Meet your Artificial Self: Generate text that sounds like you

13:30-17:00, January 25 @ 5A

Workshop / Overview

In this workshop, you will download your own chat logs and build a chat bot that generates text similar to your writing.

Workshop repo: https://github.com/mar-muel/artificial-self-AMLD-2020

Note: In order to use your own chat logs for the workshop, we highly recommend that you download them in advance as this process can take some time. Follow the guide here:  https://github.com/MasterScrat/Chatistics#exporting-your-chat-logs

As an alternative to using your own chat logs, we will provide other conversational datasets.

Fine-tuning large language models, such as OpenAI’s GPT-2, and making them generate tweets, poetry or even entire novels has become surprisingly simple. By feeding suitable textual data, these models do not only learn vocabulary and rules encoded in the input, but also learn to imitate the writing style and characteristics of the author.

Generating arbitrary text is interesting in itself - however we want to be able to interact with the model and get reasonable answers to our questions. For this to work, we also need to teach the model how conversations work. At the end of this workshop, you will be able to build a conversational AI system you can talk with.

Workshop / Outcome

Participants will learn the basics of language model fine-tuning for text generation and how to build a conversational AI system. Participants will be exposed to the popular PyTorch transformers library by HuggingFace.

Workshop / Difficulty

Intermediate level

Workshop / Prerequisites

  • Basic Python & Machine Learning understanding
  • Own laptop
  • Google account (we will use Google Colab)

Track / Co-organizers

Martin Müller

PhD Candidate, EPFL

Manuel Schneider

PhD Candidate, ETHZ

Florian Laurent

ML Engineer, AIcrowd

AMLD EPFL 2020 / Workshops

A Conceptual Introduction to Reinforcement Learning

With Kevin Smeyers, Katrien Van Meulder & Bram Vandendriessche

09:00-12:30 January 251ABC

Applied Machine Learning with R

With Dirk Wulff, Markus Steiner & Michael Schulte-Mecklenbeck

09:00-17:00 January 25Foyer 6

Augmenting the Web browsing experience using machine learning

With Oleksandr Paraska, Vasily Kuznetsov, Tudor Avram & Levan Tsinadze

09:00-12:30 January 253A

AMLD / Global partners