AgentPublic/evalap-mediatechs-service-public-travail-emploi-chunking-evaluation-v1-117
收藏资源简介：
该数据集名为MediaTech的公共服务与工作就业分块评估V1，旨在评估MediaTech的公共服务与工作就业数据集的多种分块策略。数据集包含来自EvalAP评估平台的24个实验，使用了特定的数据集（MFS_good_dataset）和评估指标（上下文精确度、上下文召回率、上下文相关性、忠实度、法官标注者）。README还列出了各种配置及其对应的数据文件，以及MFS_good_dataset下不同模型的得分。
The dataset named MediaTechs Service Public & Travail Emploi Chunking Evaluation V1 is designed to evaluate several chunking strategies for MediaTechs Service Public & Travail Emploi datasets. It contains 24 experiments from the EvalAP evaluation platform, using specific datasets (MFS_good_dataset) and metrics (contextual_precision, contextual_recall, contextual_relevancy, faithfulness, judge_notator) for evaluation. The README also lists various configurations and their corresponding data files, along with scores for different models under the MFS_good_dataset.
