five

open-llm-leaderboard/details_tricktreat__Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora

收藏
Hugging Face2024-04-16 更新2024-06-12 收录
下载链接:
https://hf-mirror.com/datasets/open-llm-leaderboard/details_tricktreat__Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora
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tricktreat/Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora时自动创建的,评估过程在Open LLM Leaderboard上进行。数据集由63个配置组成,每个配置对应一个评估任务。数据集从1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,分割以运行的时间戳命名。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上的聚合指标。

该数据集是在评估模型tricktreat/Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora时自动创建的,评估过程在Open LLM Leaderboard上进行。数据集由63个配置组成,每个配置对应一个评估任务。数据集从1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,分割以运行的时间戳命名。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上的聚合指标。
提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型 tricktreat/Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora 时自动创建的,用于 Open LLM Leaderboard 的评估。

数据集结构

  • 配置数量:63个配置,每个配置对应一个评估任务。
  • 创建来源:数据集由1次运行创建,每个运行在每个配置中作为一个特定的分割存在,分割名称使用运行的时间戳。
  • 最新结果:"train" 分割始终指向最新结果。
  • 汇总结果: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 上的聚合指标。

数据加载示例

python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_tricktreat__Llama-2-7b-chat-hf-guanaco-freeze-embed-tokens-q-v-proj-lora", "harness_winogrande_5", split="train")

最新结果

以下是 2024-04-16T04:38:13.243008 运行的最新结果

python { "all": { "acc": 0.46892131898965095, "acc_stderr": 0.03444062283174299, "acc_norm": 0.47499636851082255, "acc_norm_stderr": 0.03522696874836134, "mc1": 0.2876376988984088, "mc1_stderr": 0.015846315101394816, "mc2": 0.4250572562925463, "mc2_stderr": 0.015188747225079806 }, "harness|arc:challenge|25": { "acc": 0.4812286689419795, "acc_stderr": 0.014601090150633964, "acc_norm": 0.515358361774744, "acc_norm_stderr": 0.014604496129394913 }, "harness|hellaswag|10": { "acc": 0.5830511850229038, "acc_stderr": 0.00492046593606861, "acc_norm": 0.7651862178848835, "acc_norm_stderr": 0.004230160814469385 }, # 其他任务结果省略 }

配置详情

  • 配置名称:harness_arc_challenge_25

    • 数据文件
      • 分割:2024_04_16T04_38_13.243008
        • 路径:**/details_harness|arc:challenge|25_2024-04-16T04-38-13.243008.parquet
      • 分割:latest
        • 路径:**/details_harness|arc:challenge|25_2024-04-16T04-38-13.243008.parquet
  • 配置名称:harness_gsm8k_5

    • 数据文件
      • 分割:2024_04_16T04_38_13.243008
        • 路径:**/details_harness|gsm8k|5_2024-04-16T04-38-13.243008.parquet
      • 分割:latest
        • 路径:**/details_harness|gsm8k|5_2024-04-16T04-38-13.243008.parquet
  • 配置名称:harness_hellaswag_10

    • 数据文件
      • 分割:2024_04_16T04_38_13.243008
        • 路径:**/details_harness|hellaswag|10_2024-04-16T04-38-13.243008.parquet
      • 分割:latest
        • 路径:**/details_harness|hellaswag|10_2024-04-16T04-38-13.243008.parquet
  • 配置名称:harness_hendrycksTest_5

    • 数据文件
      • 分割:2024_04_16T04_38_13.243008
        • 路径:
          • **/details_harness|hendrycksTest-abstract_algebra|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-anatomy|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-astronomy|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-business_ethics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-clinical_knowledge|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_biology|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_chemistry|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_computer_science|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_mathematics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_medicine|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-college_physics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-computer_security|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-conceptual_physics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-econometrics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-electrical_engineering|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-elementary_mathematics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-formal_logic|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-global_facts|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_biology|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_chemistry|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_computer_science|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_european_history|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_geography|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_mathematics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_microeconomics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_physics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_psychology|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_statistics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_us_history|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-high_school_world_history|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-human_aging|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-human_sexuality|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-international_law|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-jurisprudence|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-logical_fallacies|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-machine_learning|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-management|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-marketing|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-medical_genetics|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-miscellaneous|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-moral_disputes|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-moral_scenarios|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-nutrition|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-philosophy|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-prehistory|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-professional_accounting|5_2024-04-16T04-38-13.243008.parquet
          • **/details_harness|hendrycksTest-professional_law|5_2024-04-16T04-38-13.243008.parquet
          • `**/details_harness|hendrycksTest-professional
二维码
社区交流群
二维码
科研交流群
商业服务