Train RL agents

With the local installation

RL experiments can be launched with the script run_experiments.py.

Usage

Flag	Description	Type	Example
--exp-id	Unique experiment ID	int	99
--algo	RL algorithm	str	a2c, ddpg, her, ppo, sac, td3
--env	Training environment ID	str	widowx_reacher-v1
--n-timesteps	Number of training timesteps	int	103 to 1012
--n-seeds	Number of runs with different initialisation seeds	int	2 to 10

Example:

python run_experiments.py --exp-id 99 --algo ppo --env widowx_reacher-v1 --n-timesteps 10000 --n-seeds 3

A Shell script that launches multiple experiments is provided for convenience:

./run_all_exp.sh

With Docker

RL experiments can be launched from the Docker container (CPU or GPU version). The log files will generated in the working directory of the host machine.

# CPU
docker run -it --rm --network host --ipc=host --mount src=$(pwd),target=/root/rl_reach/,type=bind rlreach/rlreach-cpu:latest bash -c "python run_experiments.py --exp-id 99 --algo ppo --env widowx_reacher-v1 --n-timesteps 10000 --n-seeds 2"
# GPU
docker run -it --rm --runtime=nvidia --network host --ipc=host --mount src=$(pwd),target=/root/rl_reach/,type=bind rlreach/rlreach-gpu:latest bash -c "python run_experiments.py --exp-id 99 --algo ppo --env widowx_reacher-v1 --n-timesteps 10000 --n-seeds 2"

A Shell script is provided for ease of usability.

# CPU
./docker/run_docker_cpu.sh python run_experiments.py --exp-id 99 --algo ppo --env widowx_reacher-v1 --n-timesteps 10000 --n-seeds 2
# GPU
./docker/run_docker_gpu.sh python run_experiments.py --exp-id 99 --algo ppo --env widowx_reacher-v1 --n-timesteps 10000 --n-seeds 2