mirror of
https://github.com/wassname/pytorch-soft-actor-critic.git
synced 2026-08-16 11:26:43 +08:00
44 lines
1.2 KiB
Markdown
44 lines
1.2 KiB
Markdown
### Description
|
|
------------
|
|
Reimplementation of [Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement
|
|
Learning with a Stochastic Actor](https://arxiv.org/pdf/1801.01290.pdf).
|
|
|
|
Contributions are welcome. If you find any mistake (very likely) or know how to make it more stable, don't hesitate to send a pull request.
|
|
|
|
### Requirements
|
|
------------
|
|
|
|
- [mujoco-py](https://github.com/openai/mujoco-py)
|
|
- [Plotly](https://plot.ly/)
|
|
- [PyTorch](http://pytorch.org/)
|
|
|
|
### Run
|
|
------------
|
|
Use the default hyperparameters.
|
|
|
|
#### For SAC :
|
|
|
|
```
|
|
python main.py --env-name Humanoid-v2 --scale_R 20
|
|
```
|
|
|
|
#### For SAC (Hard Update):
|
|
|
|
```
|
|
python main.py --env-name Humanoid-v2 --scale_R 20 --tau 1 --value_update 1000
|
|
```
|
|
|
|
#### For SAC (Deterministic, Hard Update):
|
|
|
|
```
|
|
python main.py --env-name Humanoid-v2 --scale_R 20 --deterministic True --tau 1 --value_update 1000
|
|
```
|
|
|
|
### Results
|
|
------------
|
|
My results on Humanoid-v2 environment using SAC, SAC(hard update) and SAC(deterministic, hard update).
|
|
This is a plot of average rewards at every 10000 step interval
|
|
|
|
|
|
|