Wrappers

We provide a set of wrappers based on rllab to wrap OpenAI Gym environments, these typically will filter observations, etc.

These wrappers require rllab as a dependency, but we’re working on removing this dependency. Please submit a Pull Request if you want it done sooner!

Current Wrappers

We only provide one wrapper for now:

ObservationTransformWrapper - Provides a list of transformers and runs all observations through the transformers.

Current Transformers

SimpleNormalizePixelIntensitiesTransformer - Normalizes pixels (divide by 255)

ResizeImageTransformer - Resizes an image by a given percentage

RandomSensorMaskTransformer - Randomly occludes some given percentage of the observations at every timestep

Example Usage

gymenv = gym.make("Hopper-v1")
transformers = [SimpleNormalizePixelIntensitiesTransformer(), ResizeImageTransformer(fraction_of_current_size=.35)]
transformed_env = ObservationTransformWrapper(gymenv, transformers)

gym-extensions

This repo is intended as an extension for OpenAI Gym for auxiliary tasks (multitask learning, transfer learning, inverse reinforcement learning, etc.)

Wrappers

Current Wrappers

Current Transformers

Example Usage