Legacy fine-tuning inaccuracy

I just noticed this while looking through the docs on legacy GPT3.0 models - the page at OpenAI Platform still claims that these models are the only ones available to fine-tune, despite 3.5 fine-tuning being available for a while now. Also note the typo “are current the only …”

1 Like

Welcome to the forum.

Good catch! Things are moving so quickly, it’s gotta be hard to keep everything up to date. I’m not sure there’s a specific form for documentation errors, but you might want to report to help.openai.com if you have time.

Hope you stick around the community we’ve got growing here. Once you hit Trust Level 3, you get access to AI tools here and more.

Thanks again.

ETA: Although, it could be read as those are the only ones available for non-Instruct models? Should probably be “currently” not “current” tho…