Fine tuning saves cost

Prompting
1

Just came across fine-tuning, as I am a beginner, how much cost does it save approximately?
Also, what do they mean by -
" Once you fine-tune a model, you’ll be billed only for the tokens you use, similar to our [base models] pay-as-you-go pricing."

Regardless, Thank you for the active response! Love this community :slight_smile:

1 Like
2

Savings come from being able to use smaller finetuned models with cheaper pricing, and the fact that you can exclude prompt instructions/examples- which usually consume the majority of tokens in a api call.

image

3 Likes