UffParser: Unsupported number of graph 0

Should I post my errors there?

No. I do not suggest doing that. You can check if each topic’s case/solution can provide hints with you.

More, for you case, I still suggest you to git clone deepstream_tao_apps/configs at release/tao3.0 · NVIDIA-AI-IOT/deepstream_tao_apps · GitHub and follow readme step by step to run yolo models again.

The Git link that you have shared does not have YOLOV4_TINY…
I am working with YOLOV4_TINY models…

So, do you mean your model is trained with yolov4_tiny network?

Yes, That’s right.
“yolov4_cspdarknet_tiny”… My etlt model says it.
It is in my config file that I have shared

OK. For yolov4_tiny, please generate tensorrt engine firstly.
See YOLOv4-tiny - NVIDIA Docs

Currently, YOLOv4-tiny etlt files are not supported by DeepStream

For YOLOv4-tiny, you will need to build the TensorRT open-source plugins and custom bounding-box parser. The instructions to build TensorRT open-source plugins are provided in the TensorRT Open Source Software (OSS) section above. The instructions to build a custom bounding-box parser are provided in the Prerequisites for YOLOv4-tiny Model section below, and the required code can be found in this GitHub repo.

okay, I was trying something similar…
I will try that.

After building the TRT OSS plugin, I copied the “libnvinfer_plugin.so.7.p.q” to the below path

/usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so.7.p.q...

But When I run the tao app, below is the error…
Failed to load plugin ‘/usr/lib/x86_64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_infer.so’: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory

Please run below and share result.
$ ll /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so*

I did!!
This is the output…
rwxr-xr-x 1 root root 3711136 Mar 17 13:39 /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so*
lrwxrwxrwx 1 root root 26 Mar 17 13:55 /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so.7 → libnvinfer_plugin.so.7.2.2*
-rwxr-xr-x 1 root root 3711136 Mar 17 12:13 /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so.7.0.0*
-rwxr-xr-x 1 root root 4829728 Mar 17 13:55 /usr/lib/x86_64-linux-gnu/libnvinfer_plugin.so.7.2.2*

When I pass an Image while running the ./ds-tao-detection…
The error message is “One element could not be created. Exiting.”

When I pass a video or a stream while running the ./ds-tao-detection…
The error message is “input file only support H264 and JPEG”

Please strictly follow https://github.com/NVIDIA-AI-IOT/deepstream_tao_apps/tree/master/TRT-OSS/x86#build-tensorrt-oss-plugin---libnvinfer_pluginso

You can also refer to

I have followed all the steps for TRT-OSS build, I have not missed any steps, as the steps clearly convey what they are doing, and are not difficult to understand…
I have solved this error “/usr/lib/x86_64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_infer.so’: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory”

When I pass an Image while running the ./ds-tao-detection…
The error message is “One element could not be created. Exiting.”

When I pass a video or a stream while running the ./ds-tao-detection…
The error message is “input file only support H264 and JPEG”

I am having problem understanding the above error message because the message is very VAGUE.
If you can help me decode the error message, it will be a great help.

See GitHub - NVIDIA-AI-IOT/deepstream_tao_apps: Sample apps to demonstrate how to deploy models trained with TAO on DeepStream , the input can be an Image or an h264 video file.

-i: H264 or JPEG input file

Well, I have tried using both…
I am ending up with errors.

Please share both logs and commands

These are the error messages…
The commands are…

./ds-tao-detection -c …/…/configs/yolov4-tiny_tao/yolov4_TINY.txt -i …/…/…/sample_videos/test.mp4
./ds-tao-detection -c …/…/configs/yolov4-tiny_tao/yolov4_TINY.txt -i …/…/…/sample_videos/1.jpg

Could you please share all the logs and commands?

/ds-tao-detection -c …/…/configs/yolov4-tiny_tao/yolov4_TINY.txt -i …/…/…/sample_videos/test.mp4 -b 13
input file only support H264 and JPEG

./ds-tao-detection -c …/…/configs/yolov4-tiny_tao/yolov4_TINY.txt -i …/…/…/sample_videos/1.jpg -b 13
One element could not be created. Exiting.