WYS1998

Results 2 issues of WYS1998

Thank you very much for your contribution. I would like to know why the accuracy is particularly low when I replace rsnet50 and Vit with a lightweight backbone network, only 0.1mAP

Hello, after I adapted the training framework with mobilenetv2, the mAP was only 0.1. Is there something wrong with this process? Do I need to change some parameters to ensure...

各位专家们你们好，我想问一下我经过量化感知训练之后的YOLO11n导出onnx的时候会有警告，并且没办法正常转出为rknn

1 comment

你们遇到量化感知训练的问题了么，在转onnx的时候会有TracerWaring，然后在RKNN-Toolkit2的docker上面将onnx转换成rknn的时候会有非法指令（内核错误）