Closed Strux-DK closed 8 months ago
can you get a shell in the container and see what exists at /
?
It needs to be from within the frigate container
Okay can you ls /openvino-model
Well, that makes a lot of sense
Could it be a permission issue? I just migrated from HAOS add-on, so all this is a bit new to me. If so, please tell me how to proceed
I had the same problem. The issue is around the format of the config file that is putting an extra CR character at the end of the file name. To get around it i put the model block at the end of the file so there is no more lines after. Someone could give proper fix but this worked for me.
@GobbleTurkey unfortunately that did not work for me
@Strux-DK: The model filename is "coco nine one c l" (the letter L), not "coco nine one c one".
This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.
I've been getting the same issue as the above - @Strux-DK @cgils did you ever get to the bottom of what this error means?
2024-07-14 14:38:37.776116763 [35mE: [xLinkUsb] [ 317776] [frigate.detecto] usb_find_device_with_bcd:266 Library has not been initialized when loaded[0m
2024-07-14 14:38:37.776584314 [35mE: [xLinkUsb] [ 317776] [frigate.detecto] usb_find_device_with_bcd:266 Library has not been initialized when loaded[0m
2024-07-14 14:38:38.010492509 [2024-07-14 14:38:38] frigate.detectors.plugins.openvino INFO : Model Input Shape: [1,300,300,3]
2024-07-14 14:38:38.010668642 [2024-07-14 14:38:38] frigate.detectors.plugins.openvino INFO : Model Output-0 Shape: [1,1,100,7]
2024-07-14 14:38:38.010671199 [2024-07-14 14:38:38] frigate.detectors.plugins.openvino INFO : Model has 1 Output Tensors
2024-07-14 14:38:45.730888726 Process detector:ov:
2024-07-14 14:38:45.730892171 Traceback (most recent call last):
2024-07-14 14:38:45.730893635 File "/usr/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
2024-07-14 14:38:45.730894575 self.run()
2024-07-14 14:38:45.730895689 File "/usr/lib/python3.9/multiprocessing/process.py", line 108, in run
2024-07-14 14:38:45.730899327 self._target(*self._args, **self._kwargs)
2024-07-14 14:38:45.730919398 File "/opt/frigate/frigate/object_detection.py", line 125, in run_detector
2024-07-14 14:38:45.730921077 detections = object_detector.detect_raw(input_frame)
2024-07-14 14:38:45.730922748 File "/opt/frigate/frigate/object_detection.py", line 75, in detect_raw
2024-07-14 14:38:45.730923864 return self.detect_api.detect_raw(tensor_input=tensor_input)
2024-07-14 14:38:45.730924934 File "/opt/frigate/frigate/detectors/plugins/openvino.py", line 84, in detect_raw
2024-07-14 14:38:45.730949273 infer_request.infer([tensor_input])
2024-07-14 14:38:45.730950724 File "/usr/local/lib/python3.9/dist-packages/openvino/runtime/ie_api.py", line 152, in infer
2024-07-14 14:38:45.730952391 return super().infer(normalize_inputs(self, dict(enumerate(inputs))))
2024-07-14 14:38:45.730953494 RuntimeError: The input blob size is not equal to the network input size: got 307200 expecting 270000
Describe the problem you are having
Just bought new cameras and was trying to get one of them to work. To do that, i had to comment out ffmpeg, detectors and model sections. I don't know why, but it kept throwing different errors until i did. Anyway, i got it working with those sections commented out, so i wanted to restore to the config i had before trying to get the camera to work, but now i can't, because apparently there is a file it can't find.
I run frigate in docker in a Proxmox LXC. The server has a i7-7700.
The only reason i put in the 'model' block is because it says so in the documentation: "Use the model configuration shown below when using the OpenVINO detector with the default model." so i did. I also tried to comment it out, but then it throws another error: "RuntimeError: The input blob size is not equal to the network input size: got 307200 expecting 270000"
As i have mentioned in previous issues i opened, i struggle with understanding all the terms, what they do and how they work together. I'm not sure what a detector or a model is or whether i need it or not, i just want hardware acceleration to work, so i can utilize my hardware as much as possible.
Version
0.12.1-367d724
Frigate config file
docker-compose file or Docker CLI command
Relevant log output
Operating system
Debian
Install method
Docker Compose
Coral version
CPU (no coral)
Any other information that may be helpful
No response