• DepthAI

  • Log: 'softMaxNClasses' '157' 'CMX memory is not enough!'.

Hi, I have trained a YoloV8 model and converted its weights .pt file to blob using tools.luxonis.com. The converted blob file when I use in pipeline of YoloNetwork. It is thowing below error.

[184430101107FB0F00] [192.168.1.102] [5.721] [DetectionNetwork(1)] [critical] Fatal error in openvino 'universal'. Likely because the model was compiled for different openvino version. If you want to select an explicit openvino version use: setOpenVINOVersion while creating pipeline. If error persists please report to developers. Log: 'softMaxNClasses' '157' 'CMX memory is not enough!'.

[184430101107FB0F00] [192.168.1.102] [10.516] [system] [critical] Fatal error. Please report to developers. Log: 'Fatal error on MSS CPU: trap: 00, address: 00000000' '0'

I tried to change the version using

pipeline.setOpenVINOVersion to ('VERSION_2021_4', 'VERSION_2022_1', 'VERSION_UNIVERSAL') . But with all these versions its throwing same above error. How to resolve this?

    Can you let me know which V8 model you are using and how many classes do you have

    NileshBawiskar

      I have trained yolov8n model with image size 960x960 and that is sized around 32 MB -.pt file and blob is 25MB. I had trained yolov8s before which was sized 125MB (.pt file) . I thaught model size is heavy so I trained yolov8n which is of very small size, still the same error.
      Classes are 1 only.