icedpanda/msmarco_passage_aug_training_gpro_phase1

该数据集包含不同配置的训练数据，每个配置都有唯一的query和相关的提示(prompt)、能力(ability)、奖励模型(reward_model)等信息。数据集适用于机器学习模型训练，特别是自然语言处理任务，其中可能涉及查询理解、角色扮演、能力评估和基于标签及分数的奖励模型训练。具体描述未在README中提供。

The dataset contains training data with different configurations, each with unique queries and associated prompts, abilities, reward models, etc. The dataset is suitable for machine learning model training, especially for natural language processing tasks, which may involve query understanding, role-playing, ability assessment, and reward model training based on labels and scores. Specific description is not provided in the README.