Q Learning Agent Playing FrozenLake-v1-4x4-no_slippery

This is a trained model of a Q Learning agent playing FrozenLake.

Using Google Colab, I trained a Q Learning agent that learns to go from the starting state (S) to the goal state (G) by walking only on frozen tiles (F) and avoiding holes (H).

Visual of agent

replay.mp4

Information about the model

Environment: FrozenLake-v1-4x4-no_slippery Model: Q Learning Mean Reward +/- Std. Dev.: 1.00 +/- 0.00