Show HN: Watch a neural net learn to play Snake

cc1b 1 day ago 6 commentsRead Article on ppo.gradexp.xyz

HI 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.

⚡ Community Insights

Discussion Sentiment

57% Positive

Analyzed from 90 words in the discussion.

Discussion (6 Comments)Read Original on HackerNews

simedw•about 2 hours ago

Cool project!

I noticed that if you go from training to watch and then back, the training temporarily drop significantly in score.

ldoughty•32 minutes ago

Really cool! But right as it was nearing 4,000, it seems to have corrupted itself and no longer got any scores above 0. Not sure if that's a code bug or a neural net issue.

avg500 -4.6 last 500 episodes

peak 3959.3 best window

roll/s 20.68 20-step avg

progress 4388 562749 episodes

jmclnx•11 minutes ago

> WebGPU not available in this browser

Looks like this is for Linux and Windows, on NetBSD I get this issue :(

beardsciences•about 1 hour ago

My average eventually made it to about 3900, and then stagnated between 3600-3900. I'm curious if this is universal behavior or not. I'm up to about 5k steps.

neduma•about 2 hours ago

More details and implementation notes please?

th1nhng0•about 1 hour ago

cool project