open-llm-leaderboard/details_eachadea__vicuna-13b-1.1

该数据集是在模型eachadea/vicuna-13b-1.1的评估运行中自动创建的。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，每次运行都可以在特定配置中找到，分割名称使用运行的时间戳。train分割始终指向最新的结果。此外，还有一个results配置存储了所有运行的聚合结果，并用于计算和显示在Open LLM Leaderboard上的聚合指标。

This dataset was automatically created during the evaluation runs of the model eachadea/vicuna-13b-1.1. The dataset consists of 64 configurations, each corresponding to one evaluation task. The dataset is constructed from two runs, each of which can be found under a specific configuration, and the split names use the timestamp of the run. The train split always points to the most recent results. Additionally, there is a "results" configuration that stores the aggregated results of all runs, and is used to calculate and display the aggregated metrics on the Open LLM Leaderboard.

原始信息汇总

数据集概述

数据集简介

该数据集是在评估模型 eachadea/vicuna-13b-1.1 在 Open LLM Leaderboard 上的自动创建的。数据集包含64个配置，每个配置对应一个评估任务。

数据集结构

数据集由2次运行结果组成，每个运行结果可以在每个配置中作为一个特定的分割找到，分割名称使用运行的时间戳。"train" 分割总是指向最新的结果。

数据集加载示例

以下是加载数据集的示例代码： python from datasets import load_dataset data = load_dataset("open-llm-leaderboard/details_eachadea__vicuna-13b-1.1", "harness_winogrande_5", split="train")

最新结果

以下是 最新结果 的示例： python { "all": { "em": 0.029677013422818792, "em_stderr": 0.0017378324714143493, "f1": 0.09310612416107406, "f1_stderr": 0.002167792401176146, "acc": 0.4141695683211732, "acc_stderr": 0.010019161585538096 }, "harness|drop|3": { "em": 0.029677013422818792, "em_stderr": 0.0017378324714143493, "f1": 0.09310612416107406, "f1_stderr": 0.002167792401176146 }, "harness|gsm8k|5": { "acc": 0.08642911296436695, "acc_stderr": 0.00774004433710381 }, "harness|winogrande|5": { "acc": 0.7419100236779794, "acc_stderr": 0.012298278833972384 } }

配置详情

以下是数据集的部分配置详情：

• harness_arc_challenge_25

  • 分割：2023_07_19T18_54_56.836268
  • 路径：**/details_harness|arc:challenge|25_2023-07-19T18:54:56.836268.parquet
  • 分割：latest
  • 路径：**/details_harness|arc:challenge|25_2023-07-19T18:54:56.836268.parquet

• harness_drop_3

  • 分割：2023_10_14T21_09_04.569052
  • 路径：**/details_harness|drop|3_2023-10-14T21-09-04.569052.parquet
  • 分割：latest
  • 路径：**/details_harness|drop|3_2023-10-14T21-09-04.569052.parquet

• harness_gsm8k_5

  • 分割：2023_10_14T21_09_04.569052
  • 路径：**/details_harness|gsm8k|5_2023-10-14T21-09-04.569052.parquet
  • 分割：latest
  • 路径：**/details_harness|gsm8k|5_2023-10-14T21-09-04.569052.parquet

• harness_hellaswag_10

  • 分割：2023_07_19T18_54_56.836268
  • 路径：**/details_harness|hellaswag|10_2023-07-19T18:54:56.836268.parquet
  • 分割：latest
  • 路径：**/details_harness|hellaswag|10_2023-07-19T18:54:56.836268.parquet

• harness_hendrycksTest_5

  • 分割：2023_07_19T18_54_56.836268
  • 路径：多个路径，例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:54:56.836268.parquet
  • 分割：latest
  • 路径：多个路径，例如 **/details_harness|hendrycksTest-abstract_algebra|5_2023-07-19T18:54:56.836268.parquet