DE version is available. Content is displayed in original English for accuracy.
In browser PPO training demo, made possible by tinygrad: TinyJit -> WebGPU kernels.
Requires WebGPU.
DE version is available. Content is displayed in original English for accuracy.
Requires WebGPU.
Discussion (4 Comments)Read Original on HackerNews
I noticed that if you go from training to watch and then back, the training temporarily drop significantly in score.