---
title: "Discussion on \"Keeping training state when the GPU node changes\" | Hashnode"
description: "Discussion on \"Keeping training state when the GPU node changes\". A controlled Nodus node replacement preserved a checkpoint and advanced training from step 8 to 9. Evidence, timing boundaries and limitations."
url: "https://hashnode.com/posts/keeping-training-state-when-the-gpu-node-changes/6af36ff0293fffa36e514a86"
---

[Feed](https://hashnode.com/feed) /

Discussion

# [Keeping training state when the GPU node changes](https://nodus-compute.hashnode.dev/keeping-training-state-when-the-gpu-node-changes?utm_source=hashnode&utm_medium=feed)

A controlled Nodus node replacement preserved a checkpoint and advanced training from step 8 to 9. Evidence, timing boundaries and limitations.

1d ago · 6 min read · nodus-compute.hashnode.dev  
Read the article

-   [#gpu](https://hashnode.com/tag/gpu)
-   [#machine-learning](https://hashnode.com/tag/machine-learning)
-   [#cloud-computing](https://hashnode.com/tag/cloud-computing)

## Written by

[Viswaretas Kotra](https://hashnode.com/@viswa-kotra)

@viswa-kotra · San Francisco, USA

  
[View profile](https://hashnode.com/@viswa-kotra)

Founder @ Nodus Compute, previously in HFT, Distributed Systems, and AI.

Founder @ Nodus Compute

## Responses

Markdown supported · ⌘↵ to post

No responses yet.