search

facebookresearch / Pearl

A Production-ready Reinforcement Learning AI Agent Library brought by the Applied Reinforcement Learning team at Meta.
MIT License
2.68k stars 166 forks source link

Frozen-lake tutorial #64

Closed mehrdad78 closed 9 months ago

mehrdad78 commented 9 months ago

Hi there! My question is about the plot you shown in frozen-lake example. What does it show to us? How can we understand that the problem is solved? image

hongjun7 commented 9 months ago

Hello, @mehrdad78. As the number of steps increases, we can observe an increase in the average return value. Furthermore, starting from around 10,000 steps, the return value consistently stabilizes around 1. This indicates that the model has been successfully trained in the right direction.