https://github.com/r-gg/flower-async
Implementation of asynchronous federated learning in flower.
https://github.com/r-gg/flower-async
asynchronous-federated-learning distributed-machine-learning federated-learning flower-framework online-learning
Last synced: 7 months ago
JSON representation
Implementation of asynchronous federated learning in flower.
- Host: GitHub
- URL: https://github.com/r-gg/flower-async
- Owner: r-gg
- License: apache-2.0
- Created: 2024-03-29T20:21:18.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-07-27T12:28:09.000Z (11 months ago)
- Last Synced: 2024-07-27T13:42:48.757Z (11 months ago)
- Topics: asynchronous-federated-learning, distributed-machine-learning, federated-learning, flower-framework, online-learning
- Language: Python
- Homepage:
- Size: 145 KB
- Stars: 7
- Watchers: 2
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# flower-async
Implementation of asynchronous federated learning in flower currently being developed as a part of my master's thesis.
Currently offers the following modes of operation:
- Unweighted - Upon each merge into the global model, the local updates and global model are weighted equally.
- FedAsync - global_new = (1-alpha) * global_old + alpha * local_new
- Where alpha depends on the staleness and/or number of samples used
- AsyncFedED - global_new = global_old + $\eta$ * (local_new - global_old)
- $\eta$ again depends on staleness and/or number of samples
# Implementation
The implementation consists of a modified server, strategy and client manager.
The inspiration of this work stems from the following papers:
- [Privacy-Preserving Asynchronous Federated Learning Mechanism for Edge Network Computing - ](https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9022982)
- [Asynchronous Federated Optimization ~ FedAsync](https://arxiv.org/pdf/1903.03934.pdf)
- [AsyncFedED: Asynchronous Federated Learning with Euclidean Distance based Adaptive Weight Aggregation](https://arxiv.org/pdf/2205.13797.pdf)
- [Federated Learning on Non-IID Data Silos: An Experimental Study](https://arxiv.org/abs/2102.02079)
# Usage
Use the server as typical flower server with one additional argument:
- async_strategy - currently only responsible for aggregating the models.
Moreover the async client manager should be used instead of the SimpleClientManager.
The server instantiation would hence look something like this:
```
server = AsyncServer(
strategy=FedAvg(),
client_manager=AsyncClientManager(),
async_strategy=AsynchronousStrategy())
```