open-llm-leaderboard-old/details_ozayezerceli__BetterSaul-7B-slerp
收藏数据集概述
该数据集是在对模型 ozayezerceli/BetterSaul-7B-slerp 进行评估运行期间自动创建的。数据集包含 63 个配置,每个配置对应一个评估任务。数据集从 1 次运行中创建,每次运行的详细信息可以在每个配置的特定分片中找到,分片名称使用运行的时间戳。"train" 分片始终指向最新的结果。
数据集结构
配置
数据集包含以下配置:
harness_arc_challenge_25harness_gsm8k_5harness_hellaswag_10harness_hendrycksTest_5
每个配置包含多个分片,包括特定时间戳的分片和最新的分片。
分片
每个配置的分片包括:
- 特定时间戳的分片,例如
2024_03_22T01_39_20.926719 - 最新的分片,例如
latest
数据文件路径
每个分片的数据文件路径如下:
harness_arc_challenge_25:**/details_harness|arc:challenge|25_2024-03-22T01-39-20.926719.parquet
harness_gsm8k_5:**/details_harness|gsm8k|5_2024-03-22T01-39-20.926719.parquet
harness_hellaswag_10:**/details_harness|hellaswag|10_2024-03-22T01-39-20.926719.parquet
harness_hendrycksTest_5:**/details_harness|hendrycksTest-abstract_algebra|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-anatomy|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-astronomy|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_biology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-college_physics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-computer_security|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-econometrics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-global_facts|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-human_aging|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-international_law|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-management|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-marketing|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-nutrition|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-philosophy|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-prehistory|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-professional_law|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-professional_psychology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-public_relations|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-security_studies|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-sociology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-us_foreign_policy|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-virology|5_2024-03-22T01-39-20.926719.parquet**/details_harness|hendrycksTest-world_religions|5_2024-03-22T01-39-20.926719.parquet
最新结果
以下是来自 2024-03-22T01:39:20.926719 运行的最新结果:
python { "all": { "acc": 0.6469318827706652, "acc_stderr": 0.03211876738336368, "acc_norm": 0.6474374045483131, "acc_norm_stderr": 0.03277873333603153, "mc1": 0.4614443084455324, "mc1_stderr": 0.017451384104637455, "mc2": 0.6307577163869668, "mc2_stderr": 0.014942073227822787 }, "harness|arc:challenge|25": { "acc": 0.6424914675767918, "acc_stderr": 0.014005494275916573, "acc_norm": 0.6808873720136519, "acc_norm_stderr": 0.01362169611917331 }, "harness|hellaswag|10": { "acc": 0.6704839673371839, "acc_stderr": 0.004690768393854473, "acc_norm": 0.862975502887871, "acc_norm_stderr": 0.003431704298641856 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.34, "acc_stderr": 0.04760952285695236, "acc_norm": 0.34, "acc_norm_stderr": 0.04760952285695236 }, "harness|hendrycksTest-anatomy|5": { "acc": 0.



