explain device equality in README

d78d5b201a17414f1a5b8a1c8b036a8b531a8147 · 2024-07-16 02:05:50 -0700 · Alex Cheema

Files touched

M README.md

Diff

commit d78d5b201a17414f1a5b8a1c8b036a8b531a8147
Author: Alex Cheema <alexcheema123@gmail.com>
Date:   Tue Jul 16 02:05:50 2024 -0700

    explain device equality in README
---
 README.md | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/README.md b/README.md
index ca240188..054c2532 100644
--- a/README.md
+++ b/README.md
@@ -45,6 +45,17 @@ exo will automatically discover other devices using the best method available. Z
 
 exo provides a ChatGPT-compatible API for running models. It's a one-line change in your application to run models on your own hardware using exo.
 
+### Device Equality
+
+Unlike other distributed inference frameworks, exo does not use a master-worker architecture. Instead, exo devices connect p2p. As long as a device is connected somewhere in the network, it can be used to run models.
+
+Exo supports different partitioning strategies to split up a model across devices. The default partitioning strategy is [ring memory weighted partitioning](topology/ring_memory_weighted_partitioning.py). This runs an inference in a ring where each device runs a number of model layers proportional to the memory of the device.
+
+<picture>
+  <img alt="ring topology" src="docs/ring-topology.png" width="50%" height="50%">
+</picture>
+
+
 ## Installation
 
 The current recommended way to install exo is from source.

← 94b6a249 print debug only · back to Exo · ring topology image 231cde5f →