3 ms·

I've now got access to GPT-4-0314, anybody know the difference between that and the 32k model in this post?

• mirekrusin 3y ago
  8k vs 32k context.
• ec109685 3y ago
  Yours is a snapshot of the 8k token context taken on March 14th.
  • pjot 3y ago
    Does the May 3rd release use the 32k token model?
    • ec109685 3y ago
      No, they would call that out specifically in the model name. It’s just a further snapshot so you don’t have to jump straight to the next finetuned version without testing your app.