遇见数据集

open-llm-leaderboard/details_uukuguy__speechless-codellama-orca-13b

收藏
Hugging Face2023-10-23 更新2024-03-04 收录
官方服务:

资源简介:

该数据集是在模型 uukuguy/speechless-codellama-orca-13b 在 Open LLM Leaderboard 上的评估运行期间自动创建的。数据集由 64 个配置组成,每个配置对应一个被评估的任务。它由 4 次运行生成,每次运行在每个配置中表示为特定的拆分。train 拆分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 还提供了一个示例,展示了如何使用 `datasets` 库中的 `load_dataset` 函数加载运行中的详细信息。

This dataset was automatically created during the evaluation run of the model uukuguy/speechless-codellama-orca-13b on the Open LLM Leaderboard. The dataset consists of 64 configurations, each corresponding to one evaluated task. It is generated through 4 runs, with each run represented as a specific split under each configuration. 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 an example demonstrating how to use the `load_dataset` function from the `datasets` library to load detailed information from a run.

提供机构:
open-llm-leaderboard
原始信息汇总

数据集概述

数据集来源

该数据集是在评估模型 uukuguy/speechless-codellama-orca-13bOpen LLM Leaderboard 上的运行过程中自动创建的。

数据集组成

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 4 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。

数据集加载示例

以下是加载数据集的示例代码: python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_uukuguy__speechless-codellama-orca-13b", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-23T18:58:19.504304 运行 的最新结果: python { "all": { "em": 0.2686661073825503, "em_stderr": 0.004539457381903774, "f1": 0.3305505453020149, "f1_stderr": 0.00452265523617686, "acc": 0.3499942854509976, "acc_stderr": 0.010012878870777758 }, "harness|drop|3": { "em": 0.2686661073825503, "em_stderr": 0.004539457381903774, "f1": 0.3305505453020149, "f1_stderr": 0.00452265523617686 }, "harness|gsm8k|5": { "acc": 0.05989385898407885, "acc_stderr": 0.006536148151288716 }, "harness|winogrande|5": { "acc": 0.6400947119179163, "acc_stderr": 0.013489609590266799 } }

配置详情

以下是数据集的配置详情:

  • harness_arc_challenge_25

    • 分割:2023_09_04T06_25_23.128128
      • 路径:**/details_harness|arc:challenge|25_2023-09-04T06:25:23.128128.parquet
    • 分割:2023_09_12T14_20_48.062177
      • 路径:**/details_harness|arc:challenge|25_2023-09-12T14-20-48.062177.parquet
    • 分割:latest
      • 路径:**/details_harness|arc:challenge|25_2023-09-12T14-20-48.062177.parquet
  • harness_drop_3

    • 分割:2023_10_17T18_24_08.012097
      • 路径:**/details_harness|drop|3_2023-10-17T18-24-08.012097.parquet
    • 分割:2023_10_23T18_58_19.504304
      • 路径:**/details_harness|drop|3_2023-10-23T18-58-19.504304.parquet
    • 分割:latest
      • 路径:**/details_harness|drop|3_2023-10-23T18-58-19.504304.parquet
  • harness_gsm8k_5

    • 分割:2023_10_17T18_24_08.012097
      • 路径:**/details_harness|gsm8k|5_2023-10-17T18-24-08.012097.parquet
    • 分割:2023_10_23T18_58_19.504304
      • 路径:**/details_harness|gsm8k|5_2023-10-23T18-58-19.504304.parquet
    • 分割:latest
      • 路径:**/details_harness|gsm8k|5_2023-10-23T18-58-19.504304.parquet
  • harness_hellaswag_10

    • 分割:2023_09_04T06_25_23.128128
      • 路径:**/details_harness|hellaswag|10_2023-09-04T06:25:23.128128.parquet
    • 分割:2023_09_12T14_20_48.062177
      • 路径:**/details_harness|hellaswag|10_2023-09-12T14-20-48.062177.parquet
    • 分割:latest
      • 路径:**/details_harness|hellaswag|10_2023-09-12T14-20-48.062177.parquet
  • harness_hendrycksTest_5

    • 分割:2023_09_04T06_25_23.128128
      • 路径:**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-anatomy|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-astronomy|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-business_ethics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_biology|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_chemistry|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_computer_science|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_mathematics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_medicine|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-college_physics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-computer_security|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-econometrics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-formal_logic|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-global_facts|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_biology|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_geography|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_physics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-human_aging|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-human_sexuality|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-international_law|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-jurisprudence|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-machine_learning|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-management|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-marketing|5_2023-09-04T06:25:23.128128.parquet
      • 路径:**/details_harness|hendrycksTest-medical_genetics|5_2023-09-04T06:25:23.128128.parquet
      • 路径:`**/details_harness|hendrycksTest-
二维码
社区交流群
二维码
科研交流群
商业服务