Skip to content

Conversation

@WanliZhong
Copy link
Member

No description provided.

@WanliZhong
Copy link
Member Author

Fail on CANN backend.

[ WARN:[email protected]] global net_cann.cpp:120 initBackend DNN/CANN: layer (name=onnx_node!InstanceNormalization_16/MVN, type=MVN) is not supported by CANN backend. Going back to default backend on CPU target

@fengyuentau
Copy link
Member

Fail on CANN backend.

[ WARN:[email protected]] global net_cann.cpp:120 initBackend DNN/CANN: layer (name=onnx_node!InstanceNormalization_16/MVN, type=MVN) is not supported by CANN backend. Going back to default backend on CPU target

It should not go to MVN. We have dedicated instance norm layer already.

@WanliZhong
Copy link
Member Author

Fail on CUDA and CUDA-FP16 target.

terminate called without an active exception

@fengyuentau fengyuentau self-assigned this Dec 28, 2023
@fengyuentau fengyuentau added the benchmark Benchmark related issues & PRs label Dec 28, 2023
@fengyuentau fengyuentau added this to the 4.10.0 milestone Dec 28, 2023
@WanliZhong
Copy link
Member Author

Benchmark is already done. Raft run very very slow. I set 30s as the accepted time for the color table.

@fengyuentau
Copy link
Member

Thanks for the hard work. I propose not to put this model into the benchmark table. Lets discuss in the meeting.

@fengyuentau fengyuentau closed this Jan 5, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

benchmark Benchmark related issues & PRs

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants