lihaoxin2020/ki-Llama-3.1-8B-Instruct-temp0.0-on-mmlu_pro-0shot_cot-scillm-9258618651

---
dataset_info:
- config_name: biology
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 2648643
num_examples: 200
download_size: 2570902
dataset_size: 2648643
- config_name: chemistry
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 12155032
num_examples: 200
download_size: 12065191
dataset_size: 12155032
- config_name: computer science
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 5229658
num_examples: 200
download_size: 5161670
dataset_size: 5229658
- config_name: engineering
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 16477900
num_examples: 200
download_size: 16373309
dataset_size: 16477900
- config_name: health
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 2032617
num_examples: 200
download_size: 1967962
dataset_size: 2032617
- config_name: math
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 8740597
num_examples: 200
download_size: 8672137
dataset_size: 8740597
- config_name: physics
features:
- name: data
list:
- name: content
dtype: string
- name: role
dtype: string
- name: raw_response
struct:
- name: text
dtype: string
- name: raw_text
dtype: string
- name: knowledge_pieces
list: string
- name: doc_id
dtype: int64
- name: native_id
dtype: int64
splits:
- name: train
num_bytes: 12692108
num_examples: 200
download_size: 12614879
dataset_size: 12692108
configs:
- config_name: biology
data_files:
- split: train
path: biology/train-*
- config_name: chemistry
data_files:
- split: train
path: chemistry/train-*
- config_name: computer science
data_files:
- split: train
path: computer science/train-*
- config_name: engineering
data_files:
- split: train
path: engineering/train-*
- config_name: health
data_files:
- split: train
path: health/train-*
- config_name: math
data_files:
- split: train
path: math/train-*
- config_name: physics
data_files:
- split: train
path: physics/train-*
---

数据集信息如下：
1. 生物学（biology）配置项
特征字段包括：
- `data`：列表类型，包含两个子字段：
- `content`：内容，数据类型为字符串
- `role`：角色，数据类型为字符串
- `raw_response`：结构体类型，其子字段`text`为文本，数据类型为字符串
- `raw_text`：原始文本，数据类型为字符串
- `knowledge_pieces`：字符串列表类型
- `doc_id`：文档ID，数据类型为64位整型
- `native_id`：原生ID，数据类型为64位整型
数据拆分：仅包含训练集（train），字节数为2648643，样本数量为200；下载大小为2570902，数据集总大小为2648643
2. 化学（chemistry）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为12155032，样本数量为200；下载大小为12065191，数据集总大小为12155032
3. 计算机科学（computer science）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为5229658，样本数量为200；下载大小为5161670，数据集总大小为5229658
4. 工程学（engineering）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为16477900，样本数量为200；下载大小为16373309，数据集总大小为16477900
5. 健康学（health）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为2032617，样本数量为200；下载大小为1967962，数据集总大小为2032617
6. 数学（math）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为8740597，样本数量为200；下载大小为8672137，数据集总大小为8740597
7. 物理学（physics）配置项
特征字段与生物学配置项一致，数据拆分仅包含训练集，字节数为12692108，样本数量为200；下载大小为12614879，数据集总大小为12692108

配置项数据文件详情：
- 生物学配置：仅包含训练集拆分的数据文件，路径为`biology/train-*`
- 化学配置：仅包含训练集拆分的数据文件，路径为`chemistry/train-*`
- 计算机科学配置：仅包含训练集拆分的数据文件，路径为`computer science/train-*`
- 工程学配置：仅包含训练集拆分的数据文件，路径为`engineering/train-*`
- 健康学配置：仅包含训练集拆分的数据文件，路径为`health/train-*`
- 数学配置：仅包含训练集拆分的数据文件，路径为`math/train-*`
- 物理学配置：仅包含训练集拆分的数据文件，路径为`physics/train-*`