This is a trained model of a Reinforce agent playing Pixelcopter-PLE-v0 .
This model didn't complete training but was still far better than the first one trained with fewer timesteps.
-
