Lixing-Li/Abyme-Training-Dataset-Test-Scored-Iteration-1

---
dataset_info:
features:
- name: index
dtype: int64
- name: prompt
dtype: string
- name: status
dtype: string
- name: output
dtype: string
- name: error
dtype: string
- name: metrics
struct:
- name: actual_latency_seconds
dtype: float64
- name: total_llm_calls
dtype: int64
- name: max_tree_depth
dtype: int64
- name: max_subproblems
dtype: int64
- name: max_output_chars
dtype: int64
- name: theoretical_parallel_latency
dtype: float64
- name: trace_tree
struct:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list:
- name: prompt
dtype: string
- name: fragment
dtype: string
- name: main_problem
dtype: string
- name: parent_problem
dtype: string
- name: output
dtype: string
- name: type
dtype: string
- name: status
dtype: string
- name: difficulty
dtype: int64
- name: depth
dtype: int64
- name: index
dtype: int64
- name: latency
dtype: float64
- name: error_message
dtype: string
- name: is_cancelled
dtype: bool
- name: past
list: 'null'
- name: subproblems
list: 'null'
- name: subproblems
list: 'null'
- name: problem_index
dtype: int64
- name: level_num
dtype: int64
- name: type
dtype: string
- name: ground_truth
dtype: string
- name: original_problem
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 8350768
num_examples: 100
download_size: 3462531
dataset_size: 8350768
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---

数据集信息：
特征字段：
- 序号（index）：64位整数（int64）类型
- 提示词（prompt）：字符串类型
- 状态（status）：字符串类型
- 输出结果（output）：字符串类型
- 错误信息（error）：字符串类型
- 评估指标（metrics）：结构体类型，包含以下子字段：
- 实际延迟时长（actual_latency_seconds）：64位浮点数（float64）类型
- 大语言模型（Large Language Model，LLM）调用总次数（total_llm_calls）：64位整数类型
- 最大树深度（max_tree_depth）：64位整数类型
- 最大子问题数（max_subproblems）：64位整数类型
- 最大输出字符数（max_output_chars）：64位整数类型
- 理论并行延迟（theoretical_parallel_latency）：64位浮点数类型
- 追踪树（trace_tree）：结构体类型，包含以下子字段：
- 提示词（prompt）：字符串类型
- 片段（fragment）：字符串类型
- 主问题（main_problem）：字符串类型
- 父问题（parent_problem）：字符串类型
- 输出结果（output）：字符串类型
- 类型（type）：字符串类型
- 状态（status）：字符串类型
- 难度等级（difficulty）：64位整数类型
- 深度（depth）：64位整数类型
- 索引（index）：64位整数类型
- 延迟时长（latency）：64位浮点数类型
- 错误消息（error_message）：字符串类型
- 是否取消（is_cancelled）：布尔类型
- 历史节点（past）：列表类型，其元素为包含以下字段的结构体：
- 提示词（prompt）：字符串类型
- 片段（fragment）：字符串类型
- 主问题（main_problem）：字符串类型
- 父问题（parent_problem）：字符串类型
- 输出结果（output）：字符串类型
- 类型（type）：字符串类型
- 状态（status）：字符串类型
- 难度等级（difficulty）：64位整数类型
- 深度（depth）：64位整数类型
- 索引（index）：64位整数类型
- 延迟时长（latency）：64位浮点数类型
- 错误消息（error_message）：字符串类型
- 是否取消（is_cancelled）：布尔类型
- 历史节点（past）：列表类型，空值为null
- 子问题（subproblems）：列表类型，其元素为递归结构，每层子问题均包含与前述子问题相同的字段定义，空值为null
- 问题索引（problem_index）：64位整数类型
- 层级编号（level_num）：64位整数类型
- 类型（type）：字符串类型
- 基准真值（ground_truth）：字符串类型
- 原始问题（original_problem）：字符串类型
- 得分（score）：64位浮点数类型

数据划分：
- 训练集（train）：数据字节数为8350768，样本数量为100
下载总大小：3462531字节
数据集存储总大小：8350768字节

配置信息：
- 默认配置（config_name: default）：对应训练划分（split: train）的数据文件路径为data/train-*