You can deploy this model and start earning money today!

artnfull/
connect-ai-artnfull-mujoco-pusher-ppo

Name: connect-ai-artnfull-mujoco-pusher-ppo
Availability: InStock

stable-baselines3

reinforcement-learning

deep-reinforcement-learning

Views

No views yet

🦾 MuJoCo Pusher-v5 PPO Physical AI Robot Model

This repository contains a trained PPO (Proximal Policy Optimization) model for the Gymnasium MuJoCo Pusher-v5 continuous robot arm control task.

🎬 Robot Performance Demonstration

📊 Training Results

Task: Reinforcement Learning (Continuous Control)
Algorithm: PPO (MlpPolicy)
Environment: Pusher-v5 (7-DOF MuJoCo Robot Arm)
Total Timesteps: 262,144 steps
Framework: Stable-Baselines3 & Gymnasium
Trained Behavior: The robot arm accurately aligns its U-shaped pusher tip behind the cylinder and pushes it directly into the target goal.

🚀 How to Use

1. Install Dependencies

pip install gymnasium[mujoco] stable-baselines3[extra]

2. Load Model & Evaluate

python

import gymnasium as gym
from stable_baselines3 import PPO

# Load environment
env = gym.make("Pusher-v5", render_mode="human")

# Load model from local zip
model = PPO.load("ppo_pusher.zip")

# Evaluate
obs, _ = env.reset()
for _ in range(200):
    action, _ = model.predict(obs, deterministic=True)
    obs, reward, terminated, truncated, _ = env.step(action)
    if terminated or truncated:
        obs, _ = env.reset()
env.close()

artnfull/connect-ai-artnfull-mujoco-pusher-ppo