{"id":17633014,"url":"https://github.com/qinbinli/fedkt","last_synced_at":"2025-05-05T23:50:16.250Z","repository":{"id":194647632,"uuid":"271676958","full_name":"QinbinLi/FedKT","owner":"QinbinLi","description":"Practical One-Shot Federated Learning for Cross-Silo Setting","archived":false,"fork":false,"pushed_at":"2021-05-21T02:24:39.000Z","size":76,"stargazers_count":42,"open_issues_count":1,"forks_count":6,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-04-19T12:14:10.019Z","etag":null,"topics":["federated-learning"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/QinbinLi.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null}},"created_at":"2020-06-12T00:59:19.000Z","updated_at":"2024-11-11T11:20:56.000Z","dependencies_parsed_at":"2023-09-14T12:38:47.730Z","dependency_job_id":"e79a8525-3ac3-4c77-b6c5-c8429d8ea80c","html_url":"https://github.com/QinbinLi/FedKT","commit_stats":null,"previous_names":["qinbinli/fedkt"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QinbinLi%2FFedKT","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QinbinLi%2FFedKT/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QinbinLi%2FFedKT/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/QinbinLi%2FFedKT/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/QinbinLi","download_url":"https://codeload.github.com/QinbinLi/FedKT/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":252596370,"owners_count":21773843,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["federated-learning"],"created_at":"2024-10-23T01:47:10.468Z","updated_at":"2025-05-05T23:50:16.209Z","avatar_url":"https://github.com/QinbinLi.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Practical One-Shot Federated Learning for Cross-Silo Setting\nThis is the code for paper \"Practical One-Shot Federated Learning for Cross-Silo Setting\" [[PDF]](https://arxiv.org/pdf/2010.01017.pdf).\n\n## Dependencies\n* PyTorch 1.6.0\n* torchvision 0.2.2\n* pandas 0.24.2\n* xgboost 1.0.2\n* scikit-learn 0.22.1\n* numpy 1.18.1\n* scipy 1.4.1\n* requests 0.23.0\n\n\n## Sample Scripts\nFedKT on MNIST using a CNN with heterogenous partition and 10 parties: `sh mnist_fedkt.sh`.\n\nFedKT on SVHN using a CNN with heterogenous partition and 10 parties: `sh svhn_fedkt.sh`.\n\n\n\n## Parameters\n\n| Parameter                      | Description                                 |\n| ----------------------------- | ---------------------------------------- |\n| `model`                     | The model architecture. Options: `tree` (random forest), `gbdt_tree`, `mlp`, `simple-cnn`, `vgg-9` .|\n| `alg` | The training algorithm. Options: `fedkt`, `fedavg`, `fedprox`, `scaffold`, `local_training`, `pate`\n| `dataset`      | Dataset to use. Options: `a9a`, `cod-rna`, `mnist`, `celeba`. |\n| `lr` | Learning rate for the local models. |\n| `stu_lr` | Learning rate for the student models and the final model of FedKT. |\n| `batch-size` | Batch size. |\n| `epochs` | Number of local training epochs for FedAvg and FedProx. |\n| `stu_epochs` | Number of training epochs for the models in FedKT. |\n| `n_parties` | Number of parties. |\n| `n_partition` | The number of partition in each party for FedKT. |\n| `n_teacher_each_partition` | The number of teacher models in each partition for FedKT. |\n| `comm_round`    | Number of communication rounds to use in FedAvg and FedProx. |\n| `beta` | The concentration parameter of the Dirichlet distribution for heterogeneous partition. |\n| `mu` | The proximal term parameter for FedProx. |\n| `gamma` | The privacy parameter for FedKT-L1 and FedKT-L2. |\n| `dp_level` | set to 1 to run FedKT-L1 and 2 to run FedKT-L2. |\n| `max_tree_depth` | The tree depth for random forest and gbdt. |\n| `n_stu_trees` | The number of trees for random forest and gbdt. |\n| `datadir` | The path of the dataset. |\n| `logdir` | The path to store the logs. |\n| `device` | Specify the device to run the program. |\n| `seed` | The initial seed. |\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fqinbinli%2Ffedkt","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fqinbinli%2Ffedkt","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fqinbinli%2Ffedkt/lists"}