squeezebits/augmented-json-mode-eval

这是一个增强版的 NousResearch/json-mode-eval 数据集。原始数据集包含用于评估模型遵循 JSON 模式指令能力的示例，而这个增强版包括具有不同模式提示格式变化的额外变体。此数据集已被过滤，移除了不支持的 **xgrammar** 和 **llguidance** JSON 模式特征。数据集共有 300 个样本（92 个原始 + 208 个增强），语言为英语。数据集的特征包括：对话消息（包含角色和内容）、地面真实 JSON 响应和 JSON 模式。

This is an augmented version of the NousResearch/json-mode-eval dataset. The original dataset contains examples for evaluating models ability to follow JSON schema instructions, and this augmented version includes additional variations with different formatting of the schema prompt. The dataset has been filtered to remove samples containing JSON schema features that are not supported by xgrammar and llguidance. There are a total of 300 samples (92 original + 208 augmented) in the dataset, and the language is English. The features of the dataset include conversation messages (with role and content), ground truth JSON responses, and JSON schemas.