From the source
It covers data quality checks, formatting, and train/evaluation splits, with references to 
Amazon· 26 AugPreparing data for supervised fine-tuning Part 1: Formatting and quality
Guidance on preparing data for supervised fine-tuning (SFT) of foundation models.
From the source
This post, the first of a two-part series, covers the foundations of SFT data preparation: quality checks, formatting requirements, and train/evaluation splits.
aws.amazon.com