Vladprime/ppo-SnowballTarget

SnowballTarget

Trained from random initialization for Vlad Kolomiiets (Vladprime) using the official Hugging Face Deep RL Course Unity environment and Unity ML-Agents 1.1.0. No pretrained weights were used. This is an initial training run; it is not claimed to have converged. Units 5 and 7 require publishing a trained model without a reward threshold.

Training: ppo, 30000 configured environment steps, seed 20261101. Export: genuine ML-Agents ONNX plus PyTorch checkpoint and trainer configuration. Separate ONNX evaluation process: seed 1000000; 44 completed agent episodes, mean 6.6364, standard deviation 2.0790. Original environment, action and observation spaces were retained. See results.json for individual returns and evaluation limitations.

The .onnx file is compatible with the course Unity browser viewer. Training/evaluation code is in train-unity.py; the exact run configuration is config.yaml.