open-llm-leaderboard-old/details_gradientputri__MegaMix-A1-13B
收藏数据集概述
该数据集是在评估模型 gradientputri/MegaMix-A1-13B 的过程中自动创建的,用于 Open LLM Leaderboard。数据集包含 64 个配置,每个配置对应一个评估任务。
数据集结构
数据集由 2 次运行结果组成,每次运行的结果可以在每个配置中找到,以运行的时间戳命名的特定分片形式存储。"train" 分片始终指向最新的结果。
数据加载示例
以下是加载特定运行结果的示例代码:
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_gradientputri__MegaMix-A1-13B", "harness_winogrande_5", split="train")
最新结果
以下是 最新结果 的摘要:
python { "all": { "em": 0.02548238255033557, "em_stderr": 0.001613817848204523, "f1": 0.09458368288590573, "f1_stderr": 0.0021111807048463803, "acc": 0.5013667047038218, "acc_stderr": 0.011878597496401019 }, "harness|drop|3": { "em": 0.02548238255033557, "em_stderr": 0.001613817848204523, "f1": 0.09458368288590573, "f1_stderr": 0.0021111807048463803 }, "harness|gsm8k|5": { "acc": 0.24109173616376042, "acc_stderr": 0.011782246325099723 }, "harness|winogrande|5": { "acc": 0.7616416732438832, "acc_stderr": 0.011974948667702313 } }
配置详情
数据集包含以下配置及其对应的数据文件路径:
-
harness_arc_challenge_25
- 分片:
2023_10_04T02_29_44.126767- 路径:
**/details_harness|arc:challenge|25_2023-10-04T02-29-44.126767.parquet
- 路径:
- 分片:
latest- 路径:
**/details_harness|arc:challenge|25_2023-10-04T02-29-44.126767.parquet
- 路径:
- 分片:
-
harness_drop_3
- 分片:
2023_10_29T01_28_47.828311- 路径:
**/details_harness|drop|3_2023-10-29T01-28-47.828311.parquet
- 路径:
- 分片:
latest- 路径:
**/details_harness|drop|3_2023-10-29T01-28-47.828311.parquet
- 路径:
- 分片:
-
harness_gsm8k_5
- 分片:
2023_10_29T01_28_47.828311- 路径:
**/details_harness|gsm8k|5_2023-10-29T01-28-47.828311.parquet
- 路径:
- 分片:
latest- 路径:
**/details_harness|gsm8k|5_2023-10-29T01-28-47.828311.parquet
- 路径:
- 分片:
-
harness_hellaswag_10
- 分片:
2023_10_04T02_29_44.126767- 路径:
**/details_harness|hellaswag|10_2023-10-04T02-29-44.126767.parquet
- 路径:
- 分片:
latest- 路径:
**/details_harness|hellaswag|10_2023-10-04T02-29-44.126767.parquet
- 路径:
- 分片:
-
harness_hendrycksTest_5
- 分片:
2023_10_04T02_29_44.126767- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-anatomy|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-astronomy|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_biology|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-college_physics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-computer_security|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-econometrics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-global_facts|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_geography|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_physics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-human_aging|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-human_sexuality|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-international_law|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-jurisprudence|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-machine_learning|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-management|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-marketing|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-medical_genetics|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-miscellaneous|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-moral_disputes|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-nutrition|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-philosophy|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-prehistory|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-professional_accounting|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-professional_law|5_2023-10-04T02-29-44.126767.parquet**/details_harness|hendrycksTest-professional_medicine|5_2023-10-04T02-29-44.126767.parquet- `**/details_harness|hendrycksTest-professional_psychology|5
- 路径:
- 分片:



