open-llm-leaderboard/details_HyperbeeAI__Tulpar-7b-v1

该数据集是在Open LLM Leaderboard上对模型HyperbeeAI/Tulpar-7b-v1进行评估时自动创建的。数据集由64个配置组成，每个配置对应一个评估任务。数据集是从2次运行中创建的，每次运行在每个配置中表示为特定的拆分。train拆分始终指向最新的结果。此外，名为results的配置存储了所有运行的聚合结果，这些结果用于计算和显示Open LLM Leaderboard上的聚合指标。README还提供了如何使用`datasets`库中的`load_dataset`函数加载运行详情的示例。

This dataset was automatically created during the evaluation of the model HyperbeeAI/Tulpar-7b-v1 on the Open LLM Leaderboard. The dataset consists of 64 configurations, each corresponding to one evaluation task. The dataset is constructed from two runs, where each run is represented as a specific split under each configuration. The `train` split always points to the most recent results. In addition, the configuration named `results` stores the aggregated results across all runs, which are used to calculate and display the aggregated metrics on the Open LLM Leaderboard. The README also provides an example of how to use the `load_dataset` function from the `datasets` library to load the run details.

原始信息汇总

数据集概述

数据集简介

该数据集是在对模型 HyperbeeAI/Tulpar-7b-v1 进行评估运行期间自动创建的，用于 Open LLM Leaderboard。

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 2 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_HyperbeeAI__Tulpar-7b-v1", "harness_winogrande_5", split="train")

最新结果

以下是 2023-10-23T09:02:28.162757 运行的最新结果： python { "all": { "em": 0.2915268456375839, "em_stderr": 0.004654152691335802, "f1": 0.3658179530201351, "f1_stderr": 0.004568137923093851, "acc": 0.3656847615417434, "acc_stderr": 0.0074116135789820126 }, "harness|drop|3": { "em": 0.2915268456375839, "em_stderr": 0.004654152691335802, "f1": 0.3658179530201351, "f1_stderr": 0.004568137923093851 }, "harness|gsm8k|5": { "acc": 0.006823351023502654, "acc_stderr": 0.0022675371022544944 }, "harness|winogrande|5": { "acc": 0.7245461720599842, "acc_stderr": 0.01255569005570953 } }

配置详情

以下是数据集的部分配置详情：

配置列表

• harness_arc_challenge_25
• harness_drop_3
• harness_gsm8k_5
• harness_hellaswag_10
• harness_hendrycksTest_5
• harness_hendrycksTest_abstract_algebra_5
• harness_hendrycksTest_anatomy_5
• harness_hendrycksTest_astronomy_5
• harness_hendrycksTest_business_ethics_5
• harness_hendrycksTest_clinical_knowledge_5
• harness_hendrycksTest_college_biology_5
• harness_hendrycksTest_college_chemistry_5
• harness_hendrycksTest_college_computer_science_5
• harness_hendrycksTest_college_mathematics_5
• harness_hendrycksTest_college_medicine_5
• harness_hendrycksTest_college_physics_5
• harness_hendrycksTest_computer_security_5
• harness_hendrycksTest_conceptual_physics_5
• harness_hendrycksTest_econometrics_5
• harness_hendrycksTest_electrical_engineering_5
• harness_hendrycksTest_elementary_mathematics_5

数据文件路径

每个配置包含不同分割的数据文件路径，例如：

• harness_arc_challenge_25 配置包含 2023_09_14T07_59_39.326009 和 latest 分割的数据文件路径。
• harness_drop_3 配置包含 2023_10_23T09_02_28.162757 和 latest 分割的数据文件路径。

具体路径格式如下：

• **/details_harness|arc:challenge|25_2023-09-14T07-59-39.326009.parquet
• **/details_harness|drop|3_2023-10-23T09-02-28.162757.parquet