mikezhang95/trifinger_data

TriFinger RL Datasets

This repository provides offline reinforcement learning datasets collected on the real TriFinger platform and in a simulated version of the environment. The paper "Benchmarking Offline Reinforcement Learning on Real-Robot Hardware" provides more details on the datasets and benchmarks offline RL algorithms on them. All datasets are available with camera images as well.

More detailed information about the simulated environment, the datasets and on how to run experiments on a cluster of real TriFinger robots can be found in the documentation.

Some of the datasets were used during the Real Robot Challenge 2022.

Installation

cd trifinger_rl_datasets
pip install --upgrade pip  # make sure the most recent version of pip is installed
pip install .
cd ..

Preprocess

python preprocess.py --input_dataset trifinger-cube-lift-real-expert-v0 --output_dataset trifinger-cube-lift-real-expert-v0-masa

Usege in Python

old_env = gym.make(
           args.input_dataset,
           flatten_obs=True)
old_dataset = old_env.get_dataset(rng=(0,2))
print('Observation Shape', old_dataset['observations'][0].shape)

new_env = gym.make(
           args.input_dataset,
           flatten_obs=True,
           data_dir=f'output_datasets/{args.output_dataset}')

new_dataset = new_env.get_dataset(rng=(0,2), clip=False)
print('Observation Shape', new_dataset['observations'][0].shape)