Overview

Setup

We assume that you have access to a GPU with CUDA >=9.2 support. All dependencies can then be installed with the following commands:

conda env create -f setup/conda.yml
conda activate dmcgb
sh setup/install_envs.sh

If you don't have the right mujoco version installed:

sh setup/install_mujoco_deps.sh
sh setup/prepare_dm_control_xp.sh

Datasets

wget http://data.csail.mit.edu/places/places365/places365standard_easyformat.tar

After downloading and extracting the data, add your dataset directory to the datasets list in setup/config.cfg.

Run SVEA training:

python src/train.py --algorithm svea --seed [SEED] --domain_name [DOMAIN] --task_name [TASK];

Run SDA training:

python src/train.py --algorithm sda --seed [SEED] --sda_quantile [SDA_QUANTILE] --domain_name [DOMAIN] --task_name [TASK];

DMControl Generalization Benchmark

Benchmark for generalization in continuous control from pixels, based on DMControl.

Test environments

The DMControl Generalization Benchmark provides two distinct benchmarks for visual generalization, random colors and video backgrounds:

Both benchmarks are offered in easy and hard variants. Samples are shown below.

video_easy

video_hard