silk-road/Haruhi-Dialogue-Speaker-Extract

该数据集用于训练一个能够从小说的chunk中批量提取摘要和对话的模型。模型使用了大约30k的中文小说数据和20k的英文小说数据进行训练，并在qwen-1.8上进行了3个epoch的微调。数据是由OpenAI生成的，并进行了清洁处理。

This dataset is designed for training a model capable of batch-extracting summaries and dialogues from chunks of novels. It is trained on approximately 30k Chinese novel datasets and 20k English novel datasets, with 3 epochs of fine-tuning performed on Qwen-1.8. The data was generated by OpenAI and cleaned.

原始信息汇总

数据集概述

数据集目的

本数据集旨在训练一个模型，该模型能够从小说的片段中批量提取摘要和对话。