open-llm-leaderboard/details_luffycodes__mcq-hal-vicuna-13b-v1.5
收藏数据集概述
数据集简介
该数据集是在评估模型 luffycodes/mcq-hal-vicuna-13b-v1.5 在 Open LLM Leaderboard 上的运行过程中自动创建的。
数据集组成
- 数据集包含 64 个配置,每个配置对应一个评估任务。
- 数据集从 3 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_luffycodes__mcq-hal-vicuna-13b-v1.5", "harness_winogrande_5", split="train")
最新结果
以下是 2023-09-23T06:42:27.712950 运行的最新结果: python { "all": { "em": 0.1556208053691275, "em_stderr": 0.0037122929662357657, "f1": 0.21929530201342184, "f1_stderr": 0.0037761120976932553, "acc": 0.40820339964803165, "acc_stderr": 0.010171078364256323 }, "harness|drop|3": { "em": 0.1556208053691275, "em_stderr": 0.0037122929662357657, "f1": 0.21929530201342184, "f1_stderr": 0.0037761120976932553 }, "harness|gsm8k|5": { "acc": 0.0887035633055345, "acc_stderr": 0.007831458737058714 }, "harness|winogrande|5": { "acc": 0.7277032359905288, "acc_stderr": 0.012510697991453934 } }
配置详情
-
harness_arc_challenge_25
- 分割:2023_09_02T17_48_54.991558
- 路径:
**/details_harness|arc:challenge|25_2023-09-02T17:48:54.991558.parquet
- 路径:
- 分割:2023_09_02T20_57_06.910156
- 路径:
**/details_harness|arc:challenge|25_2023-09-02T20:57:06.910156.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|arc:challenge|25_2023-09-02T20:57:06.910156.parquet
- 路径:
- 分割:2023_09_02T17_48_54.991558
-
harness_drop_3
- 分割:2023_09_23T06_42_27.712950
- 路径:
**/details_harness|drop|3_2023-09-23T06-42-27.712950.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|drop|3_2023-09-23T06-42-27.712950.parquet
- 路径:
- 分割:2023_09_23T06_42_27.712950
-
harness_gsm8k_5
- 分割:2023_09_23T06_42_27.712950
- 路径:
**/details_harness|gsm8k|5_2023-09-23T06-42-27.712950.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|gsm8k|5_2023-09-23T06-42-27.712950.parquet
- 路径:
- 分割:2023_09_23T06_42_27.712950
-
harness_hellaswag_10
- 分割:2023_09_02T17_48_54.991558
- 路径:
**/details_harness|hellaswag|10_2023-09-02T17:48:54.991558.parquet
- 路径:
- 分割:2023_09_02T20_57_06.910156
- 路径:
**/details_harness|hellaswag|10_2023-09-02T20:57:06.910156.parquet
- 路径:
- 分割:latest
- 路径:
**/details_harness|hellaswag|10_2023-09-02T20:57:06.910156.parquet
- 路径:
- 分割:2023_09_02T17_48_54.991558
-
harness_hendrycksTest_5
- 分割:2023_09_02T17_48_54.991558
- 路径:
**/details_harness|hendrycksTest-abstract_algebra|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-anatomy|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-astronomy|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-business_ethics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-clinical_knowledge|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_biology|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_chemistry|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_computer_science|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_mathematics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_medicine|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-college_physics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-computer_security|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-conceptual_physics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-econometrics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-electrical_engineering|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-elementary_mathematics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-formal_logic|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-global_facts|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_biology|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_chemistry|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_computer_science|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_european_history|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_geography|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_government_and_politics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_macroeconomics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_mathematics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_microeconomics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_physics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_psychology|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_statistics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_us_history|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-high_school_world_history|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-human_aging|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-human_sexuality|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-international_law|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-jurisprudence|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-logical_fallacies|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-machine_learning|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-management|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-marketing|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-medical_genetics|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-miscellaneous|5_2023-09-02T17:48:54.991558.parquet - 路径:
**/details_harness|hendrycksTest-moral_disputes|5_2023-09-02T17:48:54.991558.parquet - 路径:`**/
- 路径:
- 分割:2023_09_02T17_48_54.991558



