palex19/trading-qa-data
收藏资源简介:
Trading papers QA - full v2 是一个使用 Unsloth Recipe Studio 生成的合成数据集,专注于交易论文的问答内容。该数据集包含 1,200 条生成记录,具有 3 列,其中一列是 llm_structured_1,类型为字典,表示由大型语言模型结构化的数据。数据集旨在提供高质量的交易相关问答数据,用于自然语言处理任务,如问答系统训练或评估。生成过程涉及 llm-structured 和 seed-dataset 配置,确保数据多样性和关系控制。
Trading papers QA - full v2 is a synthetic dataset generated using Unsloth Recipe Studio, focusing on question-answering content related to trading papers. This dataset contains 1,200 generated records with 3 columns, one of which is `llm_structured_1` — a dictionary type representing structured data produced by large language models (LLMs). The dataset aims to provide high-quality trading-related question-answering data for natural language processing (NLP) tasks such as training or evaluating question-answering systems. The generation process involves `llm-structured` and `seed-dataset` configurations to ensure data diversity and relational control.





