AI impara a trascinarsi su 2 gambe

AI agent actually learning to move the 2-arme body through policy gradient.

[@ThinkingThings-q4p]

Let me know if you would like a tutorial :)

[@davidyoung5256]

Do you need very expensive motors with precision and feedback to do RL?

[@ThinkingThings-q4p]

@@davidyoung5256 Those servos are actually very cheap, couple of euros each. I am planning to make a full video tutorial in the next few weeks on this topic. Would you like to see It? Check out my lastest short, I am building also a spider-bot :)

[@davidyoung5256]

@@ThinkingThings-q4p Sure that would be interesting to see. I thought that a lot of precision was required to jump from sim to real

[@ThinkingThings-q4p]

@davidyoung5256 this Is true if you want to use the same policy, but not if you train from scratch :)