Installation
pip install PettingZoo
pip install ray[tune,rllib]
pip install dm-tree
pip install lz4
pip install tensorflow-gpu
pip install gputil
pip3 install torch torchvision torchaudio
pip install -e pz_vehicle
pip install -e gym_vehicle
pip install stable-baseline3
pip uninstall gym
pip install gym
pip3 install torch==1.10.2+cu113 torchvision==0.11.3+cu113 torchaudio==0.10.2+cu113 -f https://download.pytorch.org/whl/cu113/torch_stable.html
pip install stable-baselines3
git clone https://github.com/lcy0x1/188B-MultiAgent.git
cd 188B-MultiAgent
pip install -e gym_vehicle
pip install -e gym_symmetric
git pull
python3 training/cls.py vehicle-v0 128 2 100 100 100 1
python3 training/cls.py vehicle-v0 128 3 100 100 100 1
python3 training/cls.py vehicle-v0 256 3 100 100 100 1
python3 training/cls.py vehicle-v0 256 4 100 100 100 1
python3 training/cls.py symmetric-v0 128 2 100 100 100 8
python3 training/cls.py symmetric-v0 128 3 100 100 100 8
python3 training/cls.py symmetric-v0 256 3 100 100 100 8
python3 training/cls.py symmetric-v0 256 4 100 100 100 8
python3 training/cls.py vehicle-v0 128 2 50 100 100 1
python3 training/cls.py symmetric-v0 128 2 50 100 100 4
python3 training/cls.py symmetric-v0 128 2 30 100 100 16 3
python3 training/cls.py symmetric-v0 128 3 30 100 100 16 3
python3 training/cls.py symmetric-v0 256 2 30 100 100 16 3
python3 training/cls.py symmetric-v0 128 2 30 100 100 16 30
python3 training/cls.py symmetric-v0 128 3 30 100 100 16 30
python3 training/cls.py symmetric-v0 256 2 30 100 100 16 30
python3 training/cls.py symmetric-v0 128 2 30 100 100 16 300
python3 training/cls.py symmetric-v0 128 3 30 100 100 16 300
python3 training/cls.py symmetric-v0 256 2 30 100 100 16 300
python3 training/mix.py symmetric-v0 128 3 30 100 100 16 30 "-lr3/edge_symmetric-v0/30.zip" speedup
python3 training/mix.py symmetric-v0 128 3 30 100 100 16 30 "-lr30/edge_symmetric-v0/30.zip" continue
python3 training/cls.py symmetric-v0 128 2 50 100 100 9 3
python3 training/cls.py symmetric-v0 128 3 50 100 100 9 3
python3 training/cls.py symmetric-v0 256 2 50 100 100 9 3
python3 training/cls.py symmetric-v0 128 2 50 100 100 9 30
python3 training/cls.py symmetric-v0 128 3 50 100 100 9 30
python3 training/cls.py symmetric-v0 256 2 50 100 100 9 30
python3 training/cls.py 128 2 50 100 100 1 3
python3 training/cls.py 128 2 50 100 100 9 3
python3 training/cls.py 128 3 50 100 100 9 3
python3 training/cls.py 256 2 50 100 100 9 3
python3 training/cls.py 128 2 200 100 100 8 10
python3 training/cls.py 128 2 100 100 100 8 20
python3 training/cls.py 128 2 100 100 100 8 30
python3 training/cls.py 128 2 100 100 100 1 10
python3 training/cls.py 128 2 100 100 100 1 20
python3 training/cls.py 128 2 100 100 100 1 30
reward normalization