google / dopamine

Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.
https://github.com/google/dopamine
Apache License 2.0
10.42k stars 1.36k forks source link

Preprocessing the state at the beginning of episode #202

Closed mklissa closed 1 year ago

mklissa commented 1 year ago

It seems like preprocessing is missing at the beginning of episodes for the Jax Rainbow agent. Not sure how much it affects the results but it's a quick fix.

google-cla[bot] commented 1 year ago

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

JesseFarebro commented 1 year ago

In addition, I believe preprocess_fn is also missing when initializing the network parameters.

psc-g commented 1 year ago

thanks for pointing this out!

for annoying technical reasons, i have to incorporate this change on our end and then push it out to github. i'll acknowledge this PR when doing so, and let you know when it's pushed.

thanks!

-psc

On Wed, Mar 22, 2023 at 10:00 PM Jesse Farebrother @.***> wrote:

In addition, I believe preprocess_fn is also missing when initializing the network parameters.

— Reply to this email directly, view it on GitHub https://github.com/google/dopamine/pull/202#issuecomment-1480476553, or unsubscribe https://github.com/notifications/unsubscribe-auth/AE3CCMNRSYCCFQBOF7PHF2DW5OVEJANCNFSM6AAAAAAWDPP3KE . You are receiving this because you are subscribed to this thread.Message ID: @.***>

psc-g commented 1 year ago

This has been fixed in ce36aab6528b26a699f5f1cefd330fdaf23a5d72