8576c59e6b8 · OpenDAS / apex

17 Sep, 2018 1 commit

Remove some fp16 examples that don't converge (#45) · 0ec8addb

Christian Sarofeen authored Sep 17, 2018

* Remove some fp16 examples that don't converge

Default static loss scale of 1.0 (default value) for resnet50 doesn't converge. Either remove example or put static loss scale 128 on it, which is known to converge well.

* Update README.md

0ec8addb

30 Aug, 2018 1 commit
- Update README.md · 586c507e
  mcarilli authored Aug 30, 2018
  
  586c507e
28 Aug, 2018 3 commits
- Updating imagenet README · 034b8f02
  Michael Carilli authored Aug 28, 2018
  
  034b8f02
- Cleaning up git weirdness + updating docs for Reducer · 37a4b221
  Michael Carilli authored Aug 28, 2018
  
  37a4b221
- Adjusting learning rate for batch size · dc41c5ce
  Michael Carilli authored Aug 27, 2018
  
  dc41c5ce
23 Jul, 2018 1 commit
- Ported examples to use torch.distributed.launch · aa817132
  Michael Carilli authored Jul 19, 2018
  
  aa817132
16 Jun, 2018 1 commit
- README wiring in a reasonable state, Sphinx docstrings updated · c41d9f2b
  Michael Carilli authored Jun 15, 2018
  
  c41d9f2b
15 Jun, 2018 2 commits
- More docstring + README updates · 5f8c3183
  Michael Carilli authored Jun 15, 2018
  
  5f8c3183
- Updating READMEs and examples · 82d7a3bf
  Michael Carilli authored Jun 15, 2018
  
  82d7a3bf
14 Jun, 2018 1 commit
- Added Dockerfile example, more readme updates · 0f703d13
  Michael Carilli authored Jun 14, 2018
  
  0f703d13
06 May, 2018 1 commit
- Adding main_fp16_optimizer.py to examples/imagenet and word_language_model · 0d91a65e
  Michael Carilli authored May 05, 2018
  
  0d91a65e
03 May, 2018 1 commit
- Adding examples that can serve as master copies for QA workflow moving forward. · 8ef50822
  Michael Carilli authored May 03, 2018
  
  8ef50822