Fine tune and Capture Metrics on LLM using Prepared Data set (Step 1 and Step 2)

[bora-skylight]

@knguyenrise8 @zdeveloper @arinkulshi pair to write description for this ticket.

• Create training and validation datasets using forms from Synthea.
• Set up pipeline to ingest PDF's to be used for fine tuning (clean and process data, ground truth file creation)