Published On Apr 28, 2024
Learn how to accelerate local AI projects by connecting to cloud-hosted GPUs, using Elixir's Livebook.
This tutorial describes why Elixir and the BEAM are a great fit for scaled AI tasks, and walks through the process of running the code step-by-step.
Links
- Part 1: • Transcribe Podcasts with Whisper AI &...
- Harness code repository: https://github.com/christianalexander...
- Livebook file: https://github.com/ChristianAlexander...
- The Erlang Movie: • Erlang: The Movie (Fixed Audio)
- Remote GPU post: https://fly.io/phoenix-files/easy-at-...
- Fly.io wireguard setup: https://fly.io/docs/networking/privat...
Timestamps:
Context - 00:00
Why Elixir? - 00:53
How clustering works - 02:10
Setup - 03:19
Accelerated Run - 06:38
Streaming - 06:55
Parallel Streaming - 08:07
Full Feed - 08:41
Summary - 09:00