Toybrick

Segmentation Fault during preprocess node upsample

lewis_brick

注册会员

积分
73
发表于 2019-7-4 16:02:16    查看: 5915|回复: 4 | [复制链接]    打印 | 显示全部楼层
本帖最后由 lewis_brick 于 2019-7-5 16:50 编辑



Are there some restrictions on build rknn model which has upsample operators?????

本帖子中包含更多资源

您需要 登录 才可以下载或查看,没有帐号?立即注册

x
回复

使用道具 举报

lewis_brick

注册会员

积分
73
 楼主| 发表于 2019-7-18 10:12:11 | 显示全部楼层
.....10 days passed, unbelievable
回复

使用道具 举报

troy

版主

积分
2358
发表于 2019-7-18 14:59:34 | 显示全部楼层
本帖最后由 troy 于 2019-7-18 15:04 编辑
lewis_brick 发表于 2019-7-18 10:12
.....10 days passed, unbelievable

What's your rknn-toolkit version? We need more information to debug this question, such as rknn-toolkit versin, your model. And you can try to close precompile config to see whether it transform passed or not.
回复

使用道具 举报

lewis_brick

注册会员

积分
73
 楼主| 发表于 2019-7-18 15:27:00 | 显示全部楼层
rknn-toolkit version is 1.0.0
i am using caffe version of HRNET which has upsample layers.
from the screen, we can see that rknn_toolkit fails to preprocess node UPSAMPLE, and throws segmentation fault.

can you provide some hints?
回复

使用道具 举报

troy

版主

积分
2358
发表于 2019-7-26 09:49:25 | 显示全部楼层
lewis_brick 发表于 2019-7-18 15:27
rknn-toolkit version is 1.0.0
i am using caffe version of HRNET which has upsample layers.
from the ...

Have you tried to close precompile configuration.
回复

使用道具 举报

您需要登录后才可以回帖 登录 | 立即注册

本版积分规则

产品中心 购买渠道 开源社区 Wiki教程 资料下载 关于Toybrick


快速回复 返回顶部 返回列表