Live data from Hacker News

Show HN: Beating Pokemon Red with RL and <10M Parameters

drubinstein.github.io

61–70 of 71 posts

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#61
post #14

Earlier quoted context omitted.

It definitely cannot be beaten using random inputs. It doesn't even get out of Pallet Town after billions of steps. We tested...

the game has been beaten by fish

Based on the other examples of random inputs not being sufficient, I dare say the fish-based attempt may have been fraudulent.

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#62
Wow nice work. 10M is a tiny model and I suspect this might be the future for specialised work. I can also imagine the progress towards AGI/ASI to have smaller models used as submodules.

brains basically have “modules” like this as well - neuronal columns that handle specialised tasks. For example when you’re driving on the road, the understanding whether the distance between you and the vehicle in front is increasing or decreasing is a finely tuned function of a specialised part of the brain.

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#63
post #53

Earlier quoted context omitted.

if youve done the work to to make the rival rl based and have the ability to go around youd probably have added basic battle controls

as it stands, battling is wholly unimportant to completing the game, as long as the agents can eventually complete the trainer battles mandatory for plot advancement. it's funny because everyone thinks about battling when they think about pokemon. my first fn i wrote, back when we were still bumping around pallet town, was a battle reward function. it was trash and didn't work and was over-complicated. the crux of th…

you missed my point.

I know all about rl. Ive read go-explore 1/2, and I have personally implemented intrinsic curiosity.

I was just commenting on what rhe other person said, which is that it would be cool to have the npcs be agents that battle and train too, to which you said they could not be made to, to which I say, we have the technology. :)

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#64
post #53

Earlier quoted context omitted.

as it stands, battling is wholly unimportant to completing the game, as long as the agents can eventually complete the trainer battles mandatory for plot advancement. it's funny because everyone thinks about battling when they think about pokemon. my first fn i wrote, back when we were still bumping around pallet town, was a battle reward function. it was trash and didn't work and was over-complicated. the crux of th…

you missed my point. I know all about rl. Ive read go-explore 1/2, and I have personally implemented intrinsic curiosity. I was just commenting on what rhe other person said, which is that it would be cool to have the npcs be agents that battle and train too, to which you said they could not be made to, to which I say, we have the technology. :)

Sounds cool to me.

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#65
post #51

Earlier quoted context omitted.

Medical field, spotting things Autonomous drones Financial fraud detection Scheduling of trains/buses/etc I personally do like chatbots but you probably don't

the only chatbot for me is smarterchild

I feel like that sentence aged me.

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#67
post #40

Please stream the gameplay to twitch so people can compare.

We have a shared community map where you can watch hundreds of agents from multiple peoples training runs playing in real time! https://pwhiddy.github.io/pokerl-map-viz/

Can you make a twitch stream of a single agent playing?

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#68

Earlier quoted context omitted.

you dont port it you wrap it. you can put anything in an rl environment. usually emulators are done with bizhawk, and some lua. worst case theres ffi or screen capture.

Right, my thought was that this would be way too slow for episode rollout (versus an accelerated implementation in jax or something), but I guess not!

well thats the golden issue with rl, sample efficiency. it is env bounded, so you want an architecture that extracts the max possible information from each collected sample, avoiding catastrophic forgetting, prioritizing samples according to relevance

Re: Show HN: Beating Pokemon Red with RL and <10M Parameters

#69
post #8

Can't Pokemon be beaten by almost random play?

The win condition of the game is the entire state of the game configured in a certain way. So there exists a lot of win conditions, you just have to do a search.

not sure what you mean..details?
Post reply on HN