SuperAnnotate Docs

Create a model

Create a customized training model to automate your project and speed up the annotation process all while delivering quality results.

Step 1: Enter your model’s name and description

Go to Neural network and select New model. Next, enter the model’s name and description.

Step 2: Choose the annotation task

For Pixel projects:

  • Instance segmentation
  • Semantic segmentation

For Vector projects:

  • Instance segmentation
  • Keypoint detection
  • Object detection

Step 3: Choose a project

Enter the name of the project(s) you will be running your model on. All the images in your project should be annotated and completed or else they won’t be included in the training.

❗️

  • You cannot choose projects that do not contain at least one completed image.
  • You cannot run a training on folders, even if they contain completed images.

Step 4: Set your training’s parameters

FunctionDescription
Base modelThe base model upon which your new model will be refined.
Number of devices (GPUs)The details of the devices (GPUs) available to train your model. You have two options: A GPU of 12GB or a GPU of 16GB
Epoch countThe number of times the dataset undergoes training. The epoch count ranges from 10 to 200.
Batch sizeThe number of images.
Learning rateThe model’s learning speed. The learning rate ranges from 0 to 1.
Image (ROIs) per batchThe number of regions of interest (ROI) per image. The ROI ranges from 2 to 512 and is a power of 2. The smaller the ROI, the faster the learning rate. Image (ROIs) per batch works only with Instance segmentation.
Evaluation periodThe number of epochs per evaluation period. If the epoch count is 10 and the evaluation period is 2, it means that the dataset will undergo an evaluation every 2 epochs. The user will be reported of the best model that resulted from the evaluation period.
Train/Test split ratioThe number of images that need to be trained to the number of images that need to be tested. The train/test split ratio ranges from 0 to 100.
GammaGamma is needed for the learning rate. Its value ranges from 0 to 1, and it will be multiplied with the learning rate.
Epochs for gammaAfter you specify the number of epochs for gamma, the learning rate will be multiplied by gamma.

Step 5: Run training

When you’re done, select Run training in the top right corner. A pop-up message will tell you how long the training will take. Click Run training to confirm your action.

After the training is complete, your new model will be available in the Smart Prediction section.

Step 6: Model training metrics

You can download the training metrics when the model training is complete. To do that, go to Neural network, select the model training, and click Download.

📘

Click Delete model to remove a model.

📘

Only Team owners and Team admins can run model trainings.

Updated 17 days ago


Create a model


Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.