This is the $1000 version of https://github.com/karpathy/nanochat
- trained by Antigma Labs (https://antigma.ai)
This checkpoint does not have any additional fine tuning but has some more steps compared to the one @karpathy uploaded: https://huggingface.co/karpathy/nanochat-d32
- model_d20 is the smaller model for testing and benchmarking
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support