Model Training

Select Training Parameters

After completing data labeling, click the Train Model button in the upper right corner of the Label & Train page.

AI Creator provides preset training parameters for each algorithm type, designed to meet the needs of common detection scenarios. It also automatically optimizes parameters like model input size and epoch count based on the labeled data and image dimensions.

TIP

Detailed descriptions of the training parameters are available in tooltips on the interface. You can hover over the tooltip icon next to each parameter to see more details.

Since our task is to simply locate the main area of the chip, we can use the default parameters. Click the Start Training button to begin training the model.

Training Process

Once training starts, you'll automatically enter the training details interface, where you can monitor real-time logs of the training process. For example, you can see which epoch is currently being trained and observe metrics like train/validation loss and IoU in real-time.

Training Completion

AI Creator will automatically evaluate and save multiple training models. After training is complete, the system will select the best-performing model and save it as a version. The system will also automatically perform a model evaluation based on the selected model.

Next Step

After completing the model training, you can proceed to Model Evaluation to assess the performance of the model.