Chrome dino game machine learning

The goal of this project is to make the environment of the famous chrome dino game and build / train a machine learning algorithm to beat it.
This game is very simple, since your goal is to avoid obstacles by jumping over them or crouching on the floor, the score is calculated using. The version of the game I will develop will only have the jump option

• Python3
• Stablebaselines3
• Tensorboard
• Pygames
• Gym environments

[email protected]:thomas-mauran/DinoAI.git

apt install python3

apt install pip

pip install gym

pip install stable-baselines3[extra]

pip install pygame

the env is stocked and made in the env folder. You can train a model in the train.py file using the ml algorithm you want or change the values of the current one. Then after the training you can set the model path var to the model you just made and run the file to see how it performs

⚠️ make sure you change the model name in the file (save_path variable)

python3 train.py

⚠️ make sure you load the good model in the model_path variable

python3 test.py