search

PaulDanielML / MuJoCo_RL_UR5

A MuJoCo/Gym environment for robot control using Reinforcement Learning. The task of agents in this environment is pixel-wise prediction of grasp success chances.
MIT License
413 stars 54 forks source link

Segmentation fault (core dumped) #3

Closed ChenyangRan closed 3 years ago

ChenyangRan commented 3 years ago

Hi, when I test your code, a segmentation fault happended. Could you help me? Screenshot from 2020-10-23 22-13-32

PaulDanielML commented 3 years ago

Hi, what version of Python are you using? Also the command line argument you use is new, did you modify the code to support it?

ChenyangRan commented 3 years ago

Hi, what version of Python are you using? Also the command line argument you use is new, did you modify the code to support it?

Hi, my python is 3,7 and the torch is 1.6.0. I have tested the example.py and example_agent.py, which work well. I haven't modified the code. Could you please tell me where I should modify?

ChenyangRan commented 3 years ago

Hi, I have sloved the problem. The segmentation is from the sunmmarywriter. And I just change the order and it can work.

tjma005 commented 3 years ago

Hi, I met the same problem when running the example_agent.py, my python is 3.6. When I change the "render=True" to the "render=False", the problem gone. Is that the problem of the version of mujoco?