New optimizer other than SGD in gluon-ts models

[parimuns]

Hello Everyone
I have been using gluont-ts models for probabilistic forecasting. As per my understanding, by default, the models use stochastic gradient descent based optimizers. But I want to use, evolutionary algorithms in place of SGD. Is it possible to integrate different optimizers in gluonts ? If anyone has any idea, please let me know how can we do that in gluonts?