five

open-llm-leaderboard-old/details_uukuguy__speechless-llama2-hermes-orca-platypus-wizardlm-13b

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_uukuguy__speechless-llama2-hermes-orca-platypus-wizardlm-13b
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[uukuguy/speechless-llama2-hermes-orca-platypus-wizardlm-13b](https://huggingface.co/uukuguy/speechless-llama2-hermes-orca-platypus-wizardlm-13b)时自动生成的,评估是在[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上进行的。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从3次运行中创建的,每次运行都可以在特定配置中找到,分割名称为运行的时间戳。train分割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,并用于计算和显示[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上的聚合指标。

该数据集是在评估模型[uukuguy/speechless-llama2-hermes-orca-platypus-wizardlm-13b](https://huggingface.co/uukuguy/speechless-llama2-hermes-orca-platypus-wizardlm-13b)时自动生成的,评估是在[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上进行的。数据集由64个配置组成,每个配置对应一个评估任务。数据集是从3次运行中创建的,每次运行都可以在特定配置中找到,分割名称为运行的时间戳。train分割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,并用于计算和显示[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上的聚合指标。
提供机构:
open-llm-leaderboard-old
原始信息汇总

数据集概述

数据集创建

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 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,用于计算和显示聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_uukuguy__speechless-llama2-hermes-orca-platypus-wizardlm-13b", "harness_winogrande_5", split="train")

最新结果

  • 最新结果来自 2023-10-15T13:11:43.680043 运行,包含以下指标: python { "all": { "em": 0.057466442953020135, "em_stderr": 0.0023833905882384896, "f1": 0.17808829697986514, "f1_stderr": 0.002972308703760267, "acc": 0.44245449154575855, "acc_stderr": 0.010703432271512695 }, "harness|drop|3": { "em": 0.057466442953020135, "em_stderr": 0.0023833905882384896, "f1": 0.17808829697986514, "f1_stderr": 0.002972308703760267 }, "harness|gsm8k|5": { "acc": 0.13115996967399546, "acc_stderr": 0.009298499235587858 }, "harness|winogrande|5": { "acc": 0.7537490134175217, "acc_stderr": 0.012108365307437531 } }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|arc:challenge|25_2023-09-02T00:07:11.850382.parquet
  • harness_drop_3

    • 分割:2023_10_15T13_11_43.680043, latest
    • 路径:**/details_harness|drop|3_2023-10-15T13-11-43.680043.parquet
  • harness_gsm8k_5

    • 分割:2023_10_15T13_11_43.680043, latest
    • 路径:**/details_harness|gsm8k|5_2023-10-15T13-11-43.680043.parquet
  • harness_hellaswag_10

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hellaswag|10_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:多个路径,例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_abstract_algebra_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_anatomy_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_astronomy_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_business_ethics_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_clinical_knowledge_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_biology_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_chemistry_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_computer_science_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_mathematics_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_medicine_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_college_physics_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_computer_security_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_conceptual_physics_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_econometrics_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-09-02T00:07:11.850382.parquet
  • harness_hendrycksTest_electrical_engineering_5

    • 分割:2023_09_02T00_07_11.850382, latest
    • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-02T00:07:11.850382.parquet

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

二维码
社区交流群
二维码
科研交流群
商业服务