Edit model card

ppo Agent playing PushBlock

This is a trained model of a ppo agent playing PushBlock using the Unity ML-Agents Library.

Usage (with ML-Agents)

The Documentation: https://github.com/huggingface/ml-agents#get-started We wrote a complete tutorial to learn to train your first agent using ML-Agents and publish it to the Hub:

Resume the training

mlagents-learn <your_configuration_file_path.yaml> --run-id=<run_id> --resume

Watch your Agent play

You can watch your agent playing directly in your browser:.

  1. Go to https://huggingface.co/spaces/unity/ML-Agents-PushBlock
  2. Step 1: Write your model_id: rebolforces/ppo-pushblock-9M
  3. Step 2: Select your .nn /.onnx file
  4. Click on Watch the agent play 👀

Stats

"final_checkpoint": {
  "steps": 9000019,
  "file_path": "results/PushBlock Training 2/PushBlock.onnx",
  "reward": 4.981160132090251,
  "creation_time": 1661076513.4570658,
  "auxillary_file_paths": [
    "results/PushBlock Training 2/PushBlock/PushBlock-9000019.pt"
  ]
}
Downloads last month
22
Video Preview
loading