Comment on page

What is SmartML™?

Learn more about SmartML, Plainsight's model training backend
SmartML is Plainsight's own model training backend. SmartML enables a user to train a computer vision model using their labeled datasets without writing a single line of code. It takes a labeled dataset and configuration as input and outputs your model available for deployment through an API endpoint, a visual preview of your model detections, and evaluation metrics.

How can I use SmartML?

SmartML is used when you train a model in Plainsight. To get started with training a model from a labeled dataset, start here:

What output will I get from SmartML?

At this time, SmartML output is not available for download (with the exception of mobile models).

How can I see predictions from my trained model?

You can deploy your model using Plainsight Pipelines.
  1. 1.
    Build a batch pipeline with your model as a prediction block and specify inputs and outputs, including cloud buckets and Google Pub/Sub.
  2. 2.
    Build a streaming pipeline with your model as a prediction block and use the API to send and get predictions on your data, or output predictions to a cloud bucket or Google Pub/Sub topic.
Read more about using Plainsight Pipelines:

How can I see what version of SmartML my model was trained with?

To find the version of SmartML used to train your model:
  1. 1.
    Navigate to "Model Version Details" of the model version you want to check.
  2. 2.
    Click on the "Training Settings" tab.
  3. 3.
    Check the number listed for Model Training Version. This is the SmartML version used for training this particular version of your model.