How to Create a Classification Model
EN
Learn how to create an AI Studio classification project, import class folders as labels, review the generated training and test split, start training, and run a sample inference.
Open AI Studio from the top navigation, stay on the Projects tab, and click 'Create Project' to begin a new model workflow.
In Project Settings, enter a project name, select 'Classification', choose the training mode such as 'Auto', and select or create the Project Folder.
On the Dataset step, click 'Import folder' and choose one folder per class. Each folder becomes a class label, such as `BG2_01_Candida spp.` and `BG2_02_Listeria`.
Wait until the import finishes and the test set is generated, then confirm the green 'Ready' status and the Training and Test counts for each class.
In Review & Train, confirm the project name, Classification category, mode, Project Folder, and dataset summary, then click 'Save & Train'.
After the training job starts, open the project details to confirm the Classification setup and review the project metadata while the model is created.
When a model is available, open 'Test Model', go to the `test` folder, select an image, and click 'Run Inference'.
Review the Inference result panel to check the top prediction and the score for each class.
