caffe模型转rknn时出现一下问题，求解答，感谢

[复制链接] · 发表于 2019-6-19 19:47:05

--> Loading model...
done
--> Build RKNN model...
W extend elemwise_add02_15 to add will cause accuracy loss, do not extend.
W extend elemwise_add03_20 to add will cause accuracy loss, do not extend.
W extend elemwise_add05_32 to add will cause accuracy loss, do not extend.
W extend elemwise_add06_37 to add will cause accuracy loss, do not extend.
W extend elemwise_add07_42 to add will cause accuracy loss, do not extend.
W extend elemwise_add09_54 to add will cause accuracy loss, do not extend.
W extend elemwise_add10_59 to add will cause accuracy loss, do not extend.
W extend elemwise_add11_64 to add will cause accuracy loss, do not extend.
W extend elemwise_add12_69 to add will cause accuracy loss, do not extend.
W extend elemwise_add13_74 to add will cause accuracy loss, do not extend.
W extend elemwise_add15_86 to add will cause accuracy loss, do not extend.
W extend elemwise_add16_91 to add will cause accuracy loss, do not extend.
done
--> Export RKNN model to: out_rknn/recognition_feature.rknn
done

只看该作者 · 发表于 2019-6-26 08:52:51

这里不是成功了么。。。
只是warning精度有所损失而已