SebastianBodza/synthetischer_RAG_Datensatz_prototype

GPT 3.5 erzeugter RAG-Trainingsdatensatz.

```
prompt = """You have been assigned a retrieval task: {task}
Your mission is to write one text retrieval example for this task in JSON format. The JSON object must
contain the following keys:
- 'user_query': a string, a random user search query specified by the retrieval task.
- 'positive_document': a string, a relevant document for the user query.
- 'hard_negative_document': a string, a hard negative document that only appears relevant to the query.
Please adhere to the following guidelines:
- The 'user_query' should be {query_type}, {query_length}, {clarity}, and diverse in topic.
- Both the query and documents should be in German.
- The 'positive_document' should directly answer or be about the 'user_query'.
- The 'hard_negative_document' should be topically similar to the 'user_query' but should not answer or satisfy the query.
- The 'hard_negative_document' should be subtly irrelevant, meaning it appears to be related to the 'user_query' but does not provide a useful answer or information.

- Ensure that the documents are not copies of each other and contain unique content.
- The JSON object should be properly formatted and should validate against JSON standards.

Here is an example of how your JSON object might look for a retrieval task:

```json
{{
'user_query': '...',
'positive_document': '...',
'hard_negative_document': '...'
}}
```
Your output must always be just a JSON object only, do not explain yourself or output anything else. Always create it in German! You will get tiped 1000€ if you generate the right lengths!"""
```

原始信息汇总

数据集概述

数据集类型

• 该数据集是由GPT 3.5生成的RAG训练数据集。

数据格式

• 数据集中的每个条目以JSON格式呈现，包含以下键：
  • user_query: 用户搜索查询字符串。
  • positive_document: 与用户查询相关的文档。
  • hard_negative_document: 看似相关但实际不满足查询的文档。

数据内容要求

• user_query应满足以下条件：
  • 查询类型：{query_type}
  • 查询长度：{query_length}
  • 清晰度：{clarity}
  • 主题多样性
• 所有查询和文档均为德语。
• positive_document应直接回答或与user_query相关。
• hard_negative_document应与user_query主题相似，但不回答或满足查询，且应微妙地不相关。

数据集生成规则

• 文档内容必须唯一，不可重复。
• JSON对象必须符合JSON标准，格式正确。