Closed zhizhi-2 closed 2 years ago
Did you use a program with a lot of threads? What system are you on? On Linux, you can increase the capacity with ulimit -n
.
Did you use a program with a lot of threads? What system are you on? On Linux, you can increase the capacity with
ulimit -n
.
Thank you! It is indeed the cause of the problem
Did you use a program with a lot of threads? What system are you on? On Linux, you can increase the capacity with
ulimit -n
.
I have another question, is there a way to convert cfix data types to Python data types ?
Inherently not because cfix only hold value during the runtime but Python types only exist during compilation.
Inherently not because cfix only hold value during the runtime but Python types only exist during compilation. Thank for your reply. Does that mean that if MP-SPDZ is used to implement the secure aggregation of federated learning but the training phase in plaintext is not feasible
A lot of the training functionality isn't implemented for cfix
. However, emulate.x
provides functionality to run the secure computation code generated by the compile in plaintext.
This error occurred at runtime when computing between four parties using shamir-party.x