I would like to train it on every word we have ever said Mm-hmm via all the transcripts Uh, maybe even all of our tweets As well as All of the show notes that we have. And then say okay, given that you've listened to every episode What do you think about x y and z so fine tuning the model is not doable on the um gpt3 and and four right now there are other models that you can Fine tune eventually. I just did some back of the napkin math here 30 30 000 minutes of audio Of every single podcast we've ever recorded which is roughly about $400 to train out on every words we have ever spoken. That's
In this Hasty Treat, Scott and Wes talk about what can be done with the OpenAI API, how to get started with it, pricing, tuning your model, and gotchas for getting started with the OpenAI API.
Sentry - Sponsor
If you want to know what’s happening with your code, track errors and monitor performance with Sentry. Sentry’s Application Monitoring platform helps developers see performance issues, fix errors faster, and optimize their code health. Cut your time on error resolution from hours to minutes. It works with any language and integrates with dozens of other services. Syntax listeners new to Sentry can get two months for free by visiting Sentry.io and using the coupon code TASTYTREAT during sign up.
Show Notes
Tweet us your tasty treats