Ervin T.
Ervin T.
This is actually something of interest to us - and the weighting as well. One way you could do it with the current APIs is to have the weights as...
@robinerd Definitely would be interested in taking a look at your implementation. I think it's worth a PR!
An easy solution to this would be to just take the action that goes from -1 to 1, add 1, and divide by 2. That will get you a 0-1...
Good callout - I've updated the docs.
Imitation is still supported as GAIL and Behavior Cloning. Check out the docs here: https://github.com/Unity-Technologies/ml-agents/blob/release_17_docs/docs/ML-Agents-Overview.md#imitation-learning Also, the [forums](https://forum.unity.com/forums/ml-agents.453/) are a good spot for general questions - other users might be...
Thank you for your comments. We’ve documented your feedback and will prioritize when appropriate.
You'll have to update ML-Agents (Python code) to the latest version to use POCA.
Hey Howard, hope you figured this out yet. The "Root" folder is your option a). What are the errors, and do they clear if you restart Unity?
Hmm, this isn't a package used by Dodgeball or ML-Agents. Try removing the com.unity.purchasing package from the Package Manager.