search

airockchip / rknn-toolkit2

Other
1.02k stars 106 forks source link

Python 推理，结果正常，但是一直有warning #41

Open HouLingLXH opened 6 months ago

HouLingLXH commented 6 months ago

用Python测试推理yolov5， 结果正常，但是一直有warning，这个应该怎么处理？

W RKNN: [09:44:39.467] Output(output0): size_with_stride larger than model origin size, if need run OutputOperator in NPU, please call rknn_create_memory using size_with_stride. W RKNN: [09:44:39.468] Output(275): size_with_stride larger than model origin size, if need run OutputOperator in NPU, please call rknn_create_memory using size_with_stride. W RKNN: [09:44:39.468] Output(277): size_with_stride larger than model origin size, if need run OutputOperator in NPU, please call rknn_create_memory using size_with_stride.

satisl commented 6 months ago

same error

yuyun2000 commented 6 months ago

警告你怕啥，不要怂

satisl398 commented 6 months ago

最新b12的rknntoolkit已修复此问题（一直警告真的很影响调试，原本的调试信息全给警告刷走了

Survivor0113 commented 1 month ago

最新b12的rknntoolkit已修复此问题（一直警告真的很影响调试，原本的调试信息全给警告刷走了

请问一下你是使用的哪个版本的tookit2-lite呢？我将2.2.0的版本部署到香橙派5max上仍然报错了

leelightning commented 1 month ago

更新librknnrt.so文件，从官方git仓库下载 替换/usr/lib下原来的，我这样做就没有后面一串warning了 P.S.不要用conda环境