Process Stopped

[Savior1s]

Thanks for your great work here, and some problems here need your help then .

As I set the mesh frequency to 40 and I put the command in terminal, python -w run.py configs/...... std, and my process will stop after 7 times mesh file saved, how could I solve that?

[Savior1s]

or if I change the mesh freq to anything else , if will stop and never finish to 2000 once.

[zstsandy]

For visualization purpose, mesh frequency should be set to 40. If you only want to run for evaluation, keep mesh frequency as the original code.

[Savior1s]

The problem now is that when I run this program, no matter what the mesh freq is, the program will get stuck and cannot complete the run. py, and I don't know how to solve this problem. I am monitoring whether my RTX is running. When the progress bar gets stuck, the process disappears and never continue

[Savior1s]

like the pic uploaded, it stopped at 550/2000

[Savior1s]

I ignore the warnings before and I just find that it send the warning information:/home/lbw/anaconda3/envs/sni/lib/python3.7/multiprocessing/semaphore_tracker.py:144: UserWarning: semaphore_tracker: There appear to be 11 leaked semaphores to clean up at shutdown
len(cache)) and it stopped

[hangar01]

Sorry, I have the same problem. I run the dataset room0 and room1 respectively. They are stuck at different 40FPS frequencies such as 562 and 762. The problem is: the video memory usage decreases and the process is stuck.
My device is a GPU with 16G video memory, and I have never reached the video memory limit.
If I don't manually terminate the process, it will be stuck all the time. (I have waited for 4 or 5 hours).
What is the problem with this?

[Savior1s]

Sorry, I have the same problem. I run the dataset room0 and room1 respectively. They are stuck at different 40FPS frequencies such as 562 and 762. The problem is: the video memory usage decreases and the process is stuck. My device is a GPU with 16G video memory, and I have never reached the video memory limit. If I don't manually terminate the process, it will be stuck all the time. (I have waited for 4 or 5 hours). What is the problem with this?

how much is your ram

[hangar01]

Sorry, I have the same problem. I run the dataset room0 and room1 respectively. They are stuck at different 40FPS frequencies such as 562 and 762. The problem is: the video memory usage decreases and the process is stuck. My device is a GPU with 16G video memory, and I have never reached the video memory limit. If I don't manually terminate the process, it will be stuck all the time. (I have waited for 4 or 5 hours). What is the problem with this?

how much is your ram

RAM? My RAM is 32G (16*2)

[Savior1s]

Sorry, I have the same problem. I run the dataset room0 and room1 respectively. They are stuck at different 40FPS frequencies such as 562 and 762. The problem is: the video memory usage decreases and the process is stuck. My device is a GPU with 16G video memory, and I have never reached the video memory limit. If I don't manually terminate the process, it will be stuck all the time. (I have waited for 4 or 5 hours). What is the problem with this?

how much is your ram

RAM? My RAM is 32G (16*2)
I wondering if it's my inefficient ram but may be not, I didn't find the cause now.

[hangar01]

Sorry, I have the same problem. I run the dataset room0 and room1 respectively. They are stuck at different 40FPS frequencies such as 562 and 762. The problem is: the video memory usage decreases and the process is stuck. My device is a GPU with 16G video memory, and I have never reached the video memory limit. If I don't manually terminate the process, it will be stuck all the time. (I have waited for 4 or 5 hours). What is the problem with this?

how much is your ram

RAM? My RAM is 32G (16*2)
I wondering if it's my inefficient ram but may be not, I didn't find the cause now.

Maybe you are right.
I monitored my RAM and other hardware and found that it was indeed the 32GB RAM that was not enough that caused the project to stop running.
Can the author provide the RAM requirements for this project?

[jyue86]

I also have the same problem. My GPU has 24 gb of ram

[hangar01]

For visualization purpose, mesh frequency should be set to 40. If you only want to run for evaluation, keep mesh frequency as the original code.

For visualization purposes, when mesh_freq is set to 40. Can you provide the required running memory requirements?
My GPU's 16GB graphics memory is fine, but the 32GB RAM（running memory） has exploded.

[AIerOK]

Hello, what is the configuration of your computer

[zstsandy]

My computer has 64GB of RAM and an RTX 4090 GPU, with the code occupying about 10GB of the GPU memory. In SNI-SLAM, the keyframes are stored on the CPU, which results in high CPU usage. We will open source a updated code for limited GPU and CPU space.