five

open-llm-leaderboard/details_databricks__dolly-v2-7b

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_databricks__dolly-v2-7b
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在评估模型databricks/dolly-v2-7b时自动创建的,主要用于Open LLM Leaderboard的评估任务。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分割,分割名称使用运行的时间戳。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 created during the evaluation of the model databricks/dolly-v2-7b, and is primarily intended for evaluation tasks on the Open LLM Leaderboard. The dataset comprises 64 configurations, each corresponding to one individual evaluation task. It is generated from two separate runs, where the results of each run are stored as a dedicated dataset split, with the split name being the timestamp of the corresponding run. Additionally, the dataset includes a configuration named "results", which is used to store the aggregated results across all runs, as well as to compute 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

数据集简介

该数据集是在对模型 databricks/dolly-v2-7b 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

数据集结构

  • 配置数量:64个配置,每个配置对应一个评估任务。
  • 数据来源:数据集由2次运行结果组成,每次运行的结果可以在每个配置中找到,以运行的时间戳命名的特定分片形式存在。
  • 最新结果:"train" 分片始终指向最新的结果。
  • 聚合结果: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_databricks__dolly-v2-7b", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-15T13:27:34.576106 运行的最新结果: python { "all": { "em": 0.0018875838926174498, "em_stderr": 0.00044451099905589976, "f1": 0.059697986577181554, "f1_stderr": 0.0013648879248414308, "acc": 0.3060018322459733, "acc_stderr": 0.008342799872753168 }, "harness|drop|3": { "em": 0.0018875838926174498, "em_stderr": 0.00044451099905589976, "f1": 0.059697986577181554, "f1_stderr": 0.0013648879248414308 }, "harness|gsm8k|5": { "acc": 0.011372251705837756, "acc_stderr": 0.002920666198788728 }, "harness|winogrande|5": { "acc": 0.6006314127861089, "acc_stderr": 0.013764933546717607 } }

配置详情

  • harness_arc_challenge_25

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|arc:challenge|25_2023-07-18T11:46:56.588473.parquet
  • harness_drop_3

    • 分片:2023_10_15T13_27_34.576106, latest
    • 路径:**/details_harness|drop|3_2023-10-15T13-27-34.576106.parquet
  • harness_gsm8k_5

    • 分片:2023_10_15T13_27_34.576106, latest
    • 路径:**/details_harness|gsm8k|5_2023-10-15T13-27-34.576106.parquet
  • harness_hellaswag_10

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hellaswag|10_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:多个文件,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_anatomy_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_astronomy_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_biology_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_college_physics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_computer_security_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_econometrics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-18T11:46:56.588473.parquet
  • harness_hendrycksTest_elementary_mathematics_5

    • 分片:2023_07_18T11_46_56.588473, latest
    • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-18T11:46:56.588473.parquet
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