open-llm-leaderboard/details_togethercomputer__GPT-JT-6B-v1

该数据集是在模型 togethercomputer/GPT-JT-6B-v1 在 Open LLM Leaderboard 上的评估运行期间自动创建的。数据集由 64 个配置组成，每个配置对应一个评估任务。数据集由 2 次运行生成，每次运行在每种配置中表示为特定的拆分，拆分的名称使用运行的时间戳。train 拆分始终指向最新的结果。一个名为 results 的额外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 还提供了一个示例，展示了如何使用 `datasets` 库中的 `load_dataset` 函数加载运行中的详细信息。

This dataset was automatically created during the evaluation run of the model togethercomputer/GPT-JT-6B-v1 on the Open LLM Leaderboard. The dataset comprises 64 configurations, each corresponding to one evaluation task. The dataset is generated from two runs, where each run is represented as a specific split under each configuration, with the split name being the timestamp of the run. The `train` split always points to the most recent results. An additional 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 demonstrating how to use the `load_dataset` function from the `datasets` library to load detailed information from a run.

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 togethercomputer/GPT-JT-6B-v1 的过程中自动创建的，用于 Open LLM Leaderboard。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，每次运行的结果可以在每个配置中找到，并以运行的时间戳命名。每个配置的 "train" 分片始终指向最新的结果。

数据加载示例

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_togethercomputer__GPT-JT-6B-v1", "harness_winogrande_5", split="train")

最新结果

以下是 2023-09-22T13:39:48.520000 运行 的最新结果： python { "all": { "em": 0.00041946308724832214, "em_stderr": 0.00020969854707829363, "f1": 0.04423657718120805, "f1_stderr": 0.0011409456494249344, "acc": 0.3324266847298275, "acc_stderr": 0.00819810174632109 }, "harness|drop|3": { "em": 0.00041946308724832214, "em_stderr": 0.00020969854707829363, "f1": 0.04423657718120805, "f1_stderr": 0.0011409456494249344 }, "harness|gsm8k|5": { "acc": 0.012130401819560273, "acc_stderr": 0.0030152942428909413 }, "harness|winogrande|5": { "acc": 0.6527229676400947, "acc_stderr": 0.013380909249751237 } }

配置详情

数据集包含以下配置：

• 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

以上是数据集的概述和详细信息，包括数据集的创建背景、结构、加载示例、最新结果以及各个配置的详情。