Tutorial 1B: H2O-3
Overview
This tutorial walks through the process (flow) of how you can generate an AutoDoc using H2O AutoDoc for a built model in H2O-3. To generate an AutoDoc for a built model in H2O-3:
Prerequisites
- Knowledge of H2O-3
Step 1: H2O-3 MOJO
To build an AutoDoc for a supervised learning model built-in H2O-3, you need to download its MOJO first. To learn more, see Downloading and uploading in R and Python.
For purposes of this tutorial will use a preloaded MOJO.
With the above in mind, let's continue.
Step 2: AutoDoc settings
In H2O AutoDoc, click Create new AutoDoc.
In the New report list, select From H2O-3 MOJO.
In the Report name box, enter a name for the AutoDoc (e.g.,
H2O-3 AutoDoc
).Click Next: Upload model.
Click Use credit card model.
NoteInstead of uploading a MOJO, let's use one of the preloaded MOJOs (models).
Click Upload training data.
NoteAnytime you are preparing the settings for an AutoDoc for a built model in H2O-3, you need to upload the training dataset used to build the model.
Click Use credit card training data.
NoteThe
CreditCard_Cat_Train.csv
file contains the train data used to built the model.Click Upload validation data.
NoteAnytime you are preparing the settings for an AutoDoc for a built model in H2O-3, you need to upload the validation dataset used to build the model.
Click Use credit card validation data.
NoteThe
CreditCard_Cat_Valid.csv
file contains the validation data used to built the model.Click Next: Upload test data.
Click Skip test data. For purposes of this tutorial, we will not upload the test dataset.
Note- For purposes of this tutorial, we will not upload the test dataset of the built model in H2O-3.
- Anytime you want to generate an AutoDoc for a built model in H2O-3, you don't need to provide a test or used test dataset.
- Not providing a test dataset will lead to the AutoDoc (report) not containing an overview of the validation dataset.
Click Create AutoDoc.
- Submit and view feedback for this page
- Send feedback about H2O AutoDoc to cloud-feedback@h2o.ai