SeaWave ERROR: nefc under-allocation: found nefc=16 but allocated only 10

[dsdfasfaac]

When I start inferencing on the simulator, the simulator sometimes throws errors “ERROR: nefc under-allocation: found nefc=16 but allocated only 10”, causing my LLAVA reasoning and main program to get stuck, I really don't know how to fix it, Have you encountered this kind of problem?
OS: Ubuntu22.04
Python Version 3.9
Package Version cuda11.8 torch2.1.2
GPU: RTX3090-24GB

[abliao]

You can get more detailed error information by viewing the log files under HARIX_RDKSim/Saved/Logs.

[dsdfasfaac]

This document doesn't seem to offer any more information. This is the end of the file when the program crashes.
''[2025.09.09-15.27.34:781][888]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.35:591][935]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.36:433][983]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.37:241][ 29]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.38:062][ 76]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.38:670][110]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.39:491][157]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.40:317][204]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.41:143][251]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.41:779][285]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.46:813][584]LogTemp: Warning: RecreateMujocoScene
[2025.09.09-15.27.46:822][584]LogTemp: Warning: Start CreateMujocoScene
[2025.09.09-15.27.48:726][584]LogTemp: Warning: End CreateMujocoScene
[2025.09.09-15.27.49:262][618]LogTemp: Display: Action 0 Success.
[2025.09.09-15.27.49:314][621]LogTemp: Display: Action 0 Success.''

[abliao]

You can check whether the previous simulation was completely closed. For example:

ps -ef |grep HARIX

[dsdfasfaac]

Yes, the previous simulation has stopped.

[abliao]

You can try using another GPU by changing the -graphicsadapter.