open-llm-leaderboard-old/details_Gille__StrangeMerges_17-7B-dare_ties
收藏Hugging Face2024-04-02 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_Gille__StrangeMerges_17-7B-dare_ties
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在模型Gille/StrangeMerges_17-7B-dare_ties在Open LLM Leaderboard上的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。数据集由2次运行生成,每次运行可以在每个配置中找到特定的分割,分割以运行的时间戳命名。train分割始终指向最新的结果。此外,一个名为results的配置存储了所有运行的聚合结果,并用于计算和显示Open LLM Leaderboard上的聚合指标。
该数据集是在模型Gille/StrangeMerges_17-7B-dare_ties在Open LLM Leaderboard上的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。数据集由2次运行生成,每次运行可以在每个配置中找到特定的分割,分割以运行的时间戳命名。train分割始终指向最新的结果。此外,一个名为results的配置存储了所有运行的聚合结果,并用于计算和显示Open LLM Leaderboard上的聚合指标。
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Gille/StrangeMerges_17-7B-dare_ties在Open LLM Leaderboard上的评估运行期间自动创建的。
数据集结构
- 配置数量:数据集包含63个配置,每个配置对应一个评估任务。
- 运行次数:数据集从2次运行中创建。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。
- 训练分割:"train"分割始终指向最新的结果。
- 结果配置:一个额外的配置"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_Gille__StrangeMerges_17-7B-dare_ties", "harness_winogrande_5", split="train")
最新结果
- 最新结果来源:这些是最新的结果,来自2024-04-02T23:51:42.907496的运行。
- 结果示例: python { "all": { "acc": 0.6530806284796924, "acc_stderr": 0.03189301665466623, "acc_norm": 0.6544880575623332, "acc_norm_stderr": 0.0325393477652265, "mc1": 0.36107711138310894, "mc1_stderr": 0.016814312844836886, "mc2": 0.5318128401717307, "mc2_stderr": 0.01495332489726401 }, "harness|arc:challenge|25": { "acc": 0.621160409556314, "acc_stderr": 0.014175915490000324, "acc_norm": 0.6663822525597269, "acc_norm_stderr": 0.013778687054176536 }, "harness|hellaswag|10": { "acc": 0.665803624775941, "acc_stderr": 0.0047074472442006216, "acc_norm": 0.8603863772156941, "acc_norm_stderr": 0.003458773934719553 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.37, "acc_stderr": 0.048523658709391, "acc_norm": 0.37, "acc_norm_stderr": 0.048523658709391 }, # 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_04_02T23_46_01.083918
- 路径:
**/details_harness|arc:challenge|25_2024-04-02T23-46-01.083918.parquet
- 路径:
- 分割:2024_04_02T23_51_42.907496
- 路径:
**/details_harness|arc:challenge|25_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:2024_04_02T23_46_01.083918
- 数据文件:
-
配置名称:harness_gsm8k_5
- 数据文件:
- 分割:2024_04_02T23_46_01.083918
- 路径:
**/details_harness|gsm8k|5_2024-04-02T23-46-01.083918.parquet
- 路径:
- 分割:2024_04_02T23_51_42.907496
- 路径:
**/details_harness|gsm8k|5_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:2024_04_02T23_46_01.083918
- 数据文件:
-
配置名称:harness_hellaswag_10
- 数据文件:
- 分割:2024_04_02T23_46_01.083918
- 路径:
**/details_harness|hellaswag|10_2024-04-02T23-46-01.083918.parquet
- 路径:
- 分割:2024_04_02T23_51_42.907496
- 路径:
**/details_harness|hellaswag|10_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2024-04-02T23-51-42.907496.parquet
- 路径:
- 分割:2024_04_02T23_46_01.083918
- 数据文件:
-
配置名称:harness_hendrycksTest_5
- 数据文件:
- 分割:2024_04_02T23_46_01.083918
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-anatomy|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-astronomy|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_biology|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-college_physics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-computer_security|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-econometrics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-global_facts|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-human_aging|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-international_law|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-management|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-marketing|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-04-02T23-46-01.083918.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-04-02T23-46-01.083918.parquet- `**/details_harness|hendrycksTest-moral_scenarios|5_2024-04-02T23-46-01.083918.par
- 路径:
- 分割:2024_04_02T23_46_01.083918
- 数据文件:



