five

open-llm-leaderboard/details_TheBloke__Llama-2-7B-GPTQ

收藏
Hugging Face2023-10-21 更新2024-03-04 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_TheBloke__Llama-2-7B-GPTQ
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TheBloke/Llama-2-7B-GPTQ时自动创建的,包含64个配置,每个配置对应一个评估任务。数据集由3次运行生成,每次运行的结果作为特定的分割存储,且train分割始终指向最新结果。此外,数据集还包含一个results配置,用于存储所有运行的聚合结果。

This dataset was automatically generated during the evaluation of the model TheBloke/Llama-2-7B-GPTQ. It contains 64 configurations, each corresponding to a single evaluation task. The dataset is constructed from three experimental runs, where the results of each run are stored as a dedicated data split, and the "train" split always points to the most recent run's results. Furthermore, the dataset includes a "results" configuration for storing the aggregated results across all runs.
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 TheBloke/Llama-2-7B-GPTQ 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3次运行创建,每次运行在每个配置中都有特定的分割,分割名称使用运行的时间戳。
  • 最新结果:"train" 分割始终指向最新的结果。
  • 结果汇总:一个额外的配置 "results" 存储所有运行的汇总结果,用于计算和显示 Open LLM Leaderboard 上的汇总指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_TheBloke__Llama-2-7B-GPTQ", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-21T20:13:14.412039 运行的最新结果: python { "all": { "em": 0.0007340604026845638, "em_stderr": 0.00027736144573356367, "f1": 0.054487206375839085, "f1_stderr": 0.001279202944739141, "acc": 0.38965983773074364, "acc_stderr": 0.009246673557602019 }, "harness|drop|3": { "em": 0.0007340604026845638, "em_stderr": 0.00027736144573356367, "f1": 0.054487206375839085, "f1_stderr": 0.001279202944739141 }, "harness|gsm8k|5": { "acc": 0.050037907505686124, "acc_stderr": 0.006005442354577735 }, "harness|winogrande|5": { "acc": 0.7292817679558011, "acc_stderr": 0.012487904760626303 } }

配置详情

以下是部分配置及其数据文件路径:

harness_arc_challenge_25

  • 2023_08_29T12_13_30.420278
    • **/details_harness|arc:challenge|25_2023-08-29T12:13:30.420278.parquet
  • 2023_08_30T09_33_50.119005
    • **/details_harness|arc:challenge|25_2023-08-30T09:33:50.119005.parquet
  • latest
    • **/details_harness|arc:challenge|25_2023-08-30T09:33:50.119005.parquet

harness_drop_3

  • 2023_10_21T20_13_14.412039
    • **/details_harness|drop|3_2023-10-21T20-13-14.412039.parquet
  • latest
    • **/details_harness|drop|3_2023-10-21T20-13-14.412039.parquet

harness_gsm8k_5

  • 2023_10_21T20_13_14.412039
    • **/details_harness|gsm8k|5_2023-10-21T20-13-14.412039.parquet
  • latest
    • **/details_harness|gsm8k|5_2023-10-21T20-13-14.412039.parquet

harness_hellaswag_10

  • 2023_08_29T12_13_30.420278
    • **/details_harness|hellaswag|10_2023-08-29T12:13:30.420278.parquet
  • 2023_08_30T09_33_50.119005
    • **/details_harness|hellaswag|10_2023-08-30T09:33:50.119005.parquet
  • latest
    • **/details_harness|hellaswag|10_2023-08-30T09:33:50.119005.parquet

harness_hendrycksTest_5

  • 2023_08_29T12_13_30.420278
    • **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-anatomy|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-astronomy|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-business_ethics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_biology|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_chemistry|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_computer_science|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_mathematics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_medicine|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-college_physics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-computer_security|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-econometrics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-formal_logic|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-global_facts|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_biology|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_geography|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_physics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-human_aging|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-human_sexuality|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-international_law|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-jurisprudence|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-machine_learning|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-management|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-marketing|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-medical_genetics|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-miscellaneous|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-moral_disputes|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-nutrition|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-philosophy|5_2023-08-29T12:13:30.420278.parquet
    • **/details_harness|hendrycksTest-prehistory|5_2023-08-29T12:13:30.420278.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