qgallouedec/prj_gia_dataset_metaworld_plate_slide_back_v2_1111

该数据集是一个用于模仿学习的环境，具体为plate-slide-back-v2环境。它是作为Generally Intelligent Agents项目的一部分创建的，旨在支持深度强化学习、强化学习、多任务、多模态、模仿学习和离线强化学习等任务。

This dataset is an environment for imitation learning, specifically the plate-slide-back-v2 environment. It was developed as part of the Generally Intelligent Agents project, and is designed to support tasks including deep reinforcement learning, reinforcement learning, multi-task learning, multi-modal learning, imitation learning, and offline reinforcement learning.

原始信息汇总

数据集概述

数据集名称

• 名称: prj_gia_dataset_metaworld_plate_slide_back_v2_1111

数据集标签

• 标签:
  • deep-reinforcement-learning
  • reinforcement-learning
  • gia
  • multi-task
  • multi-modal
  • imitation-learning
  • offline-reinforcement-learning

数据集内容

• 环境: 模仿学习环境，针对plate-slide-back-v2环境
• 数据结构: 包含observations, actions, dones, rewards等键值

数据集加载

• 加载命令: python import numpy as np dataset = np.load("prj_gia_dataset_metaworld_plate_slide_back_v2_1111/dataset.npy", allow_pickle=True).item() print(dataset.keys()) # dict_keys([observations, actions, dones, rewards])