search

google-research / batch_rl

Offline Reinforcement Learning (aka Batch Reinforcement Learning) on Atari 2600 games
https://offline-rl.github.io/
Apache License 2.0
520 stars 73 forks source link

How to evaluate agents? #23

Closed Altriaex closed 1 year ago

Altriaex commented 3 years ago

I was able to train dqn agents using off-line data. I wonder how to evaluate agents ? e.g. reproduce the figure for Pong in Figure 3?

agarwl commented 3 years ago

So, the evaluation is done every iteration in dopamine here. Also, if you launched the code in the repository, the evaluation results should be logged in your tensorboard logs as well as dopamine logs.

agarwl commented 2 years ago

@Altriaex was this issue resolved?