five

open-llm-leaderboard-old/details_chinoll__Yi-7b-dpo

收藏
Hugging Face2023-12-04 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_chinoll__Yi-7b-dpo
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chinoll/Yi-7b-dpo进行评估时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从一次或多次运行中生成的,每次运行都作为每个配置中的一个特定分割存储。train分割始终指向最新结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了如何使用Python中的datasets库加载数据集的示例。

该数据集是在Open LLM Leaderboard上对模型chinoll/Yi-7b-dpo进行评估时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从一次或多次运行中生成的,每次运行都作为每个配置中的一个特定分割存储。train分割始终指向最新结果。一个名为results的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了如何使用Python中的datasets库加载数据集的示例。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集摘要

该数据集是在对模型 chinoll/Yi-7b-dpo 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

数据集组成

数据集由 63 个配置组成,每个配置对应一个评估任务。数据集从 1 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 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_chinoll__Yi-7b-dpo", "harness_winogrande_5", split="train")

最新结果

这些是最新结果(来自 2023-12-04T16:10:38.355372 运行)的示例: python { "all": { "acc": 0.6274780891690785, "acc_stderr": 0.03214198982171106, "acc_norm": 0.6382309545732996, "acc_norm_stderr": 0.03286487964348697, "mc1": 0.3047735618115055, "mc1_stderr": 0.016114124156882455, "mc2": 0.4551491788416383, "mc2_stderr": 0.014826375266749701 }, "harness|arc:challenge|25": { "acc": 0.39505119453924914, "acc_stderr": 0.014285898292938172, "acc_norm": 0.4308873720136519, "acc_norm_stderr": 0.014471133392642475 }, "harness|hellaswag|10": { "acc": 0.5570603465445131, "acc_stderr": 0.004957182635381807, "acc_norm": 0.7452698665604461, "acc_norm_stderr": 0.004348189459336535 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695235, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695235 }, # 其他任务的结果... }

配置详情

  • config_name: harness_arc_challenge_25

    • data_files:
      • split: 2023_12_04T16_10_38.355372
        • path: [**/details_harness|arc:challenge|25_2023-12-04T16-10-38.355372.parquet]
      • split: latest
        • path: [**/details_harness|arc:challenge|25_2023-12-04T16-10-38.355372.parquet]
  • config_name: harness_gsm8k_5

    • data_files:
      • split: 2023_12_04T16_10_38.355372
        • path: [**/details_harness|gsm8k|5_2023-12-04T16-10-38.355372.parquet]
      • split: latest
        • path: [**/details_harness|gsm8k|5_2023-12-04T16-10-38.355372.parquet]
  • config_name: harness_hellaswag_10

    • data_files:
      • split: 2023_12_04T16_10_38.355372
        • path: [**/details_harness|hellaswag|10_2023-12-04T16-10-38.355372.parquet]
      • split: latest
        • path: [**/details_harness|hellaswag|10_2023-12-04T16-10-38.355372.parquet]
  • config_name: harness_hendrycksTest_5

    • data_files:
      • split: 2023_12_04T16_10_38.355372
        • path: [ **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-anatomy|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-astronomy|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-business_ethics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_biology|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_chemistry|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_computer_science|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_mathematics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_medicine|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_physics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-computer_security|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-econometrics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-formal_logic|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-global_facts|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-high_school_biology|5_2023-12-04T16-10-38.355372.parquet,

          其他文件路径...

        ]
      • split: latest
        • path: [ **/details_harness|hendrycksTest-abstract_algebra|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-anatomy|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-astronomy|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-business_ethics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-clinical_knowledge|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_biology|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_chemistry|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_computer_science|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_mathematics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_medicine|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-college_physics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-computer_security|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-conceptual_physics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-econometrics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-electrical_engineering|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-elementary_mathematics|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-formal_logic|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-global_facts|5_2023-12-04T16-10-38.355372.parquet, **/details_harness|hendrycksTest-high_school_biology|5_2023-12-04T16-10-38.355372.parquet,

          其他文件路径...

        ]
二维码
社区交流群
二维码
科研交流群
商业服务