Flow
Personalized FL framework
A framework for personalized federated learning that creates dynamic models for each input instance and improves generalizability of global models.
Code for "Flow: Per-Instance Personalized Federated Learning" (NeurIPS 2023). Flow addresses the challenge of statistical heterogeneity in federated learning through creating a dynamic personalized model for each input instance through a routing mechanism.
8 stars
1 watching
2 forks
Language: Python
last commit: about 1 year ago Related projects:
| Repository | Description | Stars |
|---|---|---|
 krishnap25/fl_partial_personalization | A framework for federated learning with partial model personalization | 2 |
 omarfoq/knn-per | A federated learning framework with personalized memorization using deep neural networks and k-nearest neighbors for collaborative learning of statistical models | 42 |
 vaseline555/superfed | An open-source implementation of a personalized federated learning method that connects the optima of local and federated models to improve performance. | 28 |
 federatedai/fate-flow | An end-to-end federated learning workflow platform for managing data and models across multiple parties | 52 |
 aiot-mlsys-lab/fedrolex | An approach to heterogeneous federated learning allowing for model training on diverse devices with varying resources. | 61 |
 cuis15/fcfl | An implementation of Fair and Consistent Federated Learning using Python. | 20 |
 fangxiuwen/robust_fl | An implementation of a robust federated learning framework for handling noisy and heterogeneous clients in machine learning. | 41 |
 ibm/fl-arbitrary-participation | Analyzes Federated Learning with Arbitrary Client Participation using various optimization strategies and datasets. | 4 |
 litian96/ditto | A framework for personalized federated learning to balance fairness and robustness in decentralized machine learning systems. | 137 |
 yxdyc/pfedgate | An implementation of a personalized Federated Learning approach with adaptive sparse model adaptation. | 24 |
 royson/fedl2p | This project enables personalized learning models by collaborating on learning the best strategy for each client | 19 |
| omarfoq/fedem | Develops and evaluates federated learning algorithms for personalizing machine learning models across heterogeneous client data distributions. | 154 |
 dawenzi098/sfl-structural-federated-learning | A Python implementation of Personalized Federated Learning with Graph using PyTorch. | 50 |
 substra/substra | Enables the training and validation of machine learning models on distributed datasets in a secure and scalable manner. | 271 |
 hypervoyager/pfl | An implementation of heterogeneous federated learning with parallel edge and server computation | 16 |