no996

> Hi, I got this problem too. For solving it, I just set the upper and lower bound of the box to (1000,-1000) instead of (0,1), and then the simulation...

finally I find how to check the file under flow/flow/envs/merge.py or some others. There is a function named @property def observation_space(self): """See class definition.""" return Box(low=-1000, high=1000, shape=(5 * self.num_rl,...

I used another one here https://github.com/LeeXujie/ArUco/blob/master/sources/spm-slam_1.0.1/spm-slam.cpp it also have the same problems. But I find a solution to solve it. If you change all the argv inputs, and then run...

I tried to load my own map, which is built by road runner and it also gives me the same error like "Malloc Size=131160 LargeMemoryPoolOffset=327928." And I find the reason....

I have found that the huber_loss with GradientDescentOptimizer cause the memory leak, and when I changed to reduce mean with RMSPropOptimizer it disappears. Can you explain why you use the...

Thank you so much. I would try that. With Nvidia-smi I have saw that carla will use 4GB and the network should use more than 4 GB