Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jlenon7/mountaincar_gym_env
⛰️ Gym environment solving the MountainCar problem using Continuous Q-Learning.
https://github.com/jlenon7/mountaincar_gym_env
Last synced: 9 days ago
JSON representation
⛰️ Gym environment solving the MountainCar problem using Continuous Q-Learning.
- Host: GitHub
- URL: https://github.com/jlenon7/mountaincar_gym_env
- Owner: jlenon7
- License: mit
- Created: 2024-06-13T11:46:53.000Z (6 months ago)
- Default Branch: main
- Last Pushed: 2024-06-23T08:53:25.000Z (6 months ago)
- Last Synced: 2024-12-06T07:46:56.195Z (17 days ago)
- Language: Python
- Homepage:
- Size: 785 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# OpenAI Mountain Car Gym Environment ⛰️
> Gym environment solving the MountainCar problem using Continuous Q-Learning.
## Results
### Simple agent results
### Continuous Q-Learning agent results
### Epoch points log tracker
## TODOs
- [x] Create simple agent
- [x] Train the agent to land exactly on the flag
- [ ] Save the model to be reused with a library like [stable-baselines](https://stable-baselines3.readthedocs.io/en/master/)
## Running
To run the gym environment first create a new Python environment and activate it. I'm using [Anaconda](https://www.anaconda.com/) for setting the python version that pipenv should use to set up the environment. The command bellow will automatically setup the environment with conda and pipenv:
```shell
make env
```
Now install all the project dependencies:
```shell
make install-all
```
To run the game to be played by a human run:
```shell
make play
```
To run the game and run random actions run:
```shell
make sample
```
To run the agent to complete the game task run:
```shell
make agent
```
To run the simple agent run:
```shell
make simple-agent
```
> [!WARNING]
> Keep in mind that the simple agent is not using rewards to perform the actions.
> Instead is just observing the environment to use the car position and
> velocity to use gravity to accomplish the task of climbing the hill. With
> this approach, we are not completing the secondary task of the problem
> that is landing exactly at the flag.