Bayesian model average

[Kayhan M66]

Hello

I'm working on the time series problem and made some models with deep learning algorithms like LSTM, GRU, MLP, etc. I need to use the Bayesian model average (BMA) for posterior processing. I heard that BMA is an ensemble model. I search so much on the internet but I couldn't find an appropriate python package to do it. Can I use TensorFlow Probability to implement BMA? if I can, how can I use it?