Lee will be walking us, hands-on, slow, step-by-step, through OpenAI APIs, function calling + chaining, assistant modes etc. The purpose is to learn what the new platform offers. In addition to learning how to use LLMs, specifically, how can we functionalize them for epidemiology research applications (epiVerse)
Lee's notebook: https://observablehq.com/d/bf2deb02c5a87919
...
I'll be using the GC content calculation as an example in my own testing.
1) Fetch genome sequence
function description: https://github.com/episphere/gpt/blob/main/functions/funFetchUCSC.json
function code: https://github.com/episphere/gpt/blob/main/functions/funFetchUCSC.js
2) Calculate GC content
function description: https://github.com/episphere/gpt/blob/main/functions/funGCcontent.json
function code: https://github.com/episphere/gpt/blob/main/functions/funGCcontent.js
Using Azure and OpenAI with CBIIT managed tokens - update
epiVerse Layout
WebNN polyfill: https://github.com/webmachinelearning/webnn-polyfill