An imitation learning environment for the handle-pull-v2 environment, sample for the policy handle-pull-v2
This environment was created as part of the Generally Intelligent Agents project gia:
https://github.com/huggingface/gia
Then, load it with
import numpy as np
dataset = np.load("prj_gia_dataset_metaworld_handle_pull_v2_1111/dataset.npy"… See the full description on the dataset page:
https://huggingface.co/datasets/qgallouedec/prj_gia_dataset_metaworld_handle_pull_v2_1111.