Train a Model

After a model in AI Studio has been created, it can be trained. 

What is Training?

Training an AI model involves providing the model with training data to learn from. You are training an algorithm to evaluate and remember details about some type of information. This creates a model that can then be applied to other similar information, transferring the characteristics from one piece of information to another algorithmically. The training data must contain the correct answer, which is known as a target or target attribute. The learning model finds patterns in the training data that map the input data attributes to the target (the answer that you want to predict), and AI Studio outputs a trained AI model that captures these patterns.

You can use the trained model to get predictions on new data for which you do not know the target. For example, let's say that you want to train an AI model to predict if an email is spam or not spam. You provide the model with training data that contains emails for which you know the target (that is, a label that tells whether an email is spam or not spam). AI Studio trains an machine learning model using this data. The result is a model that attempts to predict whether new email will be spam or not spam.  

Labels

Labels are tags given that identify your data. 

Set Labels

In Hero_Flow, a label must be a DOUBLE data type in your Input.

Examples:

You want to add an Input with data to train to predict if a picture (BINARY data type) is a cat.

Your DataLabel
<A picture of a cat>1
<A picture of a dog>0

or

You want to add an Input with data to train to predict if an email (STRING data type) is spam. 

Your DataLabel
"Hello. Can you give me a quote..."0
"Hello. Do you want to buy..."1

Train Data

To train data in an AI model:

  1. Open the AI Models dashboard in Hero Flow.
  2. Open the model you want to train in AI Studio.  
  3. Select the version number that you want to train and click Train on the toolbar. 

  4. Click OK to begin training the model

After the model begins training, click Details/ Performance to view real time information on the model being trained. 

Training a model is an iterative process. Experimenting with different parameters/versions is a natural part of the training process.

Stop or Cancel a Model in Training

When a model has started training, it can be stopped and the current training session data is stored, or it can be canceled to dismiss all data from that training session.

The stop and cancel actions can be found in AI Studio after a model has started training by clicking Training Dashboards in the toolbar and clicking the menu icon for the selected version.

After a model has started training, it might take a minute for the session data to start processing. Once the session data is processing, that training session can be stopped or canceled. 

These actions can also be accessed from the training preview dialog box that is displayed after a model has started training. 

To stop a model during training and keep the session data:

  1. Click on Details/Performance in AI Studio.
  2. Click the square Stop icon.

To cancel a model during training and dismiss the session data:

  1. Click on Details/Performance in AI Studio.
  2. Click the X Cancel icon.