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
- README: fix dead links (2 verified 404)
- Fix CLI overrides whose values contain '='
- perf(data): vectorize training batch extraction while preserving target layout
- fix(ddp): bypass collective wrapper in master-only validation
- fix(bench): avoid CUDA-only operations for CPU benchmarks
- fix(checkpoint): atomically replace checkpoints after complete saves
- perf(model): release underlying storage after cropping context
- fix(train): preserve the training sampler RNG across validation
- perf(checkpoint): stage weights on CPU before transferring to GPU
- fix(bench): synchronize the configured CUDA device for timing
- Docs
- not yet supported