Passing arguments

[javimontero]

But how can we pass arguments to the environment? I am struggling with that. I can do it using EnvContext for the training phase but for the prediction where you used gym.make???

Any tip?

[ceteri]

Good question @javimontero ,

Will this help? => https://stackoverflow.com/questions/54259338/how-to-pass-arguments-to-openai-gym-environments-upon-init

The call to gym.make() now takes kwarg parameters
https://github.com/openai/gym/blob/643b648fe20371cd87cd332f374dcf4f172941d5/gym/envs/registration.py#L77