Datasets:

qgallouedec

/

prj_gia_dataset_metaworld_door_close_v2_1111

like 0

An imitation learning environment for the door-close-v2 environment, sample for the policy door-close-v2

This environment was created as part of the Generally Intelligent Agents project gia: https://github.com/huggingface/gia

Load dataset

First, clone it with

git clone https://huggingface.co/datasets/qgallouedec/prj_gia_dataset_metaworld_door_close_v2_1111

Then, load it with

import numpy as np
dataset = np.load("prj_gia_dataset_metaworld_door_close_v2_1111/dataset.npy", allow_pickle=True).item()
print(dataset.keys())  # dict_keys(['observations', 'actions', 'dones', 'rewards'])