saurabh5/coding-agent-synth-data

该数据集包含了对话实例的详细信息，每个实例都有一个唯一标识符instance_id，以及一系列消息，每条消息包含内容content和发送者角色role。此外，每个实例还有真实标签ground_truth、对话轮数num_turns、工具调用次数tool_calls_made、数据集来源dataset以及工具上下文信息，其中工具上下文包括修补信息patches和代码仓库名称repo_name。数据集被划分为训练集，其大小为20846854字节，包含8085个示例。The dataset contains detailed information about dialogue instances, each with a unique identifier instance_id, and a series of messages that include content and sender role. Additionally, each instance has a ground truth label, number of turns in the conversation num_turns, number of tool calls made tool_calls_made, source of the dataset dataset, and tool context information, which includes patch information patches and repository name repo_name. The dataset is split into a training set, which is 20846854 bytes in size and contains 8085 examples.