zhy520xp / mace-makefile-project

makefile 交叉编译 libmace.a,并能在嵌入式端调用GPU来跑深度学习模型
95 stars 18 forks source link

请问一下,mobilenet或者mobilenetssd在3559a cpu上跑出来的结果怎么样? #18

Closed leeburt closed 5 years ago

leeburt commented 5 years ago

问题如图。

zhy520xp commented 5 years ago

这个工程只关心移动GPU的性能,cpu的话,你可以自己去测,用ncnn框架或者mace都可以呢

leeburt commented 5 years ago

当时没有板子,依然谢谢你的答复。