Uploading trained model to Aws Model Registry

[muhammad-ahmed-ghani]

Hi @rinongal , I have finetuned the styleganNada model and saved it by torch.save(net, "model.pt") but it didn't save the ModelClass with it due to which it is not deploying on Model Registry as standalone .pt file to serve the users. Can you help me how I pack the entire model with Class ?

[rinongal]

Hi @Muhammad-Ahmad-Ghani. Unfortunately I'm not familiar with the AWS model registry, and I'm not sure what type of files it expects. Do you have some guide / official documentation for the tool you're using? I could have a quick look and see if there's any simple change that comes to mind.

[muhammad-ahmed-ghani]

@rinongal Actually I wanted to convert the trained model to torchscript but the input it takes isn't static and I am facing difficulty to convert it into torchscript because otherwise I always need to import the Model Class to load the saved model. Also wanted to boost its inference speed using ONNX but failed to do because of tracing problem.

[rinongal]

I'm honestly not familiar with ONNX or torchscript either, and I don't currently have the time to experiment with either and try to see where your issue comes up. I'll keep this issue open in case someone else reads it and can help.