README.md

Run

OpenAI Gym MuJoCo tasks

python train.py --env_name=HalfCheetah-v2 --save_dir=./tmp/

Experiment tracking with Weights and Biases

python train.py --env_name=HalfCheetah-v2 --save_dir=./tmp/ --track

DeepMind Control suite (--env-name=domain-task)

python train.py --env_name=cheetah-run --save_dir=./tmp/

For continuous control from pixels

MUJOCO_GL=egl python train_pixels.py --env_name=cheetah-run --save_dir=./tmp/

For offline RL

python train_offline.py --env_name=halfcheetah-expert-v0  --dataset_name=d4rl --save_dir=./tmp/

For RL finetuning

python train_finetuning.py --env_name=HalfCheetah-v2 --dataset_name=awac --save_dir=./tmp/

For sample efficient RL

python train.py --env_name=Hopper-v2 --start_training=5000 --max_steps=300000 --updates_per_step=20 --config=configs/redq_default.py --save_dir=./tmp/