Write and illustrate a childrens story book using Open Source AI tools
Clarendon Room E | Thu 16 Mar 10:45 a.m.–12:25 p.m.
Presented by
-
Alastair D'Silva
https://alastair.d-silva.org
Alastair is a software engineer who is passionate about open source, home automation and machine learning. For the past year, he has taken a career pivot into running short term rental apartments, to make more time available to spend with his kids.
Alastair D'Silva
https://alastair.d-silva.org
Abstract
In this workshop, attendees will write and illustrate a children's bedtime story, assisted with open source AI.
Attendees must provide their own laptop with a web browser and SSH, and should have generated an SSH public key in advance. The workshop content will be run on NVIDIA A10 GPU instances kindly provided by Lambda Labs.
Attendees should register at https://forms.gle/YY3jnnEzgbLZ2dFs7 to ensure a VM will be available for you.
In the first part of the workshop, we will use the open source LLaMA large language model from Meta (https://ai.facebook.com/blog/large-language-model-llama-meta-ai/) to generate the text of the story. Attendees will learn how to spin up an instance of the model, and craft prompts to guide the generated text.
Once a story has been created, and manually tweaked, we will focus on using Stable Diffusion from Stability AI (https://stability.ai/) to create illustrations for the story.
At the end of the workshop, you should have text and images that could be sent to a print-on-demand service for exciting bedtime reading.
In this workshop, attendees will write and illustrate a children's bedtime story, assisted with open source AI. Attendees must provide their own laptop with a web browser and SSH, and should have generated an SSH public key in advance. The workshop content will be run on NVIDIA A10 GPU instances kindly provided by Lambda Labs. Attendees should register at https://forms.gle/YY3jnnEzgbLZ2dFs7 to ensure a VM will be available for you. In the first part of the workshop, we will use the open source LLaMA large language model from Meta (https://ai.facebook.com/blog/large-language-model-llama-meta-ai/) to generate the text of the story. Attendees will learn how to spin up an instance of the model, and craft prompts to guide the generated text. Once a story has been created, and manually tweaked, we will focus on using Stable Diffusion from Stability AI (https://stability.ai/) to create illustrations for the story. At the end of the workshop, you should have text and images that could be sent to a print-on-demand service for exciting bedtime reading.