DCAgent2/aider_polyglot_daVinci_Dev_32B_20260430_164102-traces
收藏资源简介:
该数据集是一个包含多轮对话记录的数据集,主要用于任务执行和对话分析。每个数据样本包括对话内容(conversations),其中包含角色(role)和内容(content)字段,以及代理(agent)、模型(model)、模型提供者(model_provider)、日期(date)、任务(task)、集数(episode)、运行ID(run_id)、试验名称(trial_name)、结果(result)和验证器输出(verifier_output)等元数据。数据集分为训练集(train),包含670个样本,总大小约1.47GB,适用于自然语言处理、对话系统训练和评估任务。
This dataset is a collection of multi-turn conversation records, primarily used for task execution and dialogue analysis. Each data sample includes conversations with fields for role and content, along with metadata such as agent, model, model provider, date, task, episode, run ID, trial name, result, and verifier output. The dataset is split into a training set (train) with 670 examples and a total size of approximately 1.47GB, suitable for natural language processing, dialogue system training, and evaluation tasks.




