open-llm-leaderboard/details_grimpep__llama2-22b-wizard_vicuna
收藏数据集概述
数据集摘要
该数据集是在评估模型 grimpep/llama2-22b-wizard_vicuna 在 Open LLM Leaderboard 上的评估运行期间自动创建的。
数据集组成
- 数据集包含 61 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。
- 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。
加载数据集示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_grimpep__llama2-22b-wizard_vicuna", "harness_truthfulqa_mc_0", split="train")
最新结果
以下是 2023-08-17T14:12:20.144901 运行的最新结果:
python { "all": { "acc": 0.5471453447112705, "acc_stderr": 0.034452241209601206, "acc_norm": 0.550874294679223, "acc_norm_stderr": 0.03443332656790291, "mc1": 0.31334149326805383, "mc1_stderr": 0.0162380650690596, "mc2": 0.4692973392633332, "mc2_stderr": 0.0156700439246235 }, "harness|arc:challenge|25": { "acc": 0.560580204778157, "acc_stderr": 0.014503747823580123, "acc_norm": 0.5895904436860068, "acc_norm_stderr": 0.014374922192642662 }, "harness|hellaswag|10": { "acc": 0.6290579565823541, "acc_stderr": 0.004820697457420421, "acc_norm": 0.8200557657837084, "acc_norm_stderr": 0.003833559228158675 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.35, "acc_stderr": 0.04793724854411022, "acc_norm": 0.35, "acc_norm_stderr": 0.04793724854411022 }, ... }
配置详情
-
config_name: harness_arc_challenge_25
- split: 2023_08_17T14_12_20.144901
- path: **/details_harness|arc:challenge|25_2023-08-17T14:12:20.144901.parquet
- split: latest
- path: **/details_harness|arc:challenge|25_2023-08-17T14:12:20.144901.parquet
- split: 2023_08_17T14_12_20.144901
-
config_name: harness_hellaswag_10
- split: 2023_08_17T14_12_20.144901
- path: **/details_harness|hellaswag|10_2023-08-17T14:12:20.144901.parquet
- split: latest
- path: **/details_harness|hellaswag|10_2023-08-17T14:12:20.144901.parquet
- split: 2023_08_17T14_12_20.144901
-
config_name: harness_hendrycksTest_5
- split: 2023_08_17T14_12_20.144901
- path:
- **/details_harness|hendrycksTest-abstract_algebra|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-anatomy|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-astronomy|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-business_ethics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-clinical_knowledge|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_biology|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_chemistry|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_computer_science|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_mathematics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_medicine|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-college_physics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-computer_security|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-conceptual_physics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-econometrics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-electrical_engineering|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-elementary_mathematics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-formal_logic|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-global_facts|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_biology|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_chemistry|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_computer_science|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_european_history|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_geography|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_mathematics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_microeconomics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_physics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_psychology|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_statistics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_us_history|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-high_school_world_history|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-human_aging|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-human_sexuality|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-international_law|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-jurisprudence|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-logical_fallacies|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-machine_learning|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-management|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-marketing|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-medical_genetics|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-miscellaneous|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-moral_disputes|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-moral_scenarios|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-nutrition|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-philosophy|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-prehistory|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-professional_accounting|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-professional_law|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-professional_medicine|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-professional_psychology|5_2023-08-17T14:12:20.144901.parquet
- **/details_harness|hendrycksTest-public_relations|5_2023-08-17T14:12:20.144901.
- path:
- split: 2023_08_17T14_12_20.144901



