createInferRefitter return null when engine has multi profile

AI & Data Science Deep Learning (Training & Inference) TensorRT
1

createInferRefitter return null and LOG error msg: [TRT]2: [refit.cpp::initialize::95] Error Code 2: Internal Error (Assertion p.second failed. ground expressions with duplicate layer names and role) when the engine has multi profile. But it runing ok for single profile.
So when will we support refit on multi profiles, and do we have some trick solutions presently?
If I use one engine for each profile, the performance( Memory, GPU usage) would be a problem?

2

Hi,

The below links might be useful for you.

https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__STREAM.html

For multi-threading/streaming, will suggest you to use Deepstream or TRITON

For more details, we recommend you raise the query in Deepstream forum.

or

raise the query in Triton Inference Server Github instance issues section.

Thanks!