Upload your model

After you have created your model, you can easily upload it to make it available in Nx Cloud and, thereby, to all your Nx Servers.

First, click on the "Add a custom model" button at the top left of the sidebar:

If more organisations are available to your account, you may need to choose an organisation before continuing:

Next, choose a type of model that matches the model file you have. In the case of an ONNX model it will be the "ONNX" upload button:

Drag and drop your ONNX model into the area or click the area and choose a model from your file system. Then choose a name for your model, add some documentation, and if needed set its normalization values:

When you drop a model on the upload area, some details will be pre-filled based on the filename, please add the documentation with more relevant information about your model.

In the next field you may select extra conversions that you can enable for the AI accelerator architecture that you need for your implementation.

The normalization will be pre-filled with some common defaults. If your model has different settings they can be set here, or edited later.

After the model is uploaded, the selected conversions will be run to generate optimized model files for each selected target AI accelerator.

When this is completed you will get an email and the model is ready to be used on your devices.