search

ZhangGe6 / onnx-modifier

A tool to modify ONNX models in a visualization fashion, based on Netron and Flask.
MIT License
1.29k stars 158 forks source link

在shape inference後產生錯誤的output shape #59

Closed alan870616 closed 1 year ago

alan870616 commented 1 year ago

未命名1 未命名2 如圖所示，將原本的model勾選shape inference後會從原本的output shape從 float32[1, 8400, 80] -> float32[1]

ZhangGe6 commented 1 year ago

@alan870616 Sorry for the late response, and thanks for reporting.

It seems that shape inference function doesn't work correctly.

Can you provide the log in the console? It can give helpful clues to fix the problem.

alan870616 commented 1 year ago

image Thanks for reply. It seems there were not error reported in the console. The model what I use is the yolov8n which provided by offical The package version of my environment: python: 3.8.16 onnx: 1.13.0 onnx_tool: 0.6.4

ZhangGe6 commented 1 year ago

@alan870616 Hello, I have uploaded a new version to fix the problem. Please update to the latest version and have a try.