nusnlp/TIP-TriviaQA

--- configs: - config_name: TIP-TriviaQA data_files: - split: train path: "triviaqa/unfiltered_train_honesty.json" - split: validation path: "triviaqa/unfiltered_eval_honesty.json" - split: test path: "triviaqa/unfiltered_honesty.json" - config_name: HoneSet data_files: - split: test path: "HoneSet/HoneSet.json" - config_name: Answerability Classifier training set data_files: - split: train path: "triviaqa/triviaqa_train_ac.json" - split: test path: "triviaqa/triviaqa_eval_ac.json" - config_name: RAG training set data_files: - split: train path: "triviaqa/triviaqa_train_rag.json" - split: test path: "triviaqa/triviaqa_eval_rag.json" ---数据集配置如下： - 配置名称：TIP-TriviaQA 对应数据文件： - 训练集：路径为 `triviaqa/unfiltered_train_honesty.json` - 验证集：路径为 `triviaqa/unfiltered_eval_honesty.json` - 测试集：路径为 `triviaqa/unfiltered_honesty.json` - 配置名称：HoneSet 对应数据文件： - 测试集：路径为 `HoneSet/HoneSet.json` - 配置名称：可回答性分类器训练集（Answerability Classifier training set） 对应数据文件： - 训练集：路径为 `triviaqa/triviaqa_train_ac.json` - 测试集：路径为 `triviaqa/triviaqa_eval_ac.json` - 配置名称：检索增强生成（Retrieval-Augmented Generation，简称RAG）训练集 对应数据文件： - 训练集：路径为 `triviaqa/triviaqa_train_rag.json` - 测试集：路径为 `triviaqa/triviaqa_eval_rag.json`