five

open-llm-leaderboard-old/details_Writer__palmyra-base

收藏
Hugging Face2023-10-17 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_Writer__palmyra-base
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Writer/palmyra-base进行评估运行时自动创建的。数据集由122个配置组成,每个配置对应一个评估任务。数据集由3次运行生成,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。此外,还有一个results配置,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。README文件还提供了如何使用`datasets`库中的`load_dataset`函数加载数据集的示例,并提供了特定运行的最新结果。

该数据集是在Open LLM Leaderboard上对模型Writer/palmyra-base进行评估运行时自动创建的。数据集由122个配置组成,每个配置对应一个评估任务。数据集由3次运行生成,每次运行在每个配置中表示为特定的分割。train分割始终指向最新的结果。此外,还有一个results配置,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。README文件还提供了如何使用`datasets`库中的`load_dataset`函数加载数据集的示例,并提供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

该数据集是在对模型 Writer/palmyra-base 进行评估运行期间自动创建的,评估结果展示在 Open LLM Leaderboard 上。

数据集组成

  • 数据集包含 122 个配置,每个配置对应一个评估任务。
  • 数据集从 3 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_Writer__palmyra-base", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-17T01:27:06.940630 运行 的最新结果:

python { "all": { "em": 0.0010486577181208054, "em_stderr": 0.0003314581465219252, "f1": 0.04800964765100684, "f1_stderr": 0.0011968648184797989, "acc": 0.29537785734929894, "acc_stderr": 0.00829420088462589 }, "harness|drop|3": { "em": 0.0010486577181208054, "em_stderr": 0.0003314581465219252, "f1": 0.04800964765100684, "f1_stderr": 0.0011968648184797989 }, "harness|gsm8k|5": { "acc": 0.009855951478392721, "acc_stderr": 0.0027210765770416625 }, "harness|winogrande|5": { "acc": 0.5808997632202052, "acc_stderr": 0.013867325192210116 } }

配置详情

  • harness_arc_challenge_25

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|arc:challenge|25_2023-07-19T12:49:48.066230.parquet
  • harness_drop_3

    • 分割: 2023_10_17T01_27_06.940630, latest
    • 路径: **/details_harness|drop|3_2023-10-17T01-27-06.940630.parquet
  • harness_gsm8k_5

    • 分割: 2023_10_17T01_27_06.940630, latest
    • 路径: **/details_harness|gsm8k|5_2023-10-17T01-27-06.940630.parquet
  • harness_hellaswag_10

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hellaswag|10_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: 多个文件路径,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-anatomy|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-astronomy|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-business_ethics|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_biology|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_chemistry|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_computer_science|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_mathematics|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_medicine|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-college_physics|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-computer_security|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-econometrics|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-19T12:49:48.066230.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分割: 2023_07_19T12_49_48.066230, latest
    • 路径: **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-19T12:49:48.066230.parquet

以上是数据集的详细概述,包括数据集的组成、加载示例、最新结果以及各个配置的详细信息。

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