遇见数据集

open-llm-leaderboard/details_jondurbin__airoboros-13b-gpt4

收藏
Hugging Face2023-10-23 更新2024-03-04 收录
官方服务:

资源简介:

该数据集是在评估模型jondurbin/airoboros-13b-gpt4时自动生成的,主要用于在Open LLM Leaderboard上进行评估。数据集包含64个配置,每个配置对应一个评估任务。数据集由2次运行生成,每次运行的结果存储为特定的split,split的名称使用运行的时间戳。train split始终指向最新的结果。此外,数据集还包含一个名为results的配置,用于存储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。

This dataset was automatically generated during the evaluation of the model jondurbin/airoboros-13b-gpt4, and is primarily designed for evaluation on the Open LLM Leaderboard. The dataset contains 64 configurations, each corresponding to one evaluation task. It is generated from two runs, with the results of each run stored as a dedicated split, where the name of the split is the timestamp of the corresponding run. The `train` split always points to the most recent evaluation results. Additionally, the dataset includes a configuration named `results` that stores the aggregated results across all runs, and is used to calculate and display the aggregated metrics on the Open LLM Leaderboard.

提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

该数据集是在模型 jondurbin/airoboros-13b-gpt4Open LLM Leaderboard 上的评估运行期间自动创建的。

数据集组成

  • 数据集包含 64 个配置,每个配置对应一个评估任务。
  • 数据集从 2 次运行中创建,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" 分割总是指向最新的结果。
  • 额外的配置 "results" 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_jondurbin__airoboros-13b-gpt4", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-23T04:00:44.911684 运行的最新结果: python { "all": { "em": 0.014681208053691275, "em_stderr": 0.001231711314310859, "f1": 0.07406564597315451, "f1_stderr": 0.0017844772735649754, "acc": 0.4182714775789221, "acc_stderr": 0.009732871523024014 }, "harness|drop|3": { "em": 0.014681208053691275, "em_stderr": 0.001231711314310859, "f1": 0.07406564597315451, "f1_stderr": 0.0017844772735649754 }, "harness|gsm8k|5": { "acc": 0.07884761182714177, "acc_stderr": 0.00742339051987324 }, "harness|winogrande|5": { "acc": 0.7576953433307024, "acc_stderr": 0.012042352526174787 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分割: 2023_08_18T14_07_58.585031
      • 路径: **/details_harness|arc:challenge|25_2023-08-18T14:07:58.585031.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-08-18T14:07:58.585031.parquet
  • harness_drop_3

    • 分割: 2023_10_23T04_00_44.911684
      • 路径: **/details_harness|drop|3_2023-10-23T04-00-44.911684.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-23T04-00-44.911684.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_23T04_00_44.911684
      • 路径: **/details_harness|gsm8k|5_2023-10-23T04-00-44.911684.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-10-23T04-00-44.911684.parquet
  • harness_hellaswag_10

    • 分割: 2023_08_18T14_07_58.585031
      • 路径: **/details_harness|hellaswag|10_2023-08-18T14:07:58.585031.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-08-18T14:07:58.585031.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_08_18T14_07_58.585031
      • 路径:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-management|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-08-18T14:07:58.585031.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-08-18T14:07:58.585031.parquet
二维码
社区交流群
二维码
科研交流群
商业服务