open-llm-leaderboard/details_NousResearch__Redmond-Puffin-13B
收藏数据集概述
数据集简介
该数据集是在评估模型 NousResearch/Redmond-Puffin-13B 在 Open LLM Leaderboard 上的运行过程中自动创建的。
数据集结构
- 配置数量:64个配置,每个配置对应一个评估任务。
- 运行次数:数据集由4次运行创建,每次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" 分割始终指向最新的结果。
- 结果汇总:一个额外的配置 "results" 存储所有运行的汇总结果,用于计算和显示在 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_NousResearch__Redmond-Puffin-13B", "harness_winogrande_5", split="train")
最新结果示例
以下是 2023-10-19T14:40:40.594002 运行的最新结果: python { "all": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669685, "f1": 0.06032822986577185, "f1_stderr": 0.0013617956382083536, "acc": 0.4385024770026802, "acc_stderr": 0.01030687565094663 }, "harness|drop|3": { "em": 0.0019924496644295304, "em_stderr": 0.00045666764626669685, "f1": 0.06032822986577185, "f1_stderr": 0.0013617956382083536 }, "harness|gsm8k|5": { "acc": 0.11220621683093253, "acc_stderr": 0.00869374313824238 }, "harness|winogrande|5": { "acc": 0.7647987371744278, "acc_stderr": 0.011920008163650882 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_07_24T14_36_07.179231
- 路径:
**/details_harness|arc:challenge|25_2023-07-24T14:36:07.179231.parquet
- 路径:
- 分割:2023_07_25T10_58_43.573402
- 路径:
**/details_harness|arc:challenge|25_2023-07-25T10:58:43.573402.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-07-25T10:58:43.573402.parquet
- 路径:
- 分割:2023_07_24T14_36_07.179231
-
harness_drop_3
- 分割:2023_10_18T01_48_06.701008
- 路径:
**/details_harness|drop|3_2023-10-18T01-48-06.701008.parquet
- 路径:
- 分割:2023_10_19T14_40_40.594002
- 路径:
**/details_harness|drop|3_2023-10-19T14-40-40.594002.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|drop|3_2023-10-19T14-40-40.594002.parquet
- 路径:
- 分割:2023_10_18T01_48_06.701008
-
harness_gsm8k_5
- 分割:2023_10_18T01_48_06.701008
- 路径:
**/details_harness|gsm8k|5_2023-10-18T01-48-06.701008.parquet
- 路径:
- 分割:2023_10_19T14_40_40.594002
- 路径:
**/details_harness|gsm8k|5_2023-10-19T14-40-40.594002.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-10-19T14-40-40.594002.parquet
- 路径:
- 分割:2023_10_18T01_48_06.701008
-
harness_hellaswag_10
- 分割:2023_07_24T14_36_07.179231
- 路径:
**/details_harness|hellaswag|10_2023-07-24T14:36:07.179231.parquet
- 路径:
- 分割:2023_07_25T10_58_43.573402
- 路径:
**/details_harness|hellaswag|10_2023-07-25T10:58:43.573402.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-07-25T10:58:43.573402.parquet
- 路径:
- 分割:2023_07_24T14_36_07.179231
-
harness_hendrycksTest_5
- 分割:2023_07_24T14_36_07.179231
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-formal_logic|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-global_facts|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_biology|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_european_history|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_geography|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_mathematics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_physics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_psychology|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_statistics|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_us_history|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-high_school_world_history|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-human_aging|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-human_sexuality|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-international_law|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-jurisprudence|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-logical_fallacies|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-machine_learning|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-management|5_2023-07-24T14:36:07.179231.parquet - 路径:
**/details_harness|hendrycksTest-marketing|5_2023-07-24T14:36:07.179231.parquet - 路径:`**/details_harness|hendrycksTest
- 路径:
- 分割:2023_07_24T14_36_07.179231



