ด๊อบบี้

  • คำอธิบาย :

  • หน้าแรก : https://github.com/notmahi/dobb-e

  • ซอร์สโค้ด : tfds.robotics.rtx.Dobbe

  • รุ่น :

    • 0.1.0 (ค่าเริ่มต้น): การเปิดตัวครั้งแรก
  • ขนาดการดาวน์โหลด : Unknown size

  • ขนาดชุดข้อมูล : 21.21 GiB

  • แคชอัตโนมัติ ( เอกสาร ): No

  • แยก :

แยก ตัวอย่าง
'train' 5,208
  • โครงสร้างคุณสมบัติ :
FeaturesDict({
    'episode_metadata': FeaturesDict({
        'file_path': string,
    }),
    'steps': Dataset({
        'action': Tensor(shape=(7,), dtype=float32),
        'discount': Scalar(shape=(), dtype=float32),
        'is_first': bool,
        'is_last': bool,
        'is_terminal': bool,
        'language_embedding': Tensor(shape=(512,), dtype=float32),
        'language_instruction': string,
        'observation': FeaturesDict({
            'gripper': Tensor(shape=(1,), dtype=float32),
            'quat': Tensor(shape=(4,), dtype=float32),
            'rot': Tensor(shape=(3,), dtype=float32),
            'state': Tensor(shape=(7,), dtype=float32),
            'wrist_image': Image(shape=(256, 256, 3), dtype=uint8),
            'xyz': Tensor(shape=(3,), dtype=float32),
        }),
        'reward': Scalar(shape=(), dtype=float32),
    }),
})
  • เอกสารคุณสมบัติ :
คุณสมบัติ ระดับ รูปร่าง ประเภทD คำอธิบาย
คุณสมบัติDict
ตอนที่_ข้อมูลเมตา คุณสมบัติDict
ตอนที่_metadata/file_path เทนเซอร์ เชือก
ขั้นตอน ชุดข้อมูล
ขั้นตอน/การดำเนินการ เทนเซอร์ (7,) ลอย32
ขั้นตอน/ส่วนลด สเกลาร์ ลอย32
ขั้นตอน/is_first เทนเซอร์ บูล
ขั้นตอน/is_last เทนเซอร์ บูล
ขั้นตอน/is_terminal เทนเซอร์ บูล
ขั้นตอน/ภาษา_embedding เทนเซอร์ (512,) ลอย32
ขั้นตอน/Language_instruction เทนเซอร์ เชือก
ขั้นตอน/การสังเกต คุณสมบัติDict
ขั้นตอน/การสังเกต/อุปกรณ์จับยึด เทนเซอร์ (1,) ลอย32
ขั้นตอน/การสังเกต/ควอต เทนเซอร์ (4,) ลอย32
ขั้นตอน/การสังเกต/การเน่าเปื่อย เทนเซอร์ (3,) ลอย32
ขั้นตอน/การสังเกต/สถานะ เทนเซอร์ (7,) ลอย32
ขั้นตอน/การสังเกต/wrist_image ภาพ (256, 256, 3) uint8
ขั้นตอน/การสังเกต/xyz เทนเซอร์ (3,) ลอย32
ขั้นตอน/รางวัล สเกลาร์ ลอย32
  • การอ้างอิง :
@misc{shafiullah2023dobbe, title={On Bringing Robots Home}, author={Nur Muhammad Mahi Shafiullah and Anant Rai and Haritheja Etukuru and Yiqian Liu and Ishan Misra and Soumith Chintala and Lerrel Pinto}, year={2023}, eprint={2311.16098}, archivePrefix={arXiv}, primaryClass={cs.RO} }