बर्कले_gnm_recon

  • विवरण :

ऑफ-रोड नेविगेशन

विभाजित करना उदाहरण
'train' 11,834
  • फ़ीचर संरचना :
FeaturesDict({
   
'episode_metadata': FeaturesDict({
       
'file_path': Text(shape=(), dtype=string),
   
}),
   
'steps': Dataset({
       
'action': Tensor(shape=(2,), dtype=float64, description=Robot action, consists of 2x position),
       
'action_angle': Tensor(shape=(3,), dtype=float64, description=Robot action, consists of 2x position, 1x yaw),
       
'discount': Scalar(shape=(), dtype=float64, description=Discount if provided, default to 1.),
       
'is_first': bool,
       
'is_last': bool,
       
'is_terminal': bool,
       
'language_embedding': Tensor(shape=(512,), dtype=float32, description=Kona language embedding. See https://tfhub.dev/google/universal-sentence-encoder-large/5),
       
'language_instruction': Text(shape=(), dtype=string),
       
'observation': FeaturesDict({
           
'image': Image(shape=(120, 160, 3), dtype=uint8, description=Main camera RGB observation.),
           
'position': Tensor(shape=(2,), dtype=float64, description=Robot position),
           
'state': Tensor(shape=(3,), dtype=float64, description=Robot state, consists of [2x position, 1x yaw]),
           
'yaw': Tensor(shape=(1,), dtype=float64, description=Robot yaw),
       
}),
       
'reward': Scalar(shape=(), dtype=float64, description=Reward if provided, 1 on final step for demos.),
   
}),
})
  • फ़ीचर दस्तावेज़ीकरण :
विशेषता कक्षा आकार डीप्रकार विवरण
फीचर्सडिक्ट
एपिसोड_मेटाडेटा फीचर्सडिक्ट
एपिसोड_मेटाडेटा/फ़ाइल_पथ मूलपाठ डोरी मूल डेटा फ़ाइल का पथ.
कदम डेटासेट
कदम/कार्रवाई टेन्सर (2,) फ्लोट64 रोबोट क्रिया, 2x स्थिति से युक्त होती है
चरण/कार्रवाई_कोण टेन्सर (3,) फ्लोट64 रोबोट कार्रवाई में 2x स्थिति, 1x यॉ शामिल है
कदम/छूट अदिश फ्लोट64 यदि छूट प्रदान की गई है, तो डिफ़ॉल्ट 1 है।
चरण/पहला है टेन्सर बूल
चरण/अंतिम है टेन्सर बूल
चरण/is_terminal टेन्सर बूल
चरण/भाषा_एम्बेडिंग टेन्सर (512,) फ्लोट32 कोना भाषा एम्बेडिंग. https://tfhub.dev/google/universal-sentence-encoder-large/5 देखें
चरण/भाषा_निर्देश मूलपाठ डोरी भाषा निर्देश.
चरण/अवलोकन फीचर्सडिक्ट
चरण/अवलोकन/छवि छवि (120, 160, 3) uint8 मुख्य कैमरा आरजीबी अवलोकन।
चरण/अवलोकन/स्थिति टेन्सर (2,) फ्लोट64 रोबोट की स्थिति
चरण/अवलोकन/स्थिति टेन्सर (3,) फ्लोट64 रोबोट अवस्था में [2x स्थिति, 1x यॉ] शामिल है
कदम/अवलोकन/याव टेन्सर (1,) फ्लोट64 रोबोट यॉ
कदम/इनाम अदिश फ्लोट64 यदि प्रदान किया गया तो इनाम, डेमो के लिए अंतिम चरण पर 1।
  • उद्धरण :
@inproceedings{
shah2021rapid
,
title
={ {Rapid Exploration for Open-World Navigation with Latent Goal Models} },
author
={Dhruv Shah and Benjamin Eysenbach and Nicholas Rhinehart and Sergey Levine},
booktitle
={5th Annual Conference on Robot Learning },
year
={2021},
url
={https://openreview.net/forum?id=d_SWJhyKfVw}
}