FATE
FATE copied to clipboard
FederatedAI
An Industrial Grade Federated Learning Framework
(venv) [root@273deb40ece3 fate_flow]# python fate_flow_client.py -f bind -c examples/bind_model_service.json { "retcode": 100, "retmsg": "" } 这是部署在云服务器上的，也关闭了防火墙了
从[官方文档](https://fate.readthedocs.io/en/latest/federatedml_component/logistic_regression/#heterogeneous-sshe-logistic-regression)中得知，FATE 实现了一种基于 SS+HE 的混合纵向 LR 训练协议，在阅读协议流程图中，有以下两点问题想请教下： 1. SS+HE 的方案中，使用的损失函数是什么？从残差的计算方式中，猜测应该不是 Taylor Loss 2. 在 `forward` 流程中，Host Party 为何要将 `1` 同态加密后发送给 Guest Party? 仅仅是为了方便 Guest Party 计算出完整的密态残差吗？因为如果 Host Party 的 `1` 保留在本地，那么后续这两方可以分别计算...
报错信息如下： (venv) [app@vm-0-2-centos conf]$ flow test toy -gid 9999 -hid 10000 { "jobId": "202203221731177821370", "retcode": 103, "retmsg": "Traceback (most recent call last):\n File \"/data/projects/fate/fateflow/python/fate_flow/scheduler/dag_scheduler.py\", line 133, in submit\n raise Exception(\"create...
在[算法组建文档](https://github.com/FederatedAI/FATE/blob/master/doc/federatedml_component/README.zh.md)中看到有提供隐私并集计算，但没有找到相关具体实现，请问一下具体实现在哪里？ 
In hetero-LR settings, a role arbiter is needed. I would like to know which party should I assign this role. In the tutorial, some choose the guest to be the...
**Is your feature request related to a problem? Please describe.** PSI supports ecdh protocol, which may be more efficient when network bandwidth is limited.
有两个问题请教专家，还请专家指点，感谢答疑。 （1）FATE1.7.2，怎么将一个已训练好的横向secureboost模型，将其返回成一个Booster类的模型呢？ （2）FATE1.7.2，怎么将一个已训练好的横向secureboost模型转化成lightgbm模型，并返回一个Booster类的模型呢？
In some feature engineering module, judge a feature has been added or not is slow, some other operations may also have the same problem. This will be optimize in Fate-v1.9.0
**Describe the bug** To do woe transform in guest site, when training data does not have missing value, but predicting data has. Error will occur as "index of bounds", which...
**Describe the bug** In hetero nn algorithm, if we use selective back propagation but batch_size is not small, error will occur. This will be fixed in fate-v1.9
Metadata
Owner
Metadata
An Industrial Grade Federated Learning Framework