Using vision model from Julia API

Hello,

Is there any possibility of using a vision model defined in the Model Zoo of mxnet from the Julia API?

Thanks in advance

I don’t have much experience with Julia API, but it should be possible, since the engine of MXNet is always the same.

There is an example of how it should be possible to do here: https://github.com/apache/incubator-mxnet/blob/master/julia/examples/imagenet/ijulia-pretrained-predict/Prediction%20with%20Pre-trained%20Model.ipynb I tried to run it, but it seems like my Julia API is not properly binded… If you have one installed, you may want to try execute this notebook.

I have noticed that the model file is not available from the link the notebook references, though. You can download the model files from https://github.com/awslabs/mxnet-model-server/blob/master/docs/model_zoo.md. It is a so called model archive, and if you download, rename it to .zip and unpack, you will find a bunch of files, in which you need to pick up .json one and .params. Put it in the folder with your code and change the name of the model file to be loaded.

Let me know if it worked for you.