open-llm-leaderboard-old/details_NLUHOPOE__test-case-5
收藏Hugging Face2024-02-18 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_NLUHOPOE__test-case-5
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在Open LLM Leaderboard上对模型NLUHOPOE/test-case-5进行评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集包含1次运行的结果,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。此外,还有一个results配置存储了所有运行的聚合结果,用于计算和显示Open LLM Leaderboard上的聚合指标。文件还提供了如何使用`datasets`库中的`load_dataset`函数加载运行细节的示例。还包括了2024年2月18日特定运行的最新结果,显示了不同任务的各种准确率指标。
该数据集是在Open LLM Leaderboard上对模型NLUHOPOE/test-case-5进行评估运行期间自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集包含1次运行的结果,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。此外,还有一个results配置存储了所有运行的聚合结果,用于计算和显示Open LLM Leaderboard上的聚合指标。文件还提供了如何使用`datasets`库中的`load_dataset`函数加载运行细节的示例。还包括了2024年2月18日特定运行的最新结果,显示了不同任务的各种准确率指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总
数据集概述
数据集创建
- 创建背景:该数据集是在模型 NLUHOPOE/test-case-5 在 Open LLM Leaderboard 上的评估运行期间自动创建的。
- 数据集组成:包含 63 个配置,每个配置对应一个评估任务。
- 数据来源:数据集来自 1 次运行,每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。"train" 分割始终指向最新的结果。
- 额外配置:"results" 配置存储所有运行的聚合结果,用于计算和显示 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_NLUHOPOE__test-case-5", "harness_winogrande_5", split="train")
最新结果
- 最新结果时间戳:2024-02-18T04:29:41.534828
- 结果示例:
python
{
"all": {
"acc": 0.5577223560543467,
"acc_stderr": 0.03385647798613642,
"acc_norm": 0.5613979867951425,
"acc_norm_stderr": 0.03456272133508181,
"mc1": 0.33659730722154224,
"mc1_stderr": 0.01654241280949489,
"mc2": 0.4999721417565135,
"mc2_stderr": 0.01533600099002958
},
"harness|arc:challenge|25": {
"acc": 0.5358361774744027,
"acc_stderr": 0.01457381366473572,
"acc_norm": 0.5656996587030717,
"acc_norm_stderr": 0.014484703048857354
},
"harness|hellaswag|10": {
"acc": 0.5984863572993427,
"acc_stderr": 0.004892026457294713,
"acc_norm": 0.7903804023102967,
"acc_norm_stderr": 0.004062052570310901
},
"harness|hendrycksTest-abstract_algebra|5": {
"acc": 0.26,
"acc_stderr": 0.04408440022768081,
"acc_norm": 0.26,
"acc_norm_stderr": 0.04408440022768081
},
其他任务的结果...
}
配置详情
-
配置名称:harness_arc_challenge_25
- 数据文件:
- 分割:2024_02_18T04_29_41.534828
- 路径:
**/details_harness|arc:challenge|25_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:2024_02_18T04_29_41.534828
- 数据文件:
-
配置名称:harness_gsm8k_5
- 数据文件:
- 分割:2024_02_18T04_29_41.534828
- 路径:
**/details_harness|gsm8k|5_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:2024_02_18T04_29_41.534828
- 数据文件:
-
配置名称:harness_hellaswag_10
- 数据文件:
- 分割:2024_02_18T04_29_41.534828
- 路径:
**/details_harness|hellaswag|10_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2024-02-18T04-29-41.534828.parquet
- 路径:
- 分割:2024_02_18T04_29_41.534828
- 数据文件:
-
配置名称:harness_hendrycksTest_5
- 数据文件:
- 分割:2024_02_18T04_29_41.534828
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-anatomy|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-astronomy|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_biology|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-college_physics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-computer_security|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-econometrics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-global_facts|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-human_aging|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-international_law|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-management|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-marketing|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-nutrition|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-philosophy|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-prehistory|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-02-18T04-29-41.534828.parquet**/details_harness|hendrycksTest-professional_law|5_2024-02-18T04-29-41.534828.parquet- `**/details_harness|hendrycksTest-professional_medicine|5_2024-02-1
- 路径:
- 分割:2024_02_18T04_29_41.534828
- 数据文件:



