Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sunblaze-ucb/rl-generalization
Modifiable OpenAI Gym environments for studying generalization in RL
https://github.com/sunblaze-ucb/rl-generalization
Last synced: 3 months ago
JSON representation
Modifiable OpenAI Gym environments for studying generalization in RL
- Host: GitHub
- URL: https://github.com/sunblaze-ucb/rl-generalization
- Owner: sunblaze-ucb
- License: mit
- Created: 2018-10-21T01:09:24.000Z (over 6 years ago)
- Default Branch: master
- Last Pushed: 2019-01-22T05:01:26.000Z (about 6 years ago)
- Last Synced: 2024-08-02T07:06:58.905Z (6 months ago)
- Language: Python
- Homepage:
- Size: 168 KB
- Stars: 85
- Watchers: 12
- Forks: 14
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
- awesome-deep-rl - Berkeley rl-generalization - Modifiable OpenAI Gym environments for studying generalization in RL. (Environments)
README
# RL environments for evaluating generalization
This repo contains a set of environments (based on [OpenAI Gym](https://gym.openai.com) and [Roboschool](https://github.com/openai/roboschool)), designed for evaluating generalization in reinforcement learning. We also include implementations of several deep reinforcement learning algorithms (based on [OpenAI Baselines](https://github.com/openai/baselines)), which we have evaluated on these environments.
All environments tested using Python 3.
## Installation
### Using virtualenv
We recommend that you install inside a [virtualenv](https://packaging.python.org/guides/installing-using-pip-and-virtualenv/#creating-a-virtualenv).
Install the environments by checking out this repository and running:
```sh
pip3 install --process-dependency-links -e .
```
Some examples of agents using the environments are provided in the `examples`
directory. They require some additional dependencies, which can be installed by
running:
```sh
pip3 install --process-dependency-links -e .[examples]
```
To get a list of all provided environments, you can run:
```sh
python3 -m examples.list_environments
```
Install Roboschool separately following the instructions [here](https://github.com/openai/roboschool#installation).
### Using Docker
You can use Docker to avoid issues while installing dependencies such as Roboschool. You can clone the following Docker image which has all of the dependencies installed:
```sh
# download docker image
docker pull cpacker/rl-generalization
# start an interactive bash session
docker run -v /path/to/your/copy/of/rl-generalization:/rl-generalization -it cpacker/rl-generalization /bin/bash
# (inside container)
cd /rl-generalization
python3 -m examples.list_environments
```
## Using the modified environments
You can substitute our environments anywhere you would use an OpenAI Gym environment. For example, instead of:
```python
import gym
env = gym.make('CartPole-v0')
```
You can use one of the following environments:
```python
import gym
import sunblaze_envs
# Deterministic: the default version with fixed parameters
fixed_env = sunblaze_envs.make('SunblazeCartPole-v0')
# Random: parameters are sampled from a range nearby the default settings
random_env = sunblaze_envs.make('SunblazeCartPoleRandomNormal-v0')
# Extreme: parameters are sampled from an `extreme' range
extreme_env = sunblaze_envs.make('SunblazeCartPoleRandomExtreme-v0')
```
In the case of CartPole, RandomNormal and RandomExtreme will vary the strength of each actions, the mass of the pole, and the length of the pole:
Specific ranges for each environment setting are listed [here](sunblaze_envs#environment-details). See the code in [examples](/examples) for usage with example algorithms from OpenAI Baselines.
## Citations
To cite this repository in your research, you can reference the following [paper](https://arxiv.org/abs/1810.12282):
> Charles Packer, Katelyn Gao, Jernej Kos, Philipp Krähenbühl, Vladlen Koltun, and Dawn Song. Assessing Generalization in Deep Reinforcement Learning. *arXiv preprint arXiv:1810.12282* (2018).
```TeX
@misc{PackerGao:1810.12282,
Author = {Charles Packer and Katelyn Gao and Jernej Kos and Philipp Kr\"ahenb\"uhl and Vladlen Koltun and Dawn Song},
Title = {Assessing Generalization in Deep Reinforcement Learning},
Year = {2018},
Eprint = {arXiv:1810.12282},
}
```