Skip to main content

Vanderbilt University’s Data Science Institute hosting AI Summer workshop

Posted by on Tuesday, April 25, 2023 in - Generative AI, Newsletter, Vanderbilt University Data Science Institute and Vanderbilt University School of Engineering | Large Language Models, Workshop.

 

The Data Science Institute’s AI Summer workshop, from May 8 to June 2, is a free course designed to help researchers, educators, and students gain a deep understanding of the latest AI technologies and techniques, specifically deep learning and transformers, and how they can be used to solve a variety of problems.

The four-week workshop will be virtual on Zoom, and it will include both theoretical and practical sessions. The workshop has limited capacity, and priority will be given to researchers and educators who have a research goal in mind, as well as students who are interested in joining the researchers and educators on projects that arise from these workshops.

The primary goal of the AI Summer workshop is to generate new research and teaching resources on the Vanderbilt University campus using AI techniques. The recordings and resources from the workshops will be made available to everyone. The workshop will cover AI model fundamentals, including training and deployment. Participants who come to the workshop with data for a project in-hand will end the workshop with an initial model upon which to build further work. The DSI aims to stay at the forefront of discovery through data and is committed to providing opportunities for researchers, educators, and students to deepen their understanding of AI technologies and techniques.

Contact Jesse Spencer-Smith for more information on pre-workshop training if you have limited experience with Python. You can register for the Data Science Institute’s AI Summer 2023 by filling out this form. Here’s a look at the four-week course:

Week 1: Introduction to AI Models; Prompt Engineering for Generative Models

In the overview, we will introduce transformer models, the core of most modern AI models. In the remaining sessions for the week, we will focus on using generative models such as ChatGPT using prompt engineering. Many problems that once required training specialized models can now be address using prompt engineering with no coding, and far less development time. Will you be attending these sessions?

Week 2: AI-Assisted Programming for AI Applications

Yes, this is pretty meta. Whether you have little Python experience, or extensive experience, you’ll want to participate in this session and learn what coding looks like in the age of AI. We’ll be using AI to code AI frameworks to create new AI models! If you have no experience programming, you’ll want to try out some online resources to get some experience before attending. Reach out for links.

Week 3-4: Building AI Solutions / Training and Fine-Tuning Models

Building AI Solution track: Learn to create complete solutions using AI models with tools such as plugins and LangChain. Training Track: Many problems of interest cannot be addressed using Large Language Models such as ChatGPT. We’ll get hands-on experience create novel solutions by training foundation models with domain-specific data (including text, audio, and image), or training models from scratch.

In preparation for the workshop, please complete the following:

  1. Sign up for a Google Collaboratory account. The free account should be sufficient, but you will get more compute (and longer running times) if you sign up for Colab Pro at ~$5/month.
  2. Sign up for a Hugginface.co account. Again, the free account should be sufficient.
  3. Read through our paper on Prompt Engineering patterns. (We’ll be covering this in-depth, but a quick read to familiarize yourself with the concepts would be helpful.)
  4. To get a jump start on the material, visit the Huggingface.co course on Transformers and watch the introductory videos.
  5. Suggested if you will be following the model training track for the last two weeks: Preview the book Natural Language Processing with Transformers by Lewis Tunstall, Leandro von Werra and Thomas Wolf. If you are affiliated with Vanderbilt University, you can access this pre-print book (and any book by O’Reilly) free by logging into O’Reilly Media using your Vanderbilt email address. Vanderbilt licenses all content from O’Reilly. The book covers Transformers for purposes beyond text.
  6. Think about any data or problem you might want to bring to the workshop. Also begin thinking about any projects you might want to accomplish during our month. We’ll have office hours for you to work with us to get your first project off the ground!

If you have questions, or would like to discuss possible projects, please feel free to email or to schedule a time with us at https://calendly.com/dsi-data-science-team/research-consultation.

Python Primer for AI Summer

If you are new to Python or do not have much experience programming, we recommend that you go over the content below for some helpful resources before the AI Summer session begins. Altogether, the chapters linked in the document should take approximately two hours.

The following are ~10 minute video chapters on the foundations of Python. Proficiency in Python is not required for AI Summer, but we recommend you familiarize yourself to be able to read and understand Python code when needed. If you prefer to read the information, the Python course textbook is an additional resource. We have listed the most important lessons below, but you are welcome to complete other lessons as well.

Complete the following sections from FreeCodeCamp.org Scientific Computing with Python course:

  • Introduction: Python as a Language (7:48)
  • Introduction: Elements of Python (12:46)
  • Variables, Expressions, and Statements (9:41)
  • Conditional Execution (13:30)
  • Python Functions (10:30)
  • Loops and Iterations (9:59)
  • Python Lists (10:57)

Complete the following sections from FreeCodeCamp.org Data Analysis with Python course:

  • Data Analysis Example A (9:20)
  • How to use Jupyter Notebooks Intro (8:47)
  • Numpy Operations (5:03)
  • Pandas Introduction (8:07)
  • Pandas Indexing and Conditional Selection (9:20)
  • Pandas DataFrames (10:20)

 

 

Tags: , , , , , , , , , , , , , , , ,