adamo1139/basic_economics_questions_ts_test_1

Synthethic Question & Answer dataset trained on a corpus of the book Basic Economics by Thomas Sowell.
Formating could be improved, as model trained on this dataset write \n tokens as words and not as newline, so I guess it gets tokenized in a way different from expectations.
Note that prompt format isn't very consistent in every sample.

Spicyboros 7B gguf was used as a model that generated synthetic responses, so it was all generated locally without leaving the device, as opposed to how commonly GPT 3.5 Turbo or GPT 4 would be used for the purpose.

原始信息汇总

数据集概述

• 数据集名称：合成问题与答案数据集
• 训练语料：基于托马斯·索维尔（Thomas Sowell）的书籍《基础经济学》（Basic Economics）
• 格式问题：模型在训练过程中将换行符（ ）作为单词处理，而非实际换行，可能与预期分词方式不同。
• 提示格式：样本中的提示格式并不完全一致。
• 生成模型：使用Spicyboros 7B gguf模型生成合成响应，所有生成过程均在本地设备上完成，未使用常见的GPT 3.5 Turbo或GPT 4模型。