Using the "Advantage Actor Critic(A2C)" Reinforcement Learning method, the A.I is trained to play Atari's Breakout.

The game used for this Model is Atari's Breakout

The game was used by downloading ROM files.

AtariMania 'ROMS'

The A2C algortihm has been implemented for the Agent to play the game

refer the paper:

Asynchronous Methods for Deep Reinforcement Learning

The Python Libraries used:

StableBaselines-3

OpenAI-Gym

• Mark I (100k TimeSteps):

The first Model is a Preliminary Model which has been trained with 100k TimeSteps.

• Mark II (2M TimeSteps):

The Second model is an improved model that has been trained for 2 Million TimeSteps. The Performance was significantly better

These show as to how the Model kept getting better and better, where the Agent was able to score higher.

The Agent was able to score way better with the 2M TimeSteps trained Model while Testing.

If the Model were to be trained for longer TimeSteps, it would perform more effectively.

Have Fun

Yours Truely,

Anuraag Rath :P