Peter Hausel

pk11 ·

OK, just to update this thread: it seems the above is definitely the wrong track, since that code path is mainly about how to populate train_set_metadata, not about how to convert input string into tensors. Right now investigating

I’m not an expert in the matter, but if latency is paramount, have you considered serving with c++? The amount of work for implementing the preprocessing and postprocessing should be the same in c++ and Java.

good point on sug...

Thanks for the detailed explanation. I will try to create a prototype handling ludwig model serving with java(using java tf). For additional context: in our case, we would need to deploy models in java services due to latency requirements. Similar to y...

Hi, great project! Thanks for open sourcing it!

There is a discussion on model serving here this thread also offers some insights into pre and post processing. It sounds like the SavedModel as well as the mapping between user input->tensors(train_set...