hazyresearch/smoothie_data
收藏Hugging Face2024-12-10 更新2025-04-12 收录
下载链接：
https://hf-mirror.com/datasets/hazyresearch/smoothie_data
下载链接
链接失效反馈官方服务：
资源简介：
# Smoothie Data
This dataset is used for running the experiments in [Smoothie: Label Free Language Model Routing](https://github.com/HazyResearch/smoothie) (NeurIPS 2024).
The dataset contains the train/test data for the tasks used in Smoothie.
It also contains each LLM's generations, as well as the generations selected by Smoothie and other baselines.
For more information on Smoothie, check out our [paper](https://arxiv.org/abs/2412.04692), [blog](https://hazyresearch.stanford.edu/blog/2024-12-10-smoothie), and [code](https://github.com/HazyResearch/smoothie).
提供机构：
hazyresearch
