cyrusbehr


Yes that's true, although disabling that option also significantly slows down inference on machines supporting SIMD registers such as AVX2 registers. Ultimately I would like to see the AVX2 optimizations...

Do you have a build script you could share?

@linghu8812 I have an arcface r100 which is identical to the insightface r100 arcface model, with the exception that the output feature vector is 500 elements instead of 512. When...

Try they export scripts here: https://github.com/deepinsight/insightface/issues/1350 ```python import argparse import onnx import mxnet as mx import numpy as np from mxnet.contrib import onnx as onnx_mxnet import mxnet.contrib.onnx.mx2onnx.export_onnx as mx_op from...

Same issue. Any updates on this?

Essentially, in the `link_directories` command, `protobuf` must be infront of `TensorRT`, as TensorRt has it's own protobuf files in that directory. So this works: ``` link_directories( ${PROTOBUF_DIR}/lib ${TENSORRT_DIR}/lib ) ```...

Great thank you for putting this together. I will review in the next few days.

> Can we close this PR? Yes let's close this for now. I however would like to add Windows support at some point. With this particular PR, I never had...

Hey, I think that would be a great addition. Yeah feel free to add the feature and then open a PR. I'll give you contribution credit in the project once...