Summary repository for AI Summer 2024. Introduction to generative AI, with practical applications to inferencing and training
Presented by Vanderbilt Data Science Institute data scientists:
- Dr. Jesse Spencer-Smith, Chief Data Scientist
- Dr. Charreau Bell, Senior Data Scientist
- Myranda Shirk, Senior Data Scientist
- Umang Chaudhry, Data Scientist
- Dr. Abigail Petulante, DSI Postdoctoral Fellow
- Dr. Joshua Su, DSI Postdoctoral Fellow
The objective of these workshops is to develop foundational skills in understanding, inferencing and training generative AI models and other transformer models.
Practice your Python skills using the below documents. Choose either a Google Colab for interactive programming environment, or alternatively read through the Google Doc.
You’ll want to use the most advanced AI chat model that you can get access to. Microsoft just opened access to BingGPT through Bing Chat, which is based on an early version of GPT4, currently the most advanced AI chat model available to the public. You’ll need to install the Edge browser (https://www.microsoft.com › edge › download) and go to bing. com. Click on “Chat”.
Think about any data 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!
Session will run live from 9am-11am, with an office hour from 11am to noon (all times Central).
No class Friday (Vanderbilt Commencement)
Weeks 2, 5/13 - 5/17: Retrieval-Augmented Generation (RAG), Assistants, Agents, and Intro to Diffusion Models
Week 3, 5/20 - 5/24: Building AI Solutions, Running AI Securely Locally or in the Cloud, Introduction to Training Models
Monday:
-
Homework: Watch the following videos: General Backprop and (math-centric backprop](https://youtu.be/tIeHLnjs5U8?si=mnT36GTL7YqU8qBO)
Wednesday: Recording: (https://vanderbilt.zoom.us/rec/share/fswTlpFMlqAVgxRDDBza920i9brAuxaSiteHpDNUwpm9YQzedJa5g_2oZSSr2Eq1.wF73yKYGD5eY3cyY?startTime=1716392393000)
Friday: Recording: (https://vanderbilt.zoom.us/rec/share/plozihJcLFBIfjPxQ8Bsv9IdqHh39qFinkVUChsYtuiuiGAc8O2TcvTEbTE5cAUW.3XYBPJfbdZJ1GzAS?startTime=1716558902000)
No class Monday (Memorial Day)
Wednesday:
Papers/Blogs discussed:
https://arxiv.org/pdf/2405.17247
https://proceedings.mlr.press/v139/radford21a/radford21a.pdf
https://arxiv.org/pdf/2405.09818
https://arxiv.org/pdf/2304.10592
https://arxiv.org/pdf/2310.03744
https://huggingface.co/papers/2311.05437
https://arxiv.org/pdf/2311.05437
https://llava-vl.github.io/blog/2024-01-30-llava-next/
https://llava-vl.github.io/blog/2024-05-10-llava-next-stronger-llms/
https://llava-vl.github.io/blog/2024-04-30-llava-next-video/
https://arxiv.org/abs/2310.02239
Friday: Recording https://vanderbilt.zoom.us/rec/share/g2Q9l9ZAxfjPuPKSY9W9W9aZF_oq7a9rTcAQk2ppNzDTsZBCnpWdN3Dktd2THXY.G5FUcMWQPRFaenje
Remember we are all learning and exploring
- Please share your video upon entering the room and unmute
- Share your screens--someone volunteer to share their screen upon entering, and everyone be ready to share your screen to show what you’ve found
- Make notes of what you’ve discussed in the Response Reports below
- Everyone be ready to report out (random)
- Make some friends
- Breakout Rooms Worksheets
Google Docs has a limit of 100 people viewing/editing a document at one time.
Please be sure your display name is set in Zoom. If you are in one of the following special groups, please pre-pend your name with one of the following qualifiers.
- Data Science for Social Good: DSSG
- Center for AI in Protein Dynamics: Protein
- If you are in a lab and would like your own breakout room: Labname (keep it short, please!)
- If you are faculty and would like to be in a breakout room with other faculty: Faculty
For example, I might be DSSG-Jesse Spencer-Smith
Video recordings of these workshops can be found on our YouTube channel AI Summer playlist
Looking for the code resources for Summer 2023? View the 2023 repo version here.
- Prompt Engineering paper https://arxiv.org/abs/2302.11382
- Prompt Engineering Courserea Course: https://www.coursera.org/learn/prompt-engineering
- Visual overview of Generative AI from 3Blue1Brown: https://www.youtube.com/watch?v=wjZofJX0v4M
- Semester-long course on transformer models, DS 5690. Graduate students and advanced undergraduates can register by contacting me. I welcome auditing by a select number of postdoctoral fellows, and drop-ins from faculty!
DGX A100 Compute Grant: https://forms.gle/2mGfEy9DB4JU2GpZ8
- 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.
To get the most out of this workshop:
- Open Colab (workbook) notebooks and actively write code along with the instructor
- Actively participate in discussions
- Actively participate in breakout rooms
- Work on homework assignments before coming to class
- Relax your mind and ask questions
- Open the Edge browser (yes, Edge) and navigate to www.bing.com
- Select "chat". A new window should open saying you need the new Bing.
- Select "Start chatting" at the bottom of this window. This should prompt you to sign in to a Microsoft account. Do not use an organizational/school email (such as Vanderbilt). Instead, select "No account? Create a new one" and create one with your personal email. Note: if you get stuck in the "use the new Bing" window, go back to Bing.com and select "Sign in" instead. Follow instructions for Step 3.
- Sign in to your personal Microsoft account.
- Select "chat" - it should work now!
