OALL/details_cognitivecomputations__dolphin-2.9.3-llama-3-8b
收藏Hugging Face2024-07-02 更新2024-07-06 收录
下载链接:
https://hf-mirror.com/datasets/OALL/details_cognitivecomputations__dolphin-2.9.3-llama-3-8b
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是在模型 cognitivecomputations/dolphin-2.9.3-llama-3-8b 的评估运行期间自动创建的。数据集由136个配置组成,每个配置对应一个评估任务。数据集是从1次运行中生成的,每次运行在每个配置中表示为特定的分割,分割名称使用运行的时间戳。train 分割始终指向最新的结果。此外,还有一个 results 配置,存储了所有运行的聚合结果。README 还提供了如何使用 Hugging Face datasets 库加载数据集的示例。
The dataset was automatically created during the evaluation run of the model cognitivecomputations/dolphin-2.9.3-llama-3-8b. The dataset is composed of 136 configurations, each corresponding to one of the evaluated tasks. The dataset has been created from 1 run, with each run represented as a specific split in each configuration, named using the timestamp of the run. The train split always points to the latest results. Additionally, there is a results configuration that stores all the aggregated results of the run. The README also provides an example of how to load the dataset using the Hugging Face datasets library.
提供机构:
OALL
原始信息汇总
数据集概述
数据集基本信息
- 名称: Evaluation run of cognitivecomputations/dolphin-2.9.3-llama-3-8b
- 创建方式: 自动创建,用于评估模型 cognitivecomputations/dolphin-2.9.3-llama-3-8b
- 配置数量: 136 个配置,每个配置对应一个评估任务
- 创建次数: 1 次运行
数据集结构
- 分割方式: 每个运行结果作为一个特定的分割,分割名称使用运行的时间戳
- 训练分割: "train" 分割指向最新的结果
- 结果配置: 额外配置 "results" 存储所有聚合的运行结果
数据加载示例
python from datasets import load_dataset data = load_dataset("OALL/details_cognitivecomputations__dolphin-2.9.3-llama-3-8b", "lighteval_xstory_cloze_ar_0", split="train")
最新结果
- 最新运行时间: 2024-07-02T13:10:49.487651
- 结果示例: python { "all": { "acc_norm": 0.40049842866345675, "acc_norm_stderr": 0.037584270717651205, "acc": 0.613500992720053, "acc_stderr": 0.012531219943771486 }, "community|acva:Algeria|0": { "acc_norm": 0.5230769230769231, "acc_norm_stderr": 0.0358596530894741 }, ... }



