onnx_adaptee.h line 2128 can not process operator name:Pad__82 type:Pad attributes

huawei-noah / bolt

Bolt is a deep learning library with high performance and heterogeneous flexibility.

https://huawei-noah.github.io/bolt/

MIT License

910 stars 158 forks source link

Open xuezhongfei2008 opened 2 years ago

xuezhongfei2008 commented 2 years ago

What does onnx switch to bolt mean when reporting this error? Is it not supported?

yuxianzhi commented 2 years ago

can you send model to cos_wave@163.com?

yuxianzhi commented 2 years ago

If your model contains Shape operator, you can firstly use https://github.com/daquexian/onnx-simplifier simplify your onnx model.

yuxianzhi commented 2 years ago

如果已发送模型，请跟我说

xuezhongfei2008 commented 2 years ago

如果已发送模型，请跟我说

Sorry, i don't have outgoing permission

yuxianzhi commented 2 years ago

can you give netron visulization of error node and its all inputs?