In this section, we will look at a few Gym environment wrappers that are especially very useful for the Gym Atari environments. Most of the wrappers we will implement in this section can be used with other environments as well to improve the learning performance of the agents.
The following table mentions a list of the wrappers will be implementing in the following section with a brief description for each of the wrappers to give you an overview:
| Wrapper | Brief description of the purpose |
|
ClipRewardEnv |
To implement reward clipping |
| AtariRescale | To rescale the screen pixels to a 84x84x1 gray scale image |
| NormalizedEnv | To normalize the images based on the mean and variance observed ... |
