five

open-llm-leaderboard-old/details_maximuslee07__llama-2-13b-rockwellautomation

收藏
Hugging Face2024-01-04 更新2024-06-22 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard-old/details_maximuslee07__llama-2-13b-rockwellautomation
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[maximuslee07/llama-2-13b-rockwellautomation](https://huggingface.co/maximuslee07/llama-2-13b-rockwellautomation)在[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上的表现时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行都可以在特定配置中找到,分割名称使用运行的时间戳。"train"分割始终指向最新的结果。此外,"results"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[maximuslee07/llama-2-13b-rockwellautomation](https://huggingface.co/maximuslee07/llama-2-13b-rockwellautomation)在[Open LLM Leaderboard](https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard)上的表现时自动创建的。数据集由63个配置组成,每个配置对应一个评估任务。数据集是从1次运行中创建的,每次运行都可以在特定配置中找到,分割名称使用运行的时间戳。"train"分割始终指向最新的结果。此外,"results"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

该数据集是在对模型 maximuslee07/llama-2-13b-rockwellautomation 进行评估运行期间自动创建的。数据集包含 63 个配置,每个配置对应一个评估任务。数据集从 1 次运行中创建,每个运行可以在每个配置中找到一个特定的拆分,拆分名称使用运行的 timestamp。"train" 拆分始终指向最新结果。

数据集结构

数据集包含多个配置,每个配置对应不同的评估任务。以下是部分配置的详细信息:

  • harness_arc_challenge_25

    • 拆分:2024_01_04T15_04_41.345460
    • 路径:**/details_harness|arc:challenge|25_2024-01-04T15-04-41.345460.parquet
    • 拆分:latest
    • 路径:**/details_harness|arc:challenge|25_2024-01-04T15-04-41.345460.parquet
  • harness_gsm8k_5

    • 拆分:2024_01_04T15_04_41.345460
    • 路径:**/details_harness|gsm8k|5_2024-01-04T15-04-41.345460.parquet
    • 拆分:latest
    • 路径:**/details_harness|gsm8k|5_2024-01-04T15-04-41.345460.parquet
  • harness_hellaswag_10

    • 拆分:2024_01_04T15_04_41.345460
    • 路径:**/details_harness|hellaswag|10_2024-01-04T15-04-41.345460.parquet
    • 拆分:latest
    • 路径:**/details_harness|hellaswag|10_2024-01-04T15-04-41.345460.parquet
  • harness_hendrycksTest_5

    • 拆分:2024_01_04T15_04_41.345460
    • 路径:
      • **/details_harness|hendrycksTest-abstract_algebra|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-anatomy|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-astronomy|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-business_ethics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_biology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_chemistry|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_computer_science|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_mathematics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_medicine|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-college_physics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-computer_security|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-conceptual_physics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-econometrics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-electrical_engineering|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-formal_logic|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-global_facts|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_biology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_european_history|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_geography|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_physics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_psychology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_statistics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_us_history|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-high_school_world_history|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-human_aging|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-human_sexuality|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-international_law|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-jurisprudence|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-logical_fallacies|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-machine_learning|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-management|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-marketing|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-medical_genetics|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-miscellaneous|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-moral_disputes|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-moral_scenarios|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-nutrition|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-philosophy|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-prehistory|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-professional_accounting|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-professional_law|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-professional_medicine|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-professional_psychology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-public_relations|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-security_studies|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-sociology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-us_foreign_policy|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-virology|5_2024-01-04T15-04-41.345460.parquet
      • **/details_harness|hendrycksTest-world_religions|5_2024-01-04T15-04-41.345460.parquet

最新结果

以下是 2024-01-04T15:04:41.345460 运行的最新结果:

python { "all": { "acc": 0.2511081603813156, "acc_stderr": 0.030845430828452477, "acc_norm": 0.2520379726970298, "acc_norm_stderr": 0.031666273642418356, "mc1": 0.24357405140758873, "mc1_stderr": 0.015026354824910782, "mc2": NaN, "mc2_stderr": NaN }, "harness|arc:challenge|25": { "acc": 0.23378839590443687, "acc_stderr": 0.012368225378507148, "acc_norm": 0.2815699658703072, "acc_norm_stderr": 0.013143376735009015 }, "harness|hellaswag|10": { "acc": 0.2548297151961761, "acc_stderr": 0.004348748730529937, "acc_norm": 0.2577175861

二维码
社区交流群
二维码
科研交流群
商业服务