Gradient Dissent: Conversations on AI cover image

Peter & Boris — Fine-tuning OpenAI's GPT-3

Gradient Dissent: Conversations on AI

00:00

Train the Model on Multiple Languages?

Initially, when we trained y hty three, we made an concerted effort not to train it on other languages than english. But even by explicitly trying to filter out most other languages, probably like a few small percentage points of the data turned out to be in other languages. And even with that, the is just incredibly good at a translation a. It's co close the state of the art in ino translation tas you know, i'm am a native swedish speaker, but i've lost my ability to write things in swedish these days, because i'd never do ita. So what i do these days is, like, i write it in english, and

Transcript
Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!
App store bannerPlay store banner
Get the app