Vanilla policy gradients with ValueFunction to estimate value for the specific state (I use current observation, previous observation and previous action as a state). This same algorithm works fine without ValueFunction if you don't stop the learning process at step 200 and continue learning after that. OpenAI Gym's monitor stops the game at step 200 so you can't use monitor at the same time as training on more than 200 steps.
- Use this commit: https://github.com/tilarids/reinforcement_learning_playground/commit/7faeff3ff0c0397f84f391b75e301ac7dea8e69b
- Consider changing the API key :)