open-llm-leaderboard-old/details_cognitivecomputations__laserxtral
收藏数据集概述
该数据集是在对模型 cognitivecomputations/laserxtral 进行评估运行期间自动创建的,用于 Open LLM Leaderboard 上的评估。
数据集组成
- 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。
- 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 Open LLM Leaderboard 上的聚合指标。
最新结果
以下是 2024-01-15T20:05:55.052145 运行的最新结果:
python { "all": { "acc": 0.6512747160527517, "acc_stderr": 0.03203788592010298, "acc_norm": 0.6512676159278845, "acc_norm_stderr": 0.03269445186595507, "mc1": 0.4773561811505508, "mc1_stderr": 0.01748554225848965, "mc2": 0.638014374039814, "mc2_stderr": 0.01550457171837664 }, "harness|arc:challenge|25": { "acc": 0.6697952218430034, "acc_stderr": 0.013743085603760424, "acc_norm": 0.6902730375426621, "acc_norm_stderr": 0.013512058415238363 }, "harness|hellaswag|10": { "acc": 0.6931886078470424, "acc_stderr": 0.004602279238122065, "acc_norm": 0.8675562636924915, "acc_norm_stderr": 0.0033827979075230284 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.35, "acc_stderr": 0.0479372485441102, "acc_norm": 0.35, "acc_norm_stderr": 0.0479372485441102 }, # 其他任务的结果省略... }
配置详情
-
harness_arc_challenge_25
- 分割:
2024_01_15T20_05_55.052145,latest - 路径:
**/details_harness|arc:challenge|25_2024-01-15T20-05-55.052145.parquet
- 分割:
-
harness_gsm8k_5
- 分割:
2024_01_15T20_05_55.052145,latest - 路径:
**/details_harness|gsm8k|5_2024-01-15T20-05-55.052145.parquet
- 分割:
-
harness_hellaswag_10
- 分割:
2024_01_15T20_05_55.052145,latest - 路径:
**/details_harness|hellaswag|10_2024-01-15T20-05-55.052145.parquet
- 分割:
-
harness_hendrycksTest_5
- 分割:
2024_01_15T20_05_55.052145,latest - 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-anatomy|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-astronomy|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_biology|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-college_physics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-computer_security|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-econometrics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-global_facts|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-human_aging|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-international_law|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-management|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-marketing|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-nutrition|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-philosophy|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-prehistory|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-professional_law|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-professional_psychology|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-public_relations|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-security_studies|5_2024-01-15T20-05-55.052145.parquet**/details_harness|hendrycksTest-sociology|5_2024-01-15T20-05-55.052145.parquet
- 分割:



