Setting up the GPU-Based LLM Training Machine

Author(s): Leo Tisljaric, PhD

Originally published on Towards AI.

Setup guide for your local Ubuntu-based machine for training PyTorch and TensorFlow AI models using GPU acceleration

Supercomputer (Image by: Author; Source: OpenAI DALL-E)

Installing all the necessary tools and drivers on your local machine can be very frustrating. Especially when you need to track compatibilities and dependencies of different tools. In this article, you will find a guide to set up your local Ubuntu-based machine for training AI models using GPU acceleration.

Before we continue, give me a second of your time. If you want to support my work, you can do it through a secure PayPal link :

Go to paypal.me/tisljaricleo and type in the amount. Since it’s PayPal, it’s easy and secure. Don’t have a PayPal…

paypal.me

You can use this article as a knowledge base or as a learning material that will help you grasp the quite complicated GPU-based machine setup.

Contents:

Assumptions & LimitationsPreparationsInstallation — CUDA & CUDNNTest Installation

This article will show you how to setup the machine with several assumptions & limitations:

You are using Ubuntu (Ubuntu 22.04).You have Nvidia GPU/s installed on your machine.You have sudo privileges on your machine.

So, if you are aware of the mentioned assumptions, you will find this article useful. Let’s start!

Check GPU compatibility and compute capability:

CUDA compute… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI