burtenshaw/test-argilla-dataset

该数据集是使用Argilla创建的，包含与HuggingFace `datasets`兼容的记录。数据集的结构包括字段、问题、元数据、向量等。字段是数据集记录的特征或文本，问题是向注释者提出的问题，元数据提供有关数据集记录的附加信息，向量包含记录的向量表示。数据集包含一个名为`train`的拆分。

This dataset, created with Argilla, supports various annotation tasks including text classification, rating, ranking, multi-label selection, and span annotation. The dataset records are compatible with the HuggingFace `datasets` library and can be loaded directly. The dataset structure includes fields (e.g., text), questions (e.g., label selection, rating, ranking), metadata, vectors, and suggestions. The dataset contains only a single train split.

原始信息汇总

Dataset Card for test-argilla-dataset

数据集概述

• 创建工具: Argilla
• 数据集大小: n<1K
• 标签: rlfh, argilla, human-feedback

数据集结构

• 字段 (Fields):

  • 字段名称: text
  • 标题: text
  • 类型: text
  • 是否必需: True
  • 支持Markdown: False

• 问题 (Questions):

  • label:
    • 标题: label
    • 类型: label_selection
    • 是否必需: True
    • 描述: N/A
    • 值/标签: [positive, negative]
  • rating:
    • 标题: rating
    • 类型: rating
    • 是否必需: True
    • 描述: N/A
    • 值/标签: [1, 2, 3, 4, 5]
  • ranking:
    • 标题: ranking
    • 类型: ranking
    • 是否必需: True
    • 描述: N/A
    • 值/标签: [label1, label2, label3]
  • comment:
    • 标题: comment
    • 类型: text
    • 是否必需: True
    • 描述: N/A
    • 值/标签: N/A
  • topics:
    • 标题: topics
    • 类型: multi_label_selection
    • 是否必需: True
    • 描述: N/A
    • 值/标签: [topic1, topic2, topic3]
  • span:
    • 标题: span
    • 类型: span
    • 是否必需: True
    • 描述: N/A
    • 值/标签: N/A

• 元数据 (Metadata):

  • 元数据名称: comment_score
  • 标题: comment_score
  • 类型:
  • 值: None - None
  • 是否对标注者可见: True

• 向量 (Vectors):

  • 向量名称: vector
  • 标题: vector
  • 维度: [1, 3]

数据实例

• Argilla格式: json { "_server_id": "8aaf57d2-cb8e-4673-a7ce-2f684b60adf5", "fields": { "text": "Hello World, how are you?" }, "id": "4f56e32b-9582-47de-a2b1-b230732bb07b", "metadata": {}, "responses": { "label": [ { "user_id": "06f7d4c0-e048-43d2-ab3f-06f147616ac6", "value": "positive" } ] }, "suggestions": { "label": { "agent": null, "score": null, "value": "positive" }, "topics": { "agent": null, "score": [ 0.9, 0.8 ], "value": [ "topic1", "topic2" ] } }, "vectors": {} }

• HuggingFace datasets格式: json { "_server_id": "8aaf57d2-cb8e-4673-a7ce-2f684b60adf5", "comment.suggestion": null, "comment.suggestion.agent": null, "comment.suggestion.score": null, "comment_score": null, "id": "4f56e32b-9582-47de-a2b1-b230732bb07b", "label.responses": [ "positive" ], "label.responses.status": [ "draft" ], "label.responses.users": [ "06f7d4c0-e048-43d2-ab3f-06f147616ac6" ], "label.suggestion": "positive", "label.suggestion.agent": null, "label.suggestion.score": null, "ranking.suggestion": null, "ranking.suggestion.agent": null, "ranking.suggestion.score": null, "rating.suggestion": null, "rating.suggestion.agent": null, "rating.suggestion.score": null, "span.suggestion": null, "span.suggestion.agent": null, "span.suggestion.score": null, "text": "Hello World, how are you?", "topics.suggestion": [ "topic1", "topic2" ], "topics.suggestion.agent": null, "topics.suggestion.score": [ 0.9, 0.8 ], "vector": null }

数据分割

• 分割: train