Trying to perform inference on my compiled model in beagle bone ai64, using this repo :- https://github.com/WasabiFan/tidl-yolov5-custom-model-demo.

Trying to perform inference on my compiled model in beagle bone ai64, using this repo :- GitHub - WasabiFan/tidl-yolov5-custom-model-demo: A walkthrough and personal notes on running YOLOv5 models with TIDL..
but it shows this error:-
Available execution providers : [‘TIDLExecutionProvider’, ‘TIDLCompilationProvider’, ‘CPUExecutionProvider’]
libtidl_onnxrt_EP loaded 0x13171940
Final number of subgraphs created are : 1, - Offloaded Nodes - 429, Total Nodes - 429
APP: Init … !!!
MEM: Init … !!!
MEM: Initialized DMA HEAP (fd=4) !!!
MEM: Init … Done !!!
IPC: Init … !!!
IPC: Init … Done !!!
REMOTE_SERVICE: Init … !!!
REMOTE_SERVICE: Init … Done !!!
9880.406724 s: GTC Frequency = 200 MHz
APP: Init … Done !!!
9880.407152 s: VX_ZONE_INIT:Enabled
9880.407332 s: VX_ZONE_ERROR:Enabled
9880.407443 s: VX_ZONE_WARNING:Enabled
9880.408198 s: VX_ZONE_INIT:[tivxInitLocal:130] Initialization Done !!!
9880.408533 s: VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
9880.759560 s: VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
9880.759599 s: VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
9880.759612 s: VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
9880.759628 s: VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
9880.759634 s: VX_ZONE_ERROR:[ownNodeKernelInit:540] If the target callbacks have been registered, please ensure no errors are occurring within the create callback of this kernel
9880.759643 s: VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl … failed !!!
9880.759653 s: VX_ZONE_ERROR:[vxVerifyGraph:2055] Node kernel init failed
9880.759659 s: VX_ZONE_ERROR:[vxVerifyGraph:2109] Graph verify failed
TIDL_RT_OVX: ERROR: Verifying TIDL graph … Failed !!!
TIDL_RT_OVX: ERROR: Verify OpenVX graph failed
Input shape: [1, 3, 640, 640]
Input “images”: tensor(float).

any one knows, how to resolve this error.
thanks

What bbai64 os image are you on?
As I understood it has some problems with starting the tidl delegate…
Maybe you can try to ask for a help the author or the repo.
My experience limited by tensorflow v1 ssd models. Maybe I’ll make an article or something to summarize my path trough all this mess…

os details :- Linux BeagleBone-AI64 5.10.168-ti-arm64-r111 #1bullseye SMP Tue Sep 26 14:22:20 UTC 2023 aarch64 GNU/Linux.
please make an article , that will be very useful

Is it one with TI’s stuff?
Debian 11.x (Bullseye) TI EDGEAI (Xfce) Desktop Snapshot from ARM64 - Debian 11.x (Bullseye) - Monthly Snapshots
If so - try to run the setup_script.py in /opt/edge_ai_apps
It should install some TI’s onnx runtime stuff.