search

andyzeng / visual-pushing-grasping

Train robotic agents to learn to plan pushing and grasping actions for manipulation with deep reinforcement learning.
http://vpg.cs.princeton.edu/
BSD 2-Clause "Simplified" License
898 stars 315 forks source link

Not enough objects in view (value: 0)! Repositioning objects #9

Open yang08521 opened 5 years ago

yang08521 commented 5 years ago

Dear Andy Zeng， I am a newcomer to the field of reinforcement learning.I saw your study on github - Learning Synergies between Pushing and Grasping with Self-supervised Deep Reinforcement Learning, I am very interested in it, so I want to learn and try to reproduce your study. During the learning process, I encountered some problems and would like to ask you.When I tried to reproduce with vrep simulation, it produced some different pieces of wood, but it seems that these objects cannot be recognized, as shown by: Not enough objects in view (value: 0)! Repositioning objects.How can I solve this problem? I would be grateful if you could take the time to guide me! Sincerely, Yang

LinQing2017 commented 5 years ago

I have the same question. I use support-pytorch-v0.4 branch to run demo in simulation. Here is my version information: os: ubuntu 16.04 Cuda compilation tools, release 10.0, V10.0.130 Python 3.6.2 V-REP 3.6.2 rev0 versions PyTorch 1.2

Focusssss commented 4 years ago

@yang08521 @LinQing2017 I have the same question. Is this error related to the version? The question has troubled me for many days. Please tell me some methods to solve the question! Thanks!!!