# How to use raytune to do distributed hyper-parameter tuning?

**URL:** <https://discuss.huggingface.co/t/how-to-use-raytune-to-do-distributed-hyper-parameter-tuning/14695>\
**Category:** 🤗Transformers\
**Created:** [February 14, 2022, 2:45am UTC](https://discuss.huggingface.co/t/how-to-use-raytune-to-do-distributed-hyper-parameter-tuning/14695 "2022-02-14T02:45:41Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![wxp](https://avatars.discourse-cdn.com/v4/letter/w/13edae/32.png) [@wxp](https://discuss.huggingface.co/u/wxp)\
**Post date:** [February 14, 2022, 2:45am UTC](https://discuss.huggingface.co/t/how-to-use-raytune-to-do-distributed-hyper-parameter-tuning/14695/1 "2022-02-14T02:45:41Z")

</div>

Does anyone have any experience with this? I want to launch a hyper-parameter tuning job on multiple nodes, each node has 8 GPUs. The resource for each trial is {“gpu”:8}