Hello @jakaskerl,
I am current using an OAK4D PRO camera and deploying a custom yolo8n model on the camera, I directly converted the model(a .pt file) from the luxonis hub online and downloaded it and used it in my program which is same as the final depthai script in this repo https://github.com/luxonis/depthai-ml-training/blob/main/training/train_detection_model.ipynb the only difference being in the classes and the model_path

this is how i specified the model path MODEL_PATH = "model.rvc4.tar.xz"

When I run this program I get this error

[1443462065] [192.168.0.152] [1749754458.499] [NeuralNetwork(2)] [error] Couldn't build the model.

[2025-06-12 13:54:19.395] [depthai] [error] Communication exception - possible device error/misconfiguration. Original message 'Couldn't read data from stream: '__x_4_out' (X_LINK_ERROR)'

[2025-06-12 13:54:19.395] [depthai] [error] Communication exception - possible device error/misconfiguration. Original message 'Couldn't read data from stream: '__x_2_passthrough' (X_LINK_ERROR)'

[2025-06-12 13:54:22.563] [warning] FW crashed - trying to get out the crash dump

[2025-06-12 13:54:25.563] [warning] Getting the crash dump out - this can take up to a minute, because it first needs to be compressed.

[2025-06-12 13:54:28.666] [depthai] [error] Device with id 1443462065 has crashed. Crash dump logs are stored in: /home/depthai-core/examples/python/DetectionNetwork/.cache/depthai/crashdumps/f60687e9525bd2518855be48e378bf7a70622e2e/core_dump_0.0.1+c21e47f41f34e3d818fb11cfb44726c5edaf2695_20250612_135425.tar.gz - please report to developers.

I believe we are already addressing the issues you are having here