open-llm-leaderboard-old/details_Yuma42__KangalKhan-Ruby-7B-Fixed
收藏数据集概述
数据集摘要
该数据集是在评估模型 Yuma42/KangalKhan-Ruby-7B-Fixed 在 Open LLM Leaderboard 上的自动创建的。数据集包含 63 个配置,每个配置对应一个评估任务。
数据集结构
- 配置数量:63
- 创建来源:1 次运行
- 分割:每个配置包含特定分割,分割名称使用运行的时间戳。"train" 分割始终指向最新结果。
- 额外配置:"results" 存储所有运行的聚合结果,用于计算和显示在 Open LLM Leaderboard 上的聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_Yuma42__KangalKhan-Ruby-7B-Fixed", "harness_winogrande_5", split="train")
最新结果
最新结果来自 2024-02-17T00:20:49.422540 运行,包含多个任务的准确率和标准误差。以下是部分结果示例: python { "all": { "acc": 0.6347473013166217, "acc_stderr": 0.032259927653459516, "acc_norm": 0.6365178163764577, "acc_norm_stderr": 0.03290211517224385, "mc1": 0.38922888616891066, "mc1_stderr": 0.01706855268069033, "mc2": 0.5648879973341684, "mc2_stderr": 0.01540236564556069 }, "harness|arc:challenge|25": { "acc": 0.6220136518771331, "acc_stderr": 0.014169664520303098, "acc_norm": 0.6723549488054608, "acc_norm_stderr": 0.013715847940719337 }, "harness|hellaswag|10": { "acc": 0.6683927504481179, "acc_stderr": 0.004698285350019216, "acc_norm": 0.8522206731726748, "acc_norm_stderr": 0.00354155826377909 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.04512608598542128, "acc_norm": 0.28, "acc_norm_stderr": 0.04512608598542128 }, # 其他任务结果省略 }
配置详情
-
配置名称:harness_arc_challenge_25
- 数据文件:
- 分割:2024_02_17T00_20_49.422540
- 路径:
**/details_harness|arc:challenge|25_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:2024_02_17T00_20_49.422540
- 数据文件:
-
配置名称:harness_gsm8k_5
- 数据文件:
- 分割:2024_02_17T00_20_49.422540
- 路径:
**/details_harness|gsm8k|5_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:2024_02_17T00_20_49.422540
- 数据文件:
-
配置名称:harness_hellaswag_10
- 数据文件:
- 分割:2024_02_17T00_20_49.422540
- 路径:
**/details_harness|hellaswag|10_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2024-02-17T00-20-49.422540.parquet
- 路径:
- 分割:2024_02_17T00_20_49.422540
- 数据文件:
-
配置名称:harness_hendrycksTest_5
- 数据文件:
- 分割:2024_02_17T00_20_49.422540
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-anatomy|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-astronomy|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_biology|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-college_physics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-computer_security|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-econometrics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-global_facts|5_2024-02-17T00-20-49.422540.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-02-17T00-20-49.422540.parquet
其他任务路径省略
- 路径:
- 分割:2024_02_17T00_20_49.422540
- 数据文件:
}



