OpenVoiceOS/ovos-intent-keyword-bench-intents-for-eval
收藏资源简介:
这是一个名为OVOS intent_keyword bench — intents-for-eval的数据集,包含OVOS Plugin Arena中关键词范式意图分类模型的预测结果,基于`OpenVoiceOS/intents-for-eval`数据集。数据集为每种语言提供单独的分割,每个竞争者的预测结果以JSONL文件格式存储,遵循特定的基准合约(包括固定的数据集版本、插件版本、流水线阶段和精确匹配指标)。该数据集由可复现的基准脚本生成,用于评估意图分类模型的性能,支持多种语言如加泰罗尼亚语、丹麦语、德语、英语、西班牙语、巴斯克语、法语、加利西亚语、意大利语、荷兰语、巴西葡萄牙语和欧洲葡萄牙语。
This is the dataset named OVOS intent_keyword bench — intents-for-eval, which contains the prediction results of keyword-paradigm intent classification models in the OVOS Plugin Arena, based on the `OpenVoiceOS/intents-for-eval` dataset. The dataset provides separate splits for each language, with the prediction results of each competing model stored in JSONL file format, adhering to a standardized benchmark protocol that includes fixed dataset version, plugin version, pipeline stage, and exact match metric. This dataset is generated via reproducible benchmarking scripts and is used to evaluate the performance of intent classification models, supporting multiple languages including Catalan, Danish, German, English, Spanish, Basque, French, Galician, Italian, Dutch, Brazilian Portuguese, and European Portuguese.




