Hey Friends!
If you’re here, you know how amazing Bubble is…
I just built a little ChatGPT-powered app in 20 minutes.
Thought I’d share!
Have a great day!
5 Likes
Great start! But every user question is a conversation restart - chat should be maintaining the chat conversation context for up to 4096 tokens. The way to do this is to store the ongoing conversation to append the newest questions and answers.
1 Like
Well said! That’s definitely the next evolution on this. Will have to add that in an upcoming vid 
Any idea if fine-tuning is on the roadmap for chatgpt? Image 3-6-2023 at 5.55 PM
Thanks!
Doubt it, fine tuning is really for the base models. They are using more RL for tuning but the initial ChatGPT prompt can include few shot prompts to help the model.
See OpenAI API -
Fine-tuning is currently only available for the following base models: davinci, curie, babbage, and ada. These are the original models that do not have any instruction following training (like text-davinci-003 does for example).
1 Like