penfever/a2-rl-partial_ambiguity
收藏资源简介:
该数据集是一个多智能体对话数据集,包含34,626个训练示例,总大小约1.38GB。数据特征包括对话内容(conversations,其中每条对话有内容(content)和角色(role))、智能体(agent)、模型(model)、模型提供者(model_provider)、日期(date)、任务(task)、事件(episode)、运行ID(run_id)、试验名称(trial_name)、结果(result)、指令(instruction)和验证器输出(verifier_output)。这些特征表明数据集可能用于AI模型的多智能体交互训练、评估或基准测试,涵盖不同任务和模型提供者。数据集仅提供train分割,适用于机器学习应用。
This dataset is a multi-agent dialogue dataset containing 34,626 training examples, with a total size of approximately 1.38GB. The features include conversations (with content and role for each dialogue), agent, model, model_provider, date, task, episode, run_id, trial_name, result, instruction, and verifier_output. These features suggest that the dataset may be used for multi-agent interaction training, evaluation, or benchmarking of AI models, covering various tasks and model providers. The dataset only provides a train split and is suitable for machine learning applications.




