FATE
FATE copied to clipboard
An Industrial Grade Federated Learning Framework
搭建完成fate之后跑toy,抛出 
任务配置文件: ``` json { "dsl_version": 2, "initiator": { "role": "guest", "party_id": 9999 }, "role": { "host": [ 10000 ], "guest": [ 9999 ] }, "component_parameters": { "role": { "guest": {...
[ERROR] [2022-08-03 15:03:02,609] [202208031501390085770] [2852:4727107008] - [task_executor._run_] [line:243]: 'x9' Traceback (most recent call last): File "../FATE-1.7.2/python/fate_flow/worker/task_executor.py", line 195, in _run_ cpn_output = run_object.run(cpn_input) File "../FATE-1.8.0/python/federatedml/model_base.py", line 236, in run self._run(cpn_input=cpn_input)...
请问 python/fate/ml/nn/test/test_fedpass_lenet.py 是否是论文 FedPass: Privacy-Preserving Vertical Federated Deep Learning with Adaptive Obfuscation 中的算法实现?  如果是的话,有下面几个问题: 1. test_fedpass_lenet.py 似乎没有 passive party 对 feature 的 obfuscation 2. fedpass 实现在 agg_layer 中,是在 host(passive方)经过...
版本:FATE 2.0.0 对于组件的输入(dataframe_input),有些组件用的参数名是input_data(像psi、statistics、sample、datasplit等),有些用的是train_data(binning、scale、各建模算法)。 对于组件的输出(dataframe_output),nn和secureboost组件用的是train_data_output,其他组件用的都是train_output_data。 请问是否有考虑过对参数的命名进行统一呢,虽然组件输入的命名不同还好理解,但组件输出这个感觉很容易搞错……
**Issue Description:** Hello. I have discovered a performance degradation in the .loc function of pandas version 2.0.3 when .loc handling big DataFrame with non-unique indexes. When using pandas more than...
如图,fate这边会抛spark exception:org.apache.spark.shuffle.MetadataFetchFailedException: Missing an output location for shuffle 0。而在spark-ui上看的话,任务都是成功的。 请问有人碰到过类似的问题吗。求解答。 环境:Centos7 + FATE v1.11.0 + Spark + RabbitMQ。
Hello, What are the available metrics for early stopping in Logistic Regression model ? Do we currently have 'AUC' and 'ks' only ?
在2.X API接口文档中只看见了模型绑定接口,没有看见模型部署和模型加载接口, 并且模型绑定接口没有给参数说明