orca-agentinstruct-1M-v1 microsoft
Dataset Card This dataset is a fully synthetic set of instruction pairs where both the prompts and the responses have been synthetically generated, using the AgentInstruct framework. AgentInstruct is an extensible agentic framework for synthetic data generation. This dataset contains ~1 million instruction pairs generated by the AgentInstruct, using only raw text content publicly avialble on the Web as seeds. The data covers different capabilities, such as text editing, creative… See the full description on the dataset page: https://huggingface.co/datasets/microsoft/orca-agentinstruct-1M-v1.
- 类型
- dataset
- 许可
- cdla-permissive-2.0
- 语言
- en
- 下载量
- 1,584
- 点赞
- 465
- 访问
- public
- 文件
- 0
标签
- 合成数据
- 指令微调
- 对话数据
- 大语言模型
- Agent
- 文本数据集
- 机器学习
摘要
该数据集由微软使用AgentInstruct代理框架，仅以公开网页文本为种子，全合成生成约100万个指令-响应对，覆盖文本编辑、创意写作、代码、阅读理解、开放域问答等多种能力。可用于任意基础大语言模型的指令微调，其超集用于后训练Mistral-7b后在AGIEval、MMLU、GSM8K等基准上取得显著提升。
README
--- language: - en license: cdla-permissive-2.0 size_categories: - 1M<n<10M task_categories: - question-answering dataset_info: features: - name: messages dtype: string splits: - name: creative_content num_bytes: 288747542 num_examples: 50000 - name: text_modification num_bytes: 346421282 num_examp…