open-llm-leaderboard/details_augtoma__qCammel-70x
收藏数据集概述
数据集简介
该数据集是在评估模型augtoma/qCammel-70x在Open LLM Leaderboard上的自动创建的。
数据集结构
- 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2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"分割总是指向最新的结果。
- 一个额外的配置"results"存储所有运行的聚合结果,用于计算和显示聚合指标在Open LLM Leaderboard上。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_augtoma__qCammel-70x", "harness_winogrande_5", split="train")
最新结果
以下是2023-09-24T00:38:03.634221运行的最新结果: python { "all": { "em": 0.033766778523489936, "em_stderr": 0.001849802869119515, "f1": 0.10340918624161041, "f1_stderr": 0.0022106009828094797, "acc": 0.5700654570173166, "acc_stderr": 0.011407494958111332 }, "harness|drop|3": { "em": 0.033766778523489936, "em_stderr": 0.001849802869119515, "f1": 0.10340918624161041, "f1_stderr": 0.0022106009828094797 }, "harness|gsm8k|5": { "acc": 0.2971948445792267, "acc_stderr": 0.012588685966624186 }, "harness|winogrande|5": { "acc": 0.8429360694554064, "acc_stderr": 0.010226303949598479 } }
配置详情
-
harness_arc_challenge_25
- 分割: 2023_08_18T05_27_12.496393
- 路径:
**/details_harness|arc:challenge|25_2023-08-18T05:27:12.496393.parquet - 分割: latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-18T05:27:12.496393.parquet
-
harness_drop_3
- 分割: 2023_09_24T00_38_03.634221
- 路径:
**/details_harness|drop|3_2023-09-24T00-38-03.634221.parquet - 分割: latest
- 路径:
**/details_harness|drop|3_2023-09-24T00-38-03.634221.parquet
-
harness_gsm8k_5
- 分割: 2023_09_24T00_38_03.634221
- 路径:
**/details_harness|gsm8k|5_2023-09-24T00-38-03.634221.parquet - 分割: latest
- 路径:
**/details_harness|gsm8k|5_2023-09-24T00-38-03.634221.parquet
-
harness_hellaswag_10
- 分割: 2023_08_18T05_27_12.496393
- 路径:
**/details_harness|hellaswag|10_2023-08-18T05:27:12.496393.parquet - 分割: latest
- 路径:
**/details_harness|hellaswag|10_2023-08-18T05:27:12.496393.parquet
-
harness_hendrycksTest_5
- 分割: 2023_08_18T05_27_12.496393
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-anatomy|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-astronomy|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_biology|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-college_physics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-computer_security|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-econometrics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-global_facts|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_geography|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_physics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-human_aging|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-human_sexuality|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-international_law|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-jurisprudence|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-machine_learning|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-management|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-marketing|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-medical_genetics|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-miscellaneous|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-moral_disputes|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-nutrition|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-philosophy|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-prehistory|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-professional_accounting|5_2023-08-18T05:27:12.496393.parquet**/details_harness|hendrycksTest-professional_law|5_2023-08-18T05:27:12.496393.parquet- `**/details_harness|hendrycksTest-professional_medicine|5_2023-08



