I’m currently working on an image recognition project, and the pre-processing steps before the model execution take a considerable amount of time. When an error occurs in the model, I have to re-run everything from the beginning, which is very time-consuming. Is there a way to avoid running the entire process from scratch and only execute the model part?
If the code is not executed from the beginning, the package installation and labeling parts in the initial sections do not run properly.
mason is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.
Try saving the intermediate calculation results? If the input features are not involved in model training at that time, you can first process the original images into features and then train.