open-llm-leaderboard/details_jordiclive__gpt4all-alpaca-oa-codealpaca-lora-13b
收藏数据集概述
数据集简介
该数据集是在评估模型jordiclive/gpt4all-alpaca-oa-codealpaca-lora-13b在Open LLM Leaderboard上的自动创建的。
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。
- 数据来源: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。
- 最新结果:"train"分割总是指向最新的结果。
- 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_jordiclive__gpt4all-alpaca-oa-codealpaca-lora-13b", "harness_winogrande_5", split="train")
最新结果
以下是最新结果的摘要: python { "all": { "em": 0.0019924496644295304, "em_stderr": 0.000456667646266702, "f1": 0.05642302852349, "f1_stderr": 0.0012977737732540458, "acc": 0.41872834230806744, "acc_stderr": 0.009633077195432445 }, "harness|drop|3": { "em": 0.0019924496644295304, "em_stderr": 0.000456667646266702, "f1": 0.05642302852349, "f1_stderr": 0.0012977737732540458 }, "harness|gsm8k|5": { "acc": 0.0758150113722517, "acc_stderr": 0.0072912057231625796 }, "harness|winogrande|5": { "acc": 0.7616416732438832, "acc_stderr": 0.011974948667702311 } }
配置详情
以下是部分配置的详细信息:
-
harness_arc_challenge_25
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|arc:challenge|25_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-08-09T17:57:53.688517.parquet
-
harness_drop_3
- 分割:2023_09_22T20_03_30.331669
- 路径:
**/details_harness|drop|3_2023-09-22T20-03-30.331669.parquet - 分割:latest
- 路径:
**/details_harness|drop|3_2023-09-22T20-03-30.331669.parquet
-
harness_gsm8k_5
- 分割:2023_09_22T20_03_30.331669
- 路径:
**/details_harness|gsm8k|5_2023-09-22T20-03-30.331669.parquet - 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-09-22T20-03-30.331669.parquet
-
harness_hellaswag_10
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hellaswag|10_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_5
- 分割:2023_08_09T17_57_53.688517
- 路径:多个路径,例如
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-09T17:57:53.688517.parquet等 - 分割:latest
- 路径:多个路径,例如
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-09T17:57:53.688517.parquet等
-
harness_hendrycksTest_abstract_algebra_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_anatomy_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_astronomy_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_business_ethics_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_clinical_knowledge_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_biology_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_chemistry_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_computer_science_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_mathematics_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_medicine_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_college_physics_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_computer_security_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_conceptual_physics_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_econometrics_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-09T17:57:53.688517.parquet - 分割:latest
- 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-08-09T17:57:53.688517.parquet
-
harness_hendrycksTest_electrical_engineering_5
- 分割:2023_08_09T17_57_53.688517
- 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-08-09T17:57:53.688517.parquet



