open-llm-leaderboard-old/details_KnutJaegersberg__galactica-orca-wizardlm-1.3b
收藏数据集概述
该数据集是在对模型 KnutJaegersberg/galactica-orca-wizardlm-1.3b 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集组成
- 数据集由 61 个配置组成,每个配置对应一个评估任务。
- 数据集从 1 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" 分割始终指向最新的结果。
- 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。
最新结果
以下是 2023-07-27T10:33:13.422388 运行的最新结果:
python { "all": { "acc": 0.2607167828475717, "acc_stderr": 0.031890976655530354, "acc_norm": 0.2619942427603414, "acc_norm_stderr": 0.03190140774792886, "mc1": 0.25458996328029376, "mc1_stderr": 0.015250117079156494, "mc2": 0.41268090479367026, "mc2_stderr": 0.015430187488430076 }, "harness|arc:challenge|25": { "acc": 0.27303754266211605, "acc_stderr": 0.013019332762635727, "acc_norm": 0.30887372013651876, "acc_norm_stderr": 0.013501770929344004 }, "harness|hellaswag|10": { "acc": 0.32065325632344155, "acc_stderr": 0.004657738398900916, "acc_norm": 0.3601872137024497, "acc_norm_stderr": 0.004790734683704583 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.31, "acc_stderr": 0.04648231987117316, "acc_norm": 0.31, "acc_norm_stderr": 0.04648231987117316 }, # 其他任务的结果省略... }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2023_07_27T10_33_13.422388
- path: **/details_harness|arc:challenge|25_2023-07-27T10:33:13.422388.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-07-27T10:33:13.422388.parquet
- split: 2023_07_27T10_33_13.422388
-
config_name: harness_hellaswag_10
- split: 2023_07_27T10_33_13.422388
- path: **/details_harness|hellaswag|10_2023-07-27T10:33:13.422388.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-07-27T10:33:13.422388.parquet
- split: 2023_07_27T10_33_13.422388
-
config_name: harness_hendrycksTest_5
- split: 2023_07_27T10_33_13.422388
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2023-07-27T10:33:13.422388.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2023-07-27T10:33:13.422388.parquet
其他路径省略...
- path:
- split: 2023_07_27T10_33_13.422388
}



