huggingface / huggingface/open-muse
adafactor optimizer
- Dominant language
- Python
- Stars
- 359
- Forks
- 30
- PR merge metrics
- No merged PRs in 30d
Description
I'm planning to add in adafactor optimizer used in the official implementation. The main benefit of this over adam +adamw is that we don't need 3x the vram but I think a bit above 2x the vram of the models. I currently have the code up https://github.com/isamu-isozaki/adafactor-pytorch and after adding a triton version, I will bring a pr to here!
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by reviewing the linked adafactor-pytorch implementation and its planned Triton version, then compare the optimizer requirements with this repository's existing training entry points. The issue does not name files or tests; done would require an agreed integration scope, a working optimizer implementation, and validation of its memory and training behavior.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100