Clone Your Voice For Free Locally!

Author(s): Ishaan Gupta

Originally published on Towards AI.

Clone your voice using openaudio’s S1 mini model on GPU cloud.Credits: Link

You’ve read about OpenAudio S1’s impressive capabilities in my previous article; if not, then check it out using the link below.

Most realistic open-source voice cloning

medium.com

Now you’re ready to get your hands dirty with the actual implementation. In this story, I’ll walk you through setting up OpenAudio S1 on an AWS GPU instance and using it to clone your voice. Before u start, make sure you have access to the model, if not, then request access using below link — https://huggingface.co/fishaudio/openaudio-s1-mini

Let’s start by creating our cloud environment. We’ll use a g5.xlarge instance, which provides the 24GB of VRAM we need for optimal performance.

Log into AWS Console and navigate to EC2Launch an Instance with these specifications:Name: openaudio-s1-instanceInstance type: g5.xlargeAMI: Deep Learning OSS Nvidia Driver AMI GPU PyTorch 2.4.1 (Ubuntu 22.04)Key-Pair: create a key-pair and download it. You’ll need this to login to your EC2Security Group Rules: add 7860 port as custom TCP type to access gradio and 22 as SSH type for SSH into our instance.Storage: 100 GB gp3 (you’ll need the space for model files)

With this config. launch your instance. Then go to your instance info and click on connect button. In there, go to SSH client on top and… 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