five

open-llm-leaderboard/details_openlm-research__open_llama_13b

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_openlm-research__open_llama_13b
下载链接
链接失效反馈
官方服务:
资源简介:
该数据集是在模型 openlm-research/open_llama_13b 在 Open LLM Leaderboard 上的评估运行期间自动生成的。数据集由 122 个配置组成,每个配置对应一个评估任务。数据集是从 5 次运行中创建的,每次运行在每个配置中表示为特定的分割,分割名称使用运行的时间戳。train 分割始终指向最新的结果。一个名为 results 的额外配置存储了所有运行的聚合结果,这些结果用于计算和显示 Open LLM Leaderboard 上的聚合指标。README 还提供了如何使用 Python 代码加载运行中的详细信息的示例,并包含了特定运行的最新结果。

This dataset was automatically generated during the evaluation run of the model openlm-research/open_llama_13b on the Open LLM Leaderboard. The dataset consists of 122 configurations, each corresponding to one evaluation task. This dataset is compiled from 5 separate runs, where each run is represented as a dedicated split under each configuration, with split names using the run’s timestamp. The `train` split always points to the most recent results. An additional configuration named `results` stores the aggregated results across all runs, which are used to calculate and display the aggregate metrics on the Open LLM Leaderboard. The README also provides examples of how to load detailed run information using Python code, and includes the most recent results for specific 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 openlm-research/open_llama_13b 进行评估运行期间自动创建的,用于 Open LLM Leaderboard

数据集组成

  • 数据集包含 122 个配置,每个配置对应一个评估任务。
  • 数据集从 5 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。
  • 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 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_openlm-research__open_llama_13b", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-15T17:56:49.237621 运行的最新结果: python { "all": { "em": 0.0012583892617449664, "em_stderr": 0.00036305608931192504, "f1": 0.059914010067114144, "f1_stderr": 0.0014216209947324643, "acc": 0.3766001485184359, "acc_stderr": 0.008751247780672124 }, "harness|drop|3": { "em": 0.0012583892617449664, "em_stderr": 0.00036305608931192504, "f1": 0.059914010067114144, "f1_stderr": 0.0014216209947324643 }, "harness|gsm8k|5": { "acc": 0.032600454890068235, "acc_stderr": 0.004891669021939565 }, "harness|winogrande|5": { "acc": 0.7205998421468035, "acc_stderr": 0.012610826539404684 } }

配置详情

  • harness_arc_challenge_25

    • 分割:2023_07_20T10_02_18.035133
      • 路径:**/details_harness|arc:challenge|25_2023-07-20T10:02:18.035133.parquet
    • 分割:2023_07_24T14_23_47.053275
      • 路径:**/details_harness|arc:challenge|25_2023-07-24T14:23:47.053275.parquet
    • 分割:latest
      • 路径:**/details_harness|arc:challenge|25_2023-07-24T14:23:47.053275.parquet
  • harness_drop_3

    • 分割:2023_10_13T10_40_56.309063
      • 路径:**/details_harness|drop|3_2023-10-13T10-40-56.309063.parquet
    • 分割:2023_10_15T17_56_49.237621
      • 路径:**/details_harness|drop|3_2023-10-15T17-56-49.237621.parquet
    • 分割:latest
      • 路径:**/details_harness|drop|3_2023-10-15T17-56-49.237621.parquet
  • harness_gsm8k_5

    • 分割:2023_10_13T10_40_56.309063
      • 路径:**/details_harness|gsm8k|5_2023-10-13T10-40-56.309063.parquet
    • 分割:2023_10_15T17_56_49.237621
      • 路径:**/details_harness|gsm8k|5_2023-10-15T17-56-49.237621.parquet
    • 分割:latest
      • 路径:**/details_harness|gsm8k|5_2023-10-15T17-56-49.237621.parquet
  • harness_hellaswag_10

    • 分割:2023_07_20T10_02_18.035133
      • 路径:**/details_harness|hellaswag|10_2023-07-20T10:02:18.035133.parquet
    • 分割:2023_07_24T14_23_47.053275
      • 路径:**/details_harness|hellaswag|10_2023-07-24T14:23:47.053275.parquet
    • 分割:latest
      • 路径:**/details_harness|hellaswag|10_2023-07-24T14:23:47.053275.parquet
  • harness_hendrycksTest_5

    • 分割:2023_07_20T10_02_18.035133
      • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-formal_logic|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-global_facts|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_biology|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_geography|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_physics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-human_aging|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-human_sexuality|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-international_law|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-jurisprudence|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-machine_learning|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-management|5_2023-07-20T10:02:18.035133.parquet
      • 路径:**/details_harness|hendrycksTest-marketing|5_2023-07-20T10:02:18.035133.parquet
      • 路径:`**/details_harness|hendrycksTest-medical_genetics|5_2023-07-20T10:02:18.0351
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