Issue about inferece dataset structure for Tao-Toolkit-API multi-class classification model


Hello. In the multi-class classification notebook of Tao-Toolkit-API, there should be the class names in the datastrure.

However. When I do model inference, the custom dataset of mine has not classified yet.

Is it possible to do multi-class classification model inference by Tao-Toolkit-API without classifying dataset ?

Can you elaborate more? Or share the .ipynb notebook file.

As the picture below, the description of dataset structure in the multi-class calssification notebook is required to set the classes firstly.

In my case, I have to do model inference with the dataset which has not been classified yet.

Could I upload the unclassified dataset used for inference in the process of model inference?

There is no update from you for a period, assuming this is not an issue anymore. Hence we are closing this topic. If need further support, please open a new one. Thanks

Should be working. You can run the last “TAO Inference” section.

TAO API is not forcing directory structure inside images_test.
You can have just raw images under images_test.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.