Specific question about fine tuning

I experimented a bit with fine tuning and came across a very interesting question:
Is it possible to train the model so it does things like: “Transform the sentence which is in the active voice into the passive voice” by just training the model over and over again: prompt=active sentence; completion=passive sentence ?
If this example would be posssible it would help me in many different areas!

Thank you,

1 Like