NimbusML icon indicating copy to clipboard operation
NimbusML copied to clipboard

Improve inference performance with loaded TransformerChain ML.NET model

Open najeeb-kazmi opened this issue 5 years ago • 0 comments

PR #230 introduced ability to load and score ML.NET models trained in the new ML.NET TransformerChain serialization format. This was done by checking whether "TransformerChain" exists in the archive members. Currently, this is done every time test, predict, predict_proba, and decision_function methods call _predict. This can be improved by checking for "TransformerChain" only once when the model is loaded, improving inference performance.

najeeb-kazmi avatar Nov 20 '19 23:11 najeeb-kazmi