I want to test out if I can fine tune a model to produce better tabular data than the current model i have (which uses alot of prompt engineering).
I checked the Playground but wasnt sure if possible to fine tune to produce tabular data or if it was just for producing better natural language responses. Does anyone know?