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
- switch model to eval mode during inference to prevent unwanted dropout
- Realtime response generation simulation instead of one big response at the end.
- ft:Response streaming added
- Public soft prefix cache
- model: handle temperature=0 in generate with greedy argmax
- GPT.generate() crashes on temperature=0 instead of doing greedy decoding
- ruizhe-experiment – Add ablation analysis scripts, extended quantisation layers, and missing baseline configs
- fix: surface implicit contracts and silent failure modes — Hokmah arc…
- week3: add configurable module experiments and toy MoE
- Code quality scan: 9 findings (C-, 68/100)
- Docs
- not yet supported