Closed PierreExeter closed 4 years ago
It looks like the render
variable on the wrapped env is a boolean and not a function. Can you check that the render
member of your environment is a function?
Feel free to reopen this, but I am going to close it since it seems to be an issue with the env and not rlkit.
Apologies, I realized that I actually never replied. You were right, in my environment I had a boolean variable called render
that was overwriting the render()
method. I renamed it render_bool
to solve the issue.
Thanks
Hello,
First of all, thanks a lot for this very useful code! I have some issues with rendering a custom-made Gym environment after training it using the DDPG example.
I am able to visualise a trained policy using other Gym environments (for example cartpole-v0 trained using the DQN example script). I am also able to render the custom-made Gym environment with random action inputs (in the Pybullet engine). I am able to train the custom-made Gym environment using the DDPG example script. However, when running the trained policy,
I get the following error:
Do you have any idea what is causing this issue?
Many thanks in advance, Pierre