five

open-llm-leaderboard/details_facebook__xglm-564M

收藏
Hugging Face2023-10-15 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_facebook__xglm-564M
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facebook/xglm-564M模型进行评估时自动生成的。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,每次运行的结果作为特定配置中的一个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,还有一个名为results的配置,存储了所有运行的聚合结果,并用于计算和展示Open LLM Leaderboard上的聚合指标。

This dataset was automatically generated during the evaluation of the facebook/xglm-564M model on the Open LLM Leaderboard. The dataset consists of 64 configurations, each corresponding to one evaluation task. The dataset is generated from two runs, with the results of each run serving as a split under the corresponding configuration, where the name of each split is the timestamp of the run. The `train` split always points to the most recent evaluation results. In addition, there is a configuration named `results` that stores the aggregated results from all runs and is used to calculate and display the aggregate metrics on the Open LLM Leaderboard.
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 facebook/xglm-564MOpen LLM Leaderboard 上的运行过程中自动创建的。

数据集结构

  • 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作为一个特定的分片找到,分片名称使用运行的时间戳。
  • "train" 分片始终指向最新的结果。
  • 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_facebook__xglm-564M", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-15T23:39:39.394377 运行的最新结果: python { "all": { "em": 0.013422818791946308, "em_stderr": 0.0011784931108563684, "f1": 0.060359689597315525, "f1_stderr": 0.0017160396766447692, "acc": 0.2623842654231489, "acc_stderr": 0.007675207819463649 }, "harness|drop|3": { "em": 0.013422818791946308, "em_stderr": 0.0011784931108563684, "f1": 0.060359689597315525, "f1_stderr": 0.0017160396766447692 }, "harness|gsm8k|5": { "acc": 0.002274450341167551, "acc_stderr": 0.001312157814867416 }, "harness|winogrande|5": { "acc": 0.5224940805051302, "acc_stderr": 0.014038257824059881 } }

配置详情

以下是部分配置及其数据文件路径:

  • config_name: harness_arc_challenge_25

    • split: 2023_07_19T14_24_31.422133
      • path: **/details_harness|arc:challenge|25_2023-07-19T14:24:31.422133.parquet
    • split: latest
      • path: **/details_harness|arc:challenge|25_2023-07-19T14:24:31.422133.parquet
  • config_name: harness_drop_3

    • split: 2023_10_15T23_39_39.394377
      • path: **/details_harness|drop|3_2023-10-15T23-39-39.394377.parquet
    • split: latest
      • path: **/details_harness|drop|3_2023-10-15T23-39-39.394377.parquet
  • config_name: harness_gsm8k_5

    • split: 2023_10_15T23_39_39.394377
      • path: **/details_harness|gsm8k|5_2023-10-15T23-39-39.394377.parquet
    • split: latest
      • path: **/details_harness|gsm8k|5_2023-10-15T23-39-39.394377.parquet
  • config_name: harness_hellaswag_10

    • split: 2023_07_19T14_24_31.422133
      • path: **/details_harness|hellaswag|10_2023-07-19T14:24:31.422133.parquet
    • split: latest
      • path: **/details_harness|hellaswag|10_2023-07-19T14:24:31.422133.parquet
  • config_name: harness_hendrycksTest_5

    • split: 2023_07_19T14_24_31.422133
      • path:
        • **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-anatomy|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-astronomy|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_biology|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-college_physics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-computer_security|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-econometrics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-formal_logic|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-global_facts|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_biology|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_geography|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_physics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-human_aging|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-human_sexuality|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-international_law|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-jurisprudence|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-machine_learning|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-management|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-marketing|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-medical_genetics|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-miscellaneous|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-moral_disputes|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-nutrition|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-philosophy|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-prehistory|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-professional_accounting|5_2023-07-19T14:24:31.422133.parquet
        • **/details_harness|hendrycksTest-professional_law|5_2023-07-19T14:24:31.4221
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作