This dataset is a reformatted version of practical-dreamer/RPGPT_PublicDomain-alpaca into a ShareGPT-like format. Unfortunately, detailed information about the original dataset is scarce.
In total, the dataset includes 3 032 conversations, with some extending up to 50 turns.
The dataset consists of the following fields:
conversations: ShareGPT-like format representing the dialogue.
The 'system' message provides a randomized introduction… See the full description on the dataset page:
https://huggingface.co/datasets/agentlans/practical-dreamer-RPGPT_PublicDomain.