TexasInstruments
How to get the yolov7.prototxt
Hello,
I know that yolov7.prototxt is the same as yolov3 from your issue, but when i write the file, i meet some difficulties. when i export the yolov7.onnx, and i use netron tool to look the net structure to get "yolo_param(input)", and my yolov7.onnx model can't show the number and show the str, my sample is as follow:
str.zip
I inquiry a similar issue on netron official website as https://github.com/lutzroeder/netron/issues/1036
expect your reply soon thanks !
Hello, I know that yolov7.prototxt is the same as yolov3 from your issue, but when i write the file, i meet some difficulties. when i export the yolov7.onnx, and i use netron tool to look the net structure to get "yolo_param(input)", and my yolov7.onnx model can't show the number and show the str, my sample is as follow:
expect your reply soon thanks !
Hello Did you solve this issue?
i use netron tool to look the net structure to get "yolo_param(input)", and my yolov7.onnx model can't show the number and show the str, my sample is as follow:
Are you saying that in your model the layer input/output names are strings and not numbers? What is the problem if the yolov7.onnx model can't show the number and show the str? I couldn't understand the exact problem.
i use netron tool to look the net structure to get "yolo_param(input)", and my yolov7.onnx model can't show the number and show the str, my sample is as follow:
Are you saying that in your model the layer input/output names are strings and not numbers? What is the problem if the yolov7.onnx model can't show the number and show the str? I couldn't understand the exact problem.
Hello @mathmanu He is explaining that the output of every node is represented in terms of name of layer, but in ti model like yolov5 ti lite model the output of every node is named with number rather than name of output node
Having name for layer is not an issue, but too large name can cause issue in TIDL. That is why the name is replaced with number. You can see a sample function to do that here: https://github.com/TexasInstruments/edgeai-tidl-tools/blob/master/scripts/osrt_model_tools/onnx_tools/onnx_model_opt.py#L151
Having name for layer is not an issue, but too large name can cause issue in TIDL. That is why the name is replaced with number. You can see a sample function to do that here: https://github.com/TexasInstruments/edgeai-tidl-tools/blob/master/scripts/osrt_model_tools/onnx_tools/onnx_model_opt.py#L151
Hello Can we use this script in order to replace output names of node in onnx model with number? Since I am working with yolov8 I am facing the same issue. Thanks
You can use that function to do that; the whole script does other things as well; just use the function to avoid confusion.