eval-laion_stageB-channel-80-8B_DCAgent2_swebench-verified-random-100-folders-traces
收藏资源简介:
该数据集是一个结构化对话交互记录集合,包含1636个训练样本。数据核心特征为多轮对话内容(conversations字段,包含角色role和内容content子字段),并关联了丰富的元数据:执行代理(agent)、使用模型(model及模型提供商model_provider)、任务描述(task)、时间戳(date)、执行场景标识(episode, run_id, trial_name)、任务执行结果(result)、验证器输出(verifier_output)以及数据来源标识(trace_source)。数据集适用于分析多轮对话系统、智能体任务执行、模型行为评估等场景,为研究对话交互、任务完成度评估及不同代理/模型在特定任务上的表现提供了结构化数据。
This dataset is a collection of structured dialogue interaction records, containing 1636 training samples. Its core features include multi-turn dialogue content (conversations field, with subfields for role and content), along with rich metadata: execution agent (agent), model used (model and model provider model_provider), task description (task), timestamp (date), execution scenario identifiers (episode, run_id, trial_name), task execution result (result), verifier output (verifier_output), and data source identifier (trace_source). The dataset is suitable for analyzing multi-turn dialogue systems, agent task execution, model behavior evaluation, and other scenarios, providing structured data for research on dialogue interactions, task completion assessment, and the performance of different agents/models in specific tasks.
- 数据集名称:eval-laion_stageB-channel-80-8B_DCAgent2_swebench-verified-random-100-folders-traces
- 数据集链接:https://huggingface.co/datasets/laion/eval-laion_stageB-channel-80-8B_DCAgent2_swebench-verified-random-100-folders-traces
- 数据集规模:
- 总下载大小:约 140.2 MB
- 数据集总大小:约 241.8 MB
- 训练集样本数:1,636 条
- 数据特征:
conversations:对话列表,包含内容(content,字符串)和角色(role,字符串)agent:代理标识符(字符串)model:模型名称(字符串)model_provider:模型提供方(字符串)date:日期(字符串)task:任务描述(字符串)episode:轮次编号(字符串)run_id:运行ID(字符串)trial_name:试验名称(字符串)result:结果(字符串)verifier_output:验证器输出(字符串)trace_source:轨迹来源(字符串)
- 数据划分:仅包含训练集(
train) - 配置文件:默认配置(
default),数据文件路径为data/train-*




