In this experiment, I tried learning about creating our own data set, how to manpulate the data using different techniques like resizing, augumenting etc and then training a model on that data.
The best of the chapter comes in understanding how the process works from end-to-end creating a final webapp which is simple but has all the abilities to classify a image by having a chance to upload one ourselves.
Also the final model trained will be exported & saved to export.pkl file which we will use for inference later stage any where.
This file consists of code where we will be using the exported model file for inference & prediction over the images we will be uploading. And all the widgets that are required to show up this notebook as a simple webapp.
Here is the link to webapp built using viola & binder:
