Fine Tuning

This feature is currently in beta. Please reach out to [e-mail] for access to this feature!

Easily collect data, annotate, and fine tune an open source model!

What's a good use case for fine tuning?

We've seen classification or categorization workflows perform VERY well in fine tuning workflows.

Fine tuning specifically improves accuracy, speed, and cost.

Improve accuracy: We train on your specific responses, which a general model like OpenAI may not be good at.

Speed: We use smaller models, like Mistral 7B, which results in both fast training AND fast inference

Cost: Get rid of a bunch of prompt tokens that teach the prompt how to classify and completion tokens.