darwinha/details_darwinha__smollm3-dpo-aligned_private
收藏Hugging Face2025-10-17 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/darwinha/details_darwinha__smollm3-dpo-aligned_private
下载链接
链接失效反馈官方服务：
资源简介：
这个数据集是在对模型 [darwinha/smollm3-dpo-aligned](https://huggingface.co/darwinha/smollm3-dpo-aligned) 进行评估运行时自动创建的。数据集由3个配置组成，每个配置对应一个评估任务。数据集是从1次运行中创建的。每个运行都可以在各个配置中的特定分割中找到，分割的名称使用运行的时戳。train 分割始终指向最新结果。还有一个额外的配置 results 存储了所有运行的汇总结果。要从一个运行中加载详细信息，可以执行以下操作：python
from datasets import load_dataset
data = load_dataset("darwinha/details_darwinha__smollm3-dpo-aligned_private",
"results",
split="train")
Dataset automatically created during the evaluation run of model [darwinha/smollm3-dpo-aligned](https://huggingface.co/darwinha/smollm3-dpo-aligned). The dataset is composed of 3 configurations, each one corresponding to one of the evaluated tasks. The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run. The train split is always pointing to the latest results. An additional configuration results stores all the aggregated results of the run. To load the details from a run, you can for instance do the following:
python
from datasets import load_dataset
data = load_dataset("darwinha/details_darwinha__smollm3-dpo-aligned_private",
"results",
split="train")
提供机构：
darwinha
