A2C playing PongNoFrameskip-v4 from https://github.com/sgoodfriend/rl-algo-impls/tree/0760ef7d52b17f30219a27c18ba52c8895025ae3
1413f39
source benchmarks/train_loop.sh | |
ALGOS="ppo" | |
ENVS="CartPole-v1 MountainCar-v0 MountainCarContinuous-v0 Acrobot-v1 LunarLander-v2" | |
BENCHMARK_MAX_PROCS="${BENCHMARK_MAX_PROCS:-3}" | |
train_loop $ALGOS "$ENVS" | xargs -I CMD -P $BENCHMARK_MAX_PROCS bash -c CMD | |