five

open-llm-leaderboard/details_TheBloke__Platypus-30B-SuperHOT-8K-fp16

收藏
Hugging Face2023-08-27 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_TheBloke__Platypus-30B-SuperHOT-8K-fp16
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TheBloke/Platypus-30B-SuperHOT-8K-fp16的评估运行期间自动创建的,用于在Open LLM Leaderboard上进行评估。数据集由60个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行可以在每个配置中找到特定的分割,分割名称使用运行的时间戳。"train"分割始终指向最新的结果。此外,"results"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
原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型 TheBloke/Platypus-30B-SuperHOT-8K-fp16Open LLM Leaderboard 上的自动创建的。数据集包含60个配置,每个配置对应一个评估任务。

数据集结构

数据集由1次运行创建,每个运行可以在每个配置中找到特定的分割,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

一个额外的配置 "results" 存储了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_TheBloke__Platypus-30B-SuperHOT-8K-fp16", "harness_truthfulqa_mc_0", split="train")

最新结果

最新结果来自2023-08-18T16:25:34.320244运行,包含多个任务的评估结果。以下是部分结果示例: python { "all": { "acc": 0.23647488823331855, "acc_stderr": 0.030908567573023033, "acc_norm": 0.23771978116158754, "acc_norm_stderr": 0.030923042741200276, "mc1": 0.2178702570379437, "mc1_stderr": 0.014450846714123892, "mc2": 0.471292004765754, "mc2_stderr": 0.01664156844910162 }, "harness|arc:challenge|25": { "acc": 0.21843003412969283, "acc_stderr": 0.012074291605700987, "acc_norm": 0.2568259385665529, "acc_norm_stderr": 0.0127669237941168 }, "harness|hellaswag|10": { "acc": 0.2731527584146584, "acc_stderr": 0.004446680081493746, "acc_norm": 0.3082055367456682, "acc_norm_stderr": 0.004608082815535489 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.22, "acc_stderr": 0.04163331998932268, "acc_norm": 0.22, "acc_norm_stderr": 0.04163331998932268 }, "harness|truthfulqa:mc|0": { "mc1": 0.2178702570379437, "mc1_stderr": 0.014450846714123892, "mc2": 0.471292004765754, "mc2_stderr": 0.01664156844910162 } }

配置详情

  • config_name: harness_arc_challenge_25

    • 数据文件:
      • split: 2023_08_18T16_25_34.320244
        • path: **/details_harness|arc:challenge|25_2023-08-18T16:25:34.320244.parquet
      • split: latest
        • path: **/details_harness|arc:challenge|25_2023-08-18T16:25:34.320244.parquet
  • config_name: harness_hellaswag_10

    • 数据文件:
      • split: 2023_08_18T16_25_34.320244
        • path: **/details_harness|hellaswag|10_2023-08-18T16:25:34.320244.parquet
      • split: latest
        • path: **/details_harness|hellaswag|10_2023-08-18T16:25:34.320244.parquet
  • config_name: harness_hendrycksTest_5

    • 数据文件:
      • split: 2023_08_18T16_25_34.320244
        • path:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-international_law|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-management|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-marketing|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2023-08-18T16:25:34.320244.parquet
          • **/details_harness|hendrycksTest-professional_medicine|5_2023-08-18T16:25:34.320244.parquet
二维码
社区交流群
二维码
科研交流群
商业服务