What is the performance impact of interop with python

I want to use pytorch to train model on gpu and mojo for cpu (mojo to prepare data -> pytorch to train and predict-> loop). Any tips on efficiency?
1 Reply
sora
sora10mo ago
I think if your pre processing is cpu bound I don't see any reason how that can be worse than python. But I guess you will have to experiment and find out.
Want results from more Discord servers?
Add your server