Skip to content

Latest commit

 

History

History
59 lines (39 loc) · 2.46 KB

README.md

File metadata and controls

59 lines (39 loc) · 2.46 KB
Raw

Babble Labble Tutorials

In this tutorial, we will walk through the process of using Babble Labble to train a classifier for recognizing mentions of spouses in a corpus of news articles. The tutorial is broken up into 3 notebooks:

  1. Tutorial1: The Babble Labble pipeline

In the first tutorial, we walk through the full Babble Labble pipeline, going from natural language explanations to a trained classifier.

  1. Tutorial2: Writing explanations

In the second tutorial, we show you how to write your own explanations and share some best practices.

  1. Tutorial3: Exploring tradeoffs

In the third tutorial, we discuss the pros and cons of a few variations of the framework.

Example

As an example of the relation that we'll be classifying in this tutorial, in this sentence (specifically, a photograph caption):

Prime Minister Lee Hsien Loong and his wife Ho Ching leave a polling station after casting their votes in Singapore (Photo: AFP)

our goal is to extract the spouse relation pair ("Lee Hsien Loong", "Ho Ching"). These sentences come from the Signal Media dataset (Corney, et al. 2016).

Setup

[1] General Babble Labble setup:

First, follow the instructions in the main repository README for setting up your environment for Babble Labble.

[2] Launch Jupyter notebook:

Run one of the following commands from the root of the repository to launch Jupyter (choose the same option that you chose for general Babble Labble setup):

Option A: Docker

jupyter notebook --ip=0.0.0.0 --port=8080 --allow-root --NotebookApp.token='' --no-browser

Open a browser on your local computer and type http://localhost:8080/.
You should see the root directory of the Babble Labble repository.

Option B: Local

jupyter notebook

This will open a tab in your browser.
You should see the root directory of the Babble Labble repository.

[3] Select the environment

Navigate to notebooks/Tutorial1_BabbleLabble.ipynb and click on it to open it.
To ensure that the notebook is using your babble conda environment with the appropriate dependencies installed, check for Python [conda env:babble] in the upper right corner of your notebook. If you don't see it, select the following from the Jupyter notebook toolbar:

Kernel > Change kernel > Python [conda env:babble]