Federated CSV dataset

I am building a federated learning model, but I want to use my own dataset, which is represented in a CSV file.
any idea how can I use this data for simulation? and consider it distributed?
since all of the tutorials available are working for the predefined datasets


I was searching for the same thing
Can you tell me what you found?

Hello Kareem,

Yes I found an answer that worked for me, while suggests using the function ( tff.simulation.datasets.FilePerUserClientData)

You can have a look at this question in StackOverflow (python - tff.simulation.datasets.ClientData to build federated learning model from CSV files - Stack Overflow)
Wish you all the best

1 Like

Thanks a lot, but if you could share a complete example with colab that would be even better. I mean how did you handle it with different clients with any model.