add fine tuning demonstration by conducting unsupervised fine tuning on RAG documents

To help users understand differences between RAG vs fine tuning we could offer a method to fine tune a model example mistral with documents from the users workspace, then host the model and give them the ability to query it.