Re:GPT-4
by sebastianconcpt on Tuesday March 14, 2023 @01:10PM
(#35154777)
And how it works? you can build a context and then ask something in a prompt using it?
[ Parent
| on HN ]
Re:GPT-4
by teruakohatu on Tuesday March 14, 2023 @01:14PM
(#35154839)
Context is how many tokens it can be fed to produce an output. So now you can feed it up to 32k words (tokens).
Re:GPT-4
by O__________O on Tuesday March 14, 2023 @01:18PM
(#35154926)
Token is 0.75 words on average per OpenAI; 32k in tokens would be roughly 24k in words.
https://help.openai.com/en/articles/4936856-what-are-tokens-...