lightonai/ms-marco-fr-bge

ms-marco-fr-bge数据集是MS MARCO数据集的法语版本，使用了BGE-M3模型进行文档与查询的相似性挖掘，并使用bge-reranker-v2-m3模型进行评分。该数据集主要用于通过知识蒸馏训练检索模型。数据集包含三个主要部分：queries、documents和train。queries部分包含查询ID和查询文本，documents部分包含文档ID和文档文本，train部分包含查询ID、相关文档ID列表及其对应的评分。

The ms-marco-fr-bge dataset is the French version of the MS MARCO dataset, with documents similar to the query mined using BGE-M3 and then scored by bge-reranker-v2-m3. This dataset is primarily used to train a retrieval model using knowledge distillation. The dataset consists of three main parts: queries, documents, and train. The queries part contains query IDs and query texts, the documents part contains document IDs and document texts, and the train part contains query IDs, lists of relevant document IDs, and their corresponding scores.