A dataset API for Offline Reinforcement Learning.

Minari is a Python API that hosts a collection of popular Offline Reinforcement Learning datasets. The environments from which these datasets are generated follow the Gymnasium API. The datasets are publicly available in a Farama GCP bucket and can be downloaded through the Minari CLI. Minari also provides dataset handling features such as episode sampling, filtering trajectories, as well as dataset generation utilities.

// Install Minari
$ pip install "minari[all]"
---> 100%

// Show remote datasets.
$ minari list remote

                  <i>Minari datasets in Farama server</i>
┌─────────────────────┬───────────┬────────────┬───────────┬─────────┐
│                     │     <b>Total</b> │      <b>Total</b> │   <b>Dataset</b> │         │
│ <b>Name</b>                │  <b>Episodes</b> │      <b>Steps</b> │      <b>Size</b> │  <b>Author</b> │
┡─────────────────────╇───────────╇────────────╇───────────╇─────────┩
│ <font color="#A1EFE4">D4RL/door/cloned-v2</font> │      <font color="#03AC13">4356</font> │    <font color="#03AC13">1000000</font> │ <font color="#03AC13">1077.7 MB</font> │  <font color="#FF00FF">Farama</font> │
│ <font color="#A1EFE4">D4RL/door/expert-v2</font> │      <font color="#03AC13">5000</font> │    <font color="#03AC13">1000000</font> │ <font color="#03AC13">1096.4 MB</font> │  <font color="#FF00FF">Farama</font> │
│ <font color="#A1EFE4">D4RL/door/human-v2</font>  │        <font color="#03AC13">25</font> │       <font color="#03AC13">6729</font> │ <font color="#03AC13">7.1 MB</font>    │  <font color="#FF00FF">Farama</font> │
└─────────────────────┴───────────┴────────────┴───────────┴─────────┘

// Download dataset D4RL/door/cloned-v2
$ minari download D4RL/door/cloned-v2
Downloading D4RL/door/cloned-v2 from Farama servers...

   * Downloading data file 'D4RL/door/cloned-v2/data/main_data.hdf5' ...

---> 100%

Dataset D4RL/door/cloned-v2 downloaded to ~/.minari/datasets/D4RL/door/cloned-v2

"見習い - learning by observation"