Gemma-peft tutorial does not perform expected fine-tuning

[dboyker]

Executing the script described in the tutorial https://github.com/huggingface/blog/blob/main/gemma-peft.md throws this warning:

You passed a dataset that is already processed (contains an input_ids field) together with a formatting function. Therefore formatting_func will be ignored. Either remove the formatting_func or pass a dataset that is not already processed.

As formatting_func is ignored, the model is not correctly fine-tuned. The script works but there is no guarantee that the model ouput the format Quote: [...] Author: [...].

Proposed solution in this PR: #3260