open-llm-leaderboard-old/details_KnutJaegersberg__RWKV-pileplus-1B5-evol_instruct_v2
收藏数据集概述
该数据集是在对模型 KnutJaegersberg/RWKV-pileplus-1B5-evol_instruct_v2 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集组成
- 该数据集包含 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。
- 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。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_KnutJaegersberg__RWKV-pileplus-1B5-evol_instruct_v2", "harness_truthfulqa_mc_0", split="train")
最新结果
这些是最新的结果,来自 2023-10-02T09:01:04.742783 的运行: python { "all": { "acc": 0.25494922839990974, "acc_stderr": 0.0314010444904469, "acc_norm": 0.25761849126230174, "acc_norm_stderr": 0.03140672184687226, "mc1": 0.21542227662178703, "mc1_stderr": 0.014391902652427678, "mc2": 0.3521451447553084, "mc2_stderr": 0.013530448314563733 }, "harness|arc:challenge|25": { "acc": 0.2935153583617747, "acc_stderr": 0.013307250444941125, "acc_norm": 0.318259385665529, "acc_norm_stderr": 0.013611993916971453 }, "harness|hellaswag|10": { "acc": 0.4223262298346943, "acc_stderr": 0.0049292048643159725, "acc_norm": 0.5550687114120693, "acc_norm_stderr": 0.004959425421382027 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.28, "acc_stderr": 0.04512608598542127, "acc_norm": 0.28, "acc_norm_stderr": 0.04512608598542127 }, # 其他任务的结果... }
配置详情
-
harness_arc_challenge_25
- 分割: 2023_10_02T09_01_04.742783
- 路径:
**/details_harness|arc:challenge|25_2023-10-02T09-01-04.742783.parquet - 分割: latest
- 路径:
**/details_harness|arc:challenge|25_2023-10-02T09-01-04.742783.parquet
-
harness_hellaswag_10
- 分割: 2023_10_02T09_01_04.742783
- 路径:
**/details_harness|hellaswag|10_2023-10-02T09-01-04.742783.parquet - 分割: latest
- 路径:
**/details_harness|hellaswag|10_2023-10-02T09-01-04.742783.parquet
-
harness_hendrycksTest_5
- 分割: 2023_10_02T09_01_04.742783
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-anatomy|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-astronomy|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-business_ethics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_biology|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_chemistry|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_computer_science|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_mathematics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_medicine|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-college_physics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-computer_security|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-econometrics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-formal_logic|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-global_facts|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_biology|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_geography|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_physics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-human_aging|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-human_sexuality|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-international_law|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-jurisprudence|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-machine_learning|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-management|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-marketing|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-medical_genetics|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-miscellaneous|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-moral_disputes|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-nutrition|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-philosophy|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-prehistory|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-professional_accounting|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-professional_law|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-professional_medicine|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-professional_psychology|5_2023-10-02T09-01-04.742783.parquet**/details_harness|hendrycksTest-public_relations|5_2023-10-02T09-01-04.742783.parquet- `**/details_harness|hendrycksTest-security_studies|5_2023-10-02T09-01-04



