five

open-llm-leaderboard-old/details_openlm-research__open_llama_3b_v2

收藏
Hugging Face2023-10-15 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_openlm-research__open_llama_3b_v2
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 openlm-research/open_llama_3b_v2 在 Open LLM Leaderboard 上进行评估时自动生成的。数据集由 64 个配置组成,每个配置对应一个特定的评估任务。数据集是从 2 次运行中创建的,每次运行在每个配置中表示为特定的分割,分割名称由运行的时间戳命名。train 分割始终指向最新的结果。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 代码加载运行中的详细信息的示例,并包含了特定运行的最新结果。

该数据集是在模型 openlm-research/open_llama_3b_v2 在 Open LLM Leaderboard 上进行评估时自动生成的。数据集由 64 个配置组成,每个配置对应一个特定的评估任务。数据集是从 2 次运行中创建的,每次运行在每个配置中表示为特定的分割,分割名称由运行的时间戳命名。train 分割始终指向最新的结果。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 代码加载运行中的详细信息的示例,并包含了特定运行的最新结果。
提供机构:
open-llm-leaderboard-old
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openlm-research/open_llama_3b_v2Open LLM Leaderboard上的自动创建的。数据集包含64个配置,每个配置对应一个评估任务。

数据集结构

数据集由2次运行创建,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。

数据集加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_openlm-research__open_llama_3b_v2", "harness_winogrande_5", split="train")

最新结果

以下是2023-10-15T11:22:56.677003运行的最新结果: python { "all": { "em": 0.001153523489932886, "em_stderr": 0.0003476179896857095, "f1": 0.05134962248322172, "f1_stderr": 0.0012730168443049574, "acc": 0.3395923103113801, "acc_stderr": 0.007914879526646601 }, "harness|drop|3": { "em": 0.001153523489932886, "em_stderr": 0.0003476179896857095, "f1": 0.05134962248322172, "f1_stderr": 0.0012730168443049574 }, "harness|gsm8k|5": { "acc": 0.009097801364670205, "acc_stderr": 0.002615326510775673 }, "harness|winogrande|5": { "acc": 0.67008681925809, "acc_stderr": 0.013214432542517527 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分割: 2023_07_24T10_28_09.665576
      • 路径: **/details_harness|arc:challenge|25_2023-07-24T10:28:09.665576.parquet
    • 分割: latest
      • 路径: **/details_harness|arc:challenge|25_2023-07-24T10:28:09.665576.parquet
  • harness_drop_3

    • 分割: 2023_10_15T11_22_56.677003
      • 路径: **/details_harness|drop|3_2023-10-15T11-22-56.677003.parquet
    • 分割: latest
      • 路径: **/details_harness|drop|3_2023-10-15T11-22-56.677003.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_15T11_22_56.677003
      • 路径: **/details_harness|gsm8k|5_2023-10-15T11-22-56.677003.parquet
    • 分割: latest
      • 路径: **/details_harness|gsm8k|5_2023-10-15T11-22-56.677003.parquet
  • harness_hellaswag_10

    • 分割: 2023_07_24T10_28_09.665576
      • 路径: **/details_harness|hellaswag|10_2023-07-24T10:28:09.665576.parquet
    • 分割: latest
      • 路径: **/details_harness|hellaswag|10_2023-07-24T10:28:09.665576.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_07_24T10_28_09.665576
      • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-econometrics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-formal_logic|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-global_facts|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_biology|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_geography|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_physics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_psychology|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_statistics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_us_history|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-high_school_world_history|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-human_aging|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-human_sexuality|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-international_law|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-jurisprudence|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-logical_fallacies|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-machine_learning|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-management|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-marketing|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-medical_genetics|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-miscellaneous|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-moral_disputes|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-moral_scenarios|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-nutrition|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-philosophy|5_2023-07-24T10:28:09.665576.parquet
      • 路径: **/details_harness|hendrycksTest-prehistory|5_2023-07-24T10:28:09.665576.parquet
      • 路径: `**/details_harness|hendrycksTest-professional_accounting|5_2023-07-24T10
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