seankski/tool-parameters-v1-1-llama3-70B

该数据集包含三个不同的配置：augmented、passing-judgments-augmented和raw。每个配置都包含用户查询、工具参数、工具种子组、套件名称、工具名称、工具调用次数、主题、数据生成消息、微调消息、套件模式、工具模式、LLM作为裁判、响应、原始套件模式、新套件模式、理由消息、拒绝消息和增强等特征。数据集主要用于与工具调用、数据生成和微调相关的任务，可能涉及自然语言处理和机器学习模型的训练与评估。

The dataset includes three different configurations: augmented, passing-judgments-augmented, and raw. Each configuration contains features such as user query, tool parameters, tool seed group, suite name, tool names, number of tool calls, topics, messages from data generation, messages for finetuning, suite schemas, tool schemas, LLM as a judge, response, original suite schemas, new suite schemas, rationale messages, refusal messages, and augmentation. The dataset is primarily used for tasks related to tool invocation, data generation, and finetuning, potentially involving natural language processing and machine learning model training and evaluation.

原始信息汇总

数据集概述

数据集配置

配置名称：augmented

• 特征：
  • user_query: string
  • tool_parameters: string
  • tool_seed_group: string
  • suite_name: string
  • tool_names: string
  • n_tool_calls: int64
  • topics: string
  • messages_from_data_generation: string
  • messages_for_finetuning: string
  • suite_schemas: string
  • tool_schemas: string
  • llm_as_a_judge: string
  • response: string
  • original_suite_schemas: string
  • new_suite_schemas: string
  • rationale_messages: string
  • refusal_messages: string
  • augmentation: string
• 分割：
  • train:
    • num_bytes: 89767862
    • num_examples: 1159
• 下载大小：19223975
• 数据集大小：89767862

配置名称：passing-judgments-augmented

• 特征：
  • user_query: string
  • tool_parameters: string
  • tool_seed_group: string
  • suite_name: string
  • tool_names: string
  • n_tool_calls: int64
  • topics: string
  • messages_from_data_generation: string
  • messages_for_finetuning: string
  • suite_schemas: string
  • tool_schemas: string
  • llm_as_a_judge: string
  • response: string
  • original_suite_schemas: string
  • new_suite_schemas: string
  • rationale_messages: string
  • refusal_messages: string
  • augmentation: string
• 分割：
  • train:
    • num_bytes: 73652960
    • num_examples: 970
• 下载大小：15572288
• 数据集大小：73652960

配置名称：raw

• 特征：
  • user_query: string
  • tool_parameters: string
  • tool_seed_group: string
  • suite_name: string
  • tool_names: string
  • n_tool_calls: int64
  • topics: string
  • messages_from_data_generation: string
  • messages_for_finetuning: string
  • suite_schemas: string
  • tool_schemas: string
  • llm_as_a_judge: string
• 分割：
  • train:
    • num_bytes: 49993965
    • num_examples: 1157
• 下载大小：10648725
• 数据集大小：49993965