Back to search
Top Contributors
49
commits
Handy PyTorch implementation of Federated Learning (for your painless research)
Stars
469
Forks
89
Watchers
469
Open Issues
4
Overall repository health assessment
No package.json found
This might not be a Node.js project
49
commits
correct multithreading thanks to @imguangyu (#16)
a247148View on GitHubfix missing arguments in `_request` and modify logging scheme
bb6ca68View on GitHubfix missing arguments in `_request` and modify logging scheme
4e30636View on GitHubfix missing arguments in `_request` and modify logging scheme
0296e9aView on GitHubremove `lr_scheduler` for evading model copy error when using `.to(device)` inevitably in `central_evaluate`.
b8f9d30View on GitHubhofix: stale optimizer after `central_evaluate` call; due to `.to(device)` operation that copy and replace into a new model reference internally but the optimizer states are fixed from the very first round, not reflecting such change - fixed by not keeping optimizer as an attribute of `Server` object by defining server optimizer as far as it is required, i.e., inside `update` method
c862304View on GitHubhofix: due to the position of eval request in update method of the server, optimistic eval results are reported
8d87017View on GitHub