open-llm-leaderboard/details_jondurbin__airoboros-65b-gpt4-1.2
收藏数据集概述
数据集简介
该数据集是在对模型 jondurbin/airoboros-65b-gpt4-1.2 进行评估时自动创建的,评估过程在 Open LLM Leaderboard 上进行。
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。
- 数据集由 4 次运行结果组成,每次运行的结果可以在每个配置中找到,并以运行的时间戳命名。
- "train" 分割总是指向最新的结果。
额外配置
- 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_jondurbin__airoboros-65b-gpt4-1.2", "harness_winogrande_5", split="train")
最新结果
以下是 2023-10-22T19:27:20.004298 运行的最新结果: python { "all": { "em": 0.16883389261744966, "em_stderr": 0.0038363072356365218, "f1": 0.24563443791946302, "f1_stderr": 0.003921848688674613, "acc": 0.5304663251500592, "acc_stderr": 0.01174788303833344 }, "harness|drop|3": { "em": 0.16883389261744966, "em_stderr": 0.0038363072356365218, "f1": 0.24563443791946302, "f1_stderr": 0.003921848688674613 }, "harness|gsm8k|5": { "acc": 0.265352539802881, "acc_stderr": 0.012161675464069677 }, "harness|winogrande|5": { "acc": 0.7955801104972375, "acc_stderr": 0.011334090612597202 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_07_25T19_32_20.541789
- 路径:
**/details_harness|arc:challenge|25_2023-07-25T19:32:20.541789.parquet
- 路径:
- 分割:2023_08_03T17_35_02.727730
- 路径:
**/details_harness|arc:challenge|25_2023-08-03T17:35:02.727730.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-03T17:35:02.727730.parquet
- 路径:
- 分割:2023_07_25T19_32_20.541789
-
harness_drop_3
- 分割:2023_10_22T18_25_44.617546
- 路径:
**/details_harness|drop|3_2023-10-22T18-25-44.617546.parquet
- 路径:
- 分割:2023_10_22T19_27_20.004298
- 路径:
**/details_harness|drop|3_2023-10-22T19-27-20.004298.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-22T19-27-20.004298.parquet
- 路径:
- 分割:2023_10_22T18_25_44.617546
-
harness_gsm8k_5
- 分割:2023_10_22T18_25_44.617546
- 路径:
**/details_harness|gsm8k|5_2023-10-22T18-25-44.617546.parquet
- 路径:
- 分割:2023_10_22T19_27_20.004298
- 路径:
**/details_harness|gsm8k|5_2023-10-22T19-27-20.004298.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-22T19-27-20.004298.parquet
- 路径:
- 分割:2023_10_22T18_25_44.617546
-
harness_hellaswag_10
- 分割:2023_07_25T19_32_20.541789
- 路径:
**/details_harness|hellaswag|10_2023-07-25T19:32:20.541789.parquet
- 路径:
- 分割:2023_08_03T17_35_02.727730
- 路径:
**/details_harness|hellaswag|10_2023-08-03T17:35:02.727730.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-08-03T17:35:02.727730.parquet
- 路径:
- 分割:2023_07_25T19_32_20.541789
-
harness_hendrycksTest_5
- 分割:2023_07_25T19_32_20.541789
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-formal_logic|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-global_facts|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_biology|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_geography|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_physics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-human_aging|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-human_sexuality|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-international_law|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-jurisprudence|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-machine_learning|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-management|5_2023-07-25T19:32:20.541789.parquet - 路径:
**/details_harness|hendrycksTest-marketing|5_2023-07-25T19:32:20.541789.parquet - 路径:`**/details
- 路径:
- 分割:2023_07_25T19_32_20.541789



