open-llm-leaderboard-old/details_l3utterfly__tinyllama-1.1b-layla-v1
收藏数据集概述
数据集简介
该数据集是在对模型 l3utterfly/tinyllama-1.1b-layla-v1 进行评估运行期间自动创建的,用于 Open LLM Leaderboard。
数据集结构
- 数据集包含 63 个配置,每个配置对应一个评估任务。
- 数据集由 1 次运行创建,每次运行的详细结果存储在特定的分割中,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 "results" 配置存储所有运行的聚合结果,用于计算和显示聚合指标。
数据加载示例
python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_l3utterfly__tinyllama-1.1b-layla-v1", "harness_winogrande_5", split="train")
最新结果
以下是 2024-02-04T01:13:05.279665 运行的最新结果:
python { "all": { "acc": 0.25360571978457686, "acc_stderr": 0.030631027319513834, "acc_norm": 0.2545795497909492, "acc_norm_stderr": 0.03138700259514901, "mc1": 0.2533659730722154, "mc1_stderr": 0.01522589934082685, "mc2": 0.4102690094325491, "mc2_stderr": 0.014297501504936961 }, "harness|arc:challenge|25": { "acc": 0.3225255972696246, "acc_stderr": 0.01365998089427737, "acc_norm": 0.3438566552901024, "acc_norm_stderr": 0.01388064457015621 }, "harness|hellaswag|10": { "acc": 0.4600677155945031, "acc_stderr": 0.004973842670559798, "acc_norm": 0.5985859390559649, "acc_norm_stderr": 0.004891826692722825 }, "harness|hendrycksTest-abstract_algebra|5": { "acc": 0.25, "acc_stderr": 0.04351941398892446, "acc_norm": 0.25, "acc_norm_stderr": 0.04351941398892446 }, # 其他任务的结果... }
配置详情
-
harness_arc_challenge_25
- 分割: 2024_02_04T01_13_05.279665, latest
- 路径:
**/details_harness|arc:challenge|25_2024-02-04T01-13-05.279665.parquet
-
harness_gsm8k_5
- 分割: 2024_02_04T01_13_05.279665, latest
- 路径:
**/details_harness|gsm8k|5_2024-02-04T01-13-05.279665.parquet
-
harness_hellaswag_10
- 分割: 2024_02_04T01_13_05.279665, latest
- 路径:
**/details_harness|hellaswag|10_2024-02-04T01-13-05.279665.parquet
-
harness_hendrycksTest_5
- 分割: 2024_02_04T01_13_05.279665, latest
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-anatomy|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-astronomy|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-business_ethics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-clinical_knowledge|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_biology|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_chemistry|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_computer_science|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_mathematics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_medicine|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-college_physics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-computer_security|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-conceptual_physics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-econometrics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-electrical_engineering|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-elementary_mathematics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-formal_logic|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-global_facts|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_biology|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_chemistry|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_computer_science|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_european_history|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_geography|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_government_and_politics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_macroeconomics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_mathematics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_microeconomics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_physics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_psychology|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_statistics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_us_history|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-high_school_world_history|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-human_aging|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-human_sexuality|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-international_law|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-jurisprudence|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-logical_fallacies|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-machine_learning|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-management|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-marketing|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-medical_genetics|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-miscellaneous|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-moral_disputes|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-moral_scenarios|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-nutrition|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-philosophy|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-prehistory|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-professional_accounting|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-professional_law|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-professional_medicine|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-professional_psychology|5_2024-02-04T01-13-05.279665.parquet**/details_harness|hendrycksTest-public_relations|5_2024-02-04T01-13-05.279665.parquet- `**/details_harness|hendrycksTest-security_studies|5_2



