|
|
--- |
|
|
license: mit |
|
|
tags: |
|
|
- reinforcement-learning |
|
|
- deep-rl |
|
|
- ppo |
|
|
- cartpole |
|
|
library_name: torch |
|
|
framework: pytorch |
|
|
model-index: |
|
|
- name: PPO-CartPole |
|
|
results: [] |
|
|
--- |
|
|
|
|
|
# 🧠 PPO-CartPole Agent |
|
|
|
|
|
This is a PPO (Proximal Policy Optimization) agent trained to solve the `CartPole-v1` environment using PyTorch. |
|
|
|
|
|
## 🛠️ Model Details |
|
|
|
|
|
- **Algorithm**: PPO (Proximal Policy Optimization) |
|
|
- **Environment**: CartPole-v1 |
|
|
- **Framework**: PyTorch |
|
|
- **Observation Space**: Continuous (4-dim) |
|
|
- **Action Space**: Discrete (2 actions) |
|
|
- **Training Episodes**: 1000 |
|
|
- **Max Steps per Episode**: 500 |
|
|
|
|
|
## 🚀 Usage |
|
|
|
|
|
You can load the model using PyTorch: |
|
|
|
|
|
```python |
|
|
import torch |
|
|
from your_model_file import PolicyNetwork # replace with your actual class name |
|
|
|
|
|
model = PolicyNetwork() |
|
|
model.load_state_dict(torch.load("ppo_cartpole.pt")) |
|
|
model.eval() |
|
|
|
|
|
|
|
|
# PPO CartPole Agent 🏋️ |
|
|
|
|
|
This repository contains a PPO agent trained to solve the CartPole-v1 environment using PyTorch and Gymnasium. |
|
|
|
|
|
## 🎥 Episode Demo |
|
|
|
|
|
<video controls width="600"> |
|
|
<source src="ppo-episode-0.mp4" type="video/mp4"> |
|
|
Your browser does not support the video tag. |
|
|
</video> |
