Run Speech-to-Speech Models on Mac or GPU

Опубликовано: 26 Август 2024
на канале: Trelis Research

2,604

104

➡️ Get Life-time Access to the Trelis Scripts (and future improvements): https://Trelis.com/ADVANCED-transcrip...
➡️ Trelis Runpod Affiliate Link (supports the channel): https://runpod.io?ref=jmfkcdio
➡️ Trelis Vast AI Affiliate Link (supports the channel): https://cloud.vast.ai/?ref_id=98762
➡️ Newsletter: https://blog.Trelis.com
➡️ Trelis Resources/Support/Discord: https://Trelis.com/About

CREDIT: Rohan Sharma for his contribution to this video via the Trelis Internship program (https://trelis.com/internships/). Find Rohan here: https://github.com/rs545837

VIDEO RESOURCES:
HuggingFace Speech-to-Speech Repo: https://github.com/huggingface/speech...
One-click-templates: https://github.com/TrelisResearch/one...
Slides: https://docs.google.com/presentation/...
Llama 3 Paper: https://arxiv.org/abs/2407.21783

TIMESTAMPS:
0:00 Introduction to Speech to Speech AI Models like GPT-4o
0:43 Video Overview
2:02 How to build speech-to-speech models like GPT-4o
6:40 Llama 3 Speech-to-Speech Model
8:54 HuggingFace Speech-to-Speech
13:34 Running speech to speech on your Mac
24:41 Running speech-to-speech on a remote GPU (CUDA)
33:48 Reducing latency with UDP ports instead of TCP
36:02 Video Resources