Galileo
Search…
⌃K
🏁

Installing Galileo

Python Environment Prerequisite

Galileo works with python 3.7+ environments.
You can install the Galileo python client via a simple pip install.

Installing Galileo

pip install --upgrade pip dataquality
[Enterprise Only] Set your Galileo Console URL
To point to your custom Enterprise environment, simply call
import dataquality as dq
​
dq.set_console_url()

Login()

Simply run the following lines in your notebook/python runtime.
import dataquality as dq
​
dq.login()

For automated login and configuration, see Environment Variables​

Projects and Runs

In Galileo, your experimentation is organized into projects and runs.
  • Project: A top level namespace holding one or more associated Runs.
  • Run: A single instance of a model training/evaluation loop where data is logged

Start a Project and a Run

Initialize a new run in Galileo by calling dq.init().
dq.init(
task_type="text_classification",
project_name="my_awesome_project",
run_name="my_awesome_run"
)
project_name and run_name are optional. If not provided, a generated name for each will be created and used.
Or see below to integrate your own model.

Task Types supported

Choose one of the following task types supported by Galileo:
  1. 1.
    ​Text Classification - Binary or multi-class models
​
Last modified 3d ago