Other Parts Discussed in Thread: TDA4VM
Dear TI Team,
I hope this message finds you well. I am reaching out regarding an issue I encountered while attempting to run an image classification AI model on the BeagleBone AI-64 (TDA4VM) platform using the TI model composer.
Here is a brief overview of the problem:
Problem Description: I followed the image classification model training process outlined in the video tutorial provided by TI's model composer. After successful model compilation, I downloaded the compiled models and deployed them to my BeagleBone AI-64 (TDA4VM). Subsequently, I added the new model configuration to the "edge_ai_apps/configs/image_classification.yaml" file. However, when attempting to run the "app_edgeai.py" script, it failed to execute, and I received an error message.
FIY: I can run the image classification, and object detection demo using a USB cam.
Platform Details:
- Hardware: BeagleBone AI-64 (TDA4VM)
- SDK Version: 8.2
Error Message:
debian@BeagleBone:/opt/edge_ai_apps/apps_python$ sudo ./app_edgeai.py ../configs/image_classification.yaml
[sudo] password for debian:
libtidl_onnxrt_EP loaded 0x1da82f80
Final number of subgraphs created are : 1, - Offloaded Nodes - 121, Total Nodes - 121
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 !!!
5571.932945 s: GTC Frequency = 200 MHz
APP: Init ... Done !!!
5571.939610 s: VX_ZONE_INIT:Enabled
5571.939652 s: VX_ZONE_ERROR:Enabled
5571.939658 s: VX_ZONE_WARNING:Enabled
5571.940533 s: VX_ZONE_INIT:[tivxInitLocal:130] Initialization Done !!!
5571.940687 s: VX_ZONE_INIT:[tivxHostInitLocal:86] Initialization Done for HOST !!!
5572.034562 s: VX_ZONE_ERROR:[ownContextSendCmd:815] Command ack message returned failure cmd_status: -1
5572.034598 s: VX_ZONE_ERROR:[ownContextSendCmd:851] tivxEventWait() failed.
5572.034613 s: VX_ZONE_ERROR:[ownNodeKernelInit:538] Target kernel, TIVX_CMD_NODE_CREATE failed for node TIDLNode
5572.034624 s: VX_ZONE_ERROR:[ownNodeKernelInit:539] Please be sure the target callbacks have been registered for this core
5572.034635 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
5572.034647 s: VX_ZONE_ERROR:[ownGraphNodeKernelInit:583] kernel init for node 0, kernel com.ti.tidl ... failed !!!
5572.034664 s: VX_ZONE_ERROR:[vxVerifyGraph:2055] Node kernel init failed
5572.034674 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
Traceback (most recent call last):
File "/opt/edge_ai_apps/apps_python/./app_edgeai.py", line 71, in <module>
main(sys.argv)
File "/opt/edge_ai_apps/apps_python/./app_edgeai.py", line 45, in main
demo = EdgeAIDemo(config)
File "/opt/edge_ai_apps/apps_python/edge_ai_class.py", line 101, in __init__
self.infer_pipes.append(InferPipe(s, self.gst_pipe))
File "/opt/edge_ai_apps/apps_python/infer_pipe.py", line 54, in __init__
self.post_proc = PostProcess.get(sub_flow)
File "/opt/edge_ai_apps/apps_python/post_process.py", line 72, in get
return PostProcessClassification(flow)
File "/opt/edge_ai_apps/apps_python/post_process.py", line 83, in __init__
self.classnames = eval(flow.model.dataset)
File "<string>", line 1, in <module>
NameError: name 'modelmaker' is not defined
5572.927189 s: VX_ZONE_ERROR:[vxGetStatus:713] Reference is NULL
I am seeking assistance in resolving this issue so that I can successfully run my image classification AI model on the BeagleBone AI-64 platform.
Your prompt assistance and guidance would be greatly appreciated. If there are any additional details or logs needed to diagnose and resolve this problem, please let me know, and I will provide them promptly.
Thank you for your attention to this matter.
Regards,
Raveen
