Logan just announced this on twitter.
Fine tuning will be getting a UI in the coming months!
7 Likes
Nice. This will be useful for a lot of people moving forward…
2 Likes
Was this announced in this forum? I hate having to go to discord, twitter, email, etc. to get updates. Nice if there was one place.
1 Like
It was twitter, but I’ll make sure than anything on discord, twitter, the website, some random blog, gets posted to the community section.
2 Likes
It’s live for me now, not sure if it is for everyone.
1 Like
Can confirm, it’s live for me as well! 
2 Likes
Is there a rough estimate of when we’ll be able to fine tuning a fine tuned model? Will help creating personalized models that change frequently.
You can now fine-tune a new model based on an existing fine-tune.
All one needs to do is specify not that the base model is “gpt-3.5-turbo” (or others), but instead the base model is your trained model name/id that you want to continue from.
1 Like
When was this released? I know when we they launched the ability to fine-tune the “gpt-3.5-turbo” you couldn’t fine-tune a model that was already fine-tuned.
I guess you can now. The last time I look this FAQ said no.
[
Can I continue fine-tuning a model that has already been fine-tuned?
Yes, you can pass the name of a fine-tuned model into the model parameter when creating a fine-tuning job. This will start a new fine-tuning job using the fine-tuned model as the starting point.
1 Like