alihan/exo
1
0
Fork 0
mirror of https://github.com/exo-explore/exo.git synced 2025-10-23 02:57:14 +03:00
Code Issues Packages Projects Releases Wiki Activity

Updated node refs

Browse Source
This commit is contained in:
Nel Nibcord
2024-12-11 02:54:01 -08:00
parent 59af2dd592
commit 763fbf8486
4 changed files with 7 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
README.md
View File

@@ -46,7 +46,7 @@ exo [optimally splits up models](exo/topology/ring_memory_weighted_partitioning_
### Automatic Device Discovery
exo will [automatically discover](https://github.com/exo-explore/exo/blob/945f90f676182a751d2ad7bcf20987ab7fe0181e/exo/orchestration/standard_node.py#L154) other devices using the best method available. Zero manual configuration.
exo will [automatically discover](https://github.com/exo-explore/exo/blob/945f90f676182a751d2ad7bcf20987ab7fe0181e/exo/orchestration/node.py#L154) other devices using the best method available. Zero manual configuration.
### ChatGPT-compatible API
@@ -54,7 +54,7 @@ exo provides a [ChatGPT-compatible API](exo/api/chatgpt_api.py) for running mode
### Device Equality
Unlike other distributed inference frameworks, exo does not use a master-worker architecture. Instead, exo devices [connect p2p](https://github.com/exo-explore/exo/blob/945f90f676182a751d2ad7bcf20987ab7fe0181e/exo/orchestration/standard_node.py#L161). As long as a device is connected somewhere in the network, it can be used to run models.
Unlike other distributed inference frameworks, exo does not use a master-worker architecture. Instead, exo devices [connect p2p](https://github.com/exo-explore/exo/blob/945f90f676182a751d2ad7bcf20987ab7fe0181e/exo/orchestration/node.py#L161). As long as a device is connected somewhere in the network, it can be used to run models.
Exo supports different [partitioning strategies](exo/topology/partitioning_strategy.py) to split up a model across devices. The default partitioning strategy is [ring memory weighted partitioning](exo/topology/ring_memory_weighted_partitioning_strategy.py). This runs an inference in a ring where each device runs a number of model layers proportional to the memory of the device.