nanogpt
https://github.com/karpathy/nanogpt
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
not yet supported1 Subscribers
Add a CodeTriage badge to nanogpt
Help out
- Issues
- Refactor for easier configuration and overrides
- nano_gpt
- Fix: conditional use of GradScaler based on device_type and dtype in train.py
- Shouldnt the ddp check be on ZERO instead of -1
- sign descent seems to do better than adamw?
- Overfitting of the small GPU model
- fix(train.py): mfu estimation to respect CPU-GPU sync point
- fix val dataset size code comment
- Data cleaning for openwebtext
- free up state_dict variable memory after loading checkpoint
- Docs
- not yet supported