Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/Jingliang-Duan/DSAC-v1
DSAC; Distributional Soft Actor-Critic
https://github.com/Jingliang-Duan/DSAC-v1
pytorch reinforcement-learning
Last synced: 24 days ago
JSON representation
DSAC; Distributional Soft Actor-Critic
- Host: GitHub
- URL: https://github.com/Jingliang-Duan/DSAC-v1
- Owner: Jingliang-Duan
- Created: 2020-03-03T03:57:34.000Z (almost 5 years ago)
- Default Branch: master
- Last Pushed: 2024-02-22T16:52:44.000Z (10 months ago)
- Last Synced: 2024-10-11T02:02:32.777Z (2 months ago)
- Topics: pytorch, reinforcement-learning
- Language: Python
- Homepage:
- Size: 9.36 MB
- Stars: 109
- Watchers: 7
- Forks: 31
- Open Issues: 3
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
- StarryDivineSky - Jingliang-Duan/Distributional-Soft-Actor-Critic
README
## Reference
- [Distributional Soft Actor-Critic (DSAC)](https://ieeexplore.ieee.org/document/9448360)
## Requires
1. Windows 7 or greater or Linux.
2. Python 3.8.
3. The installation path must be in English.
## Installation
```bash
# Please make sure not to include Chinese characters in the installation path, as it may result in a failed execution.
# clone DSAC_v1 repository
git clone [email protected]/Jingliang-Duan/DSAC_v1
cd DSAC_v1
# create conda environment
conda env create -f DSAC1.0_environment.yml
conda activate DSAC1.0
# install DSAC1.0
pip install -e.
```
## Train
These are two examples of running DSAC-v1 on two environments.
Train the policy by running:
```bash
cd example_train
#Train a pendulum task
python main.py
#Train a humanoid task. To execute this file, Mujoco and Mujoco-py need to be installed first.
python dsac_mlp_humanoidconti_offserial.py
```
After training, the results will be stored in the "DSAC_v1/results" folder.
## Simulation
In the "DSAC-v1/results" folder, pick the path to the folder where the policy will be applied to the simulation and select the appropriate PKL file for the simulation.
```bash
python run_policy.py
#you may need to "pip install imageio-ffmpeg" before running this file on Windows.
```
After running, the simulation vedio and state&action curve figures will be stored in the "DSAC_v1/figures" folder.
## Acknowledgment
We would like to thank all members in Intelligent Driving Laboratory (iDLab), School of Vehicle and Mobility, Tsinghua University for making excellent contributions and providing helpful advices for DSAC-v1.