OALL/details_Lina-Z__arabic_llm_test_v2

该数据集是在评估模型 [Lina-Z/arabic_llm_test](https://huggingface.co/Lina-Z/arabic_llm_test) 的过程中自动创建的，包含116个配置，每个配置对应一个评估任务。数据集由1次运行的结果组成，每次运行的结果存储在以运行时间戳命名的特定分割中。train 分割始终指向最新的结果。还有一个名为 results 的额外配置，用于存储所有运行结果的汇总。内容还包括一个Python代码示例，说明如何从一个运行中加载数据集的详细信息。它还提供了关于特定运行的最新结果的信息，包括各种任务的准确性和标准误差。

The dataset is automatically created during the evaluation run of the model [Lina-Z/arabic_llm_test](https://huggingface.co/Lina-Z/arabic_llm_test) and contains 116 configurations, each corresponding to an evaluated task. The dataset includes results from 1 run, with each runs results stored in a specific split named by the timestamp of the run. The train split always points to the latest results. There is also a results configuration that stores all the aggregated results of the run. The content includes a Python code snippet on how to load the dataset details from a run. It also provides information about the latest results from a specific run, including the accuracy and standard error for various tasks.