seldon-core
seldon-core copied to clipboard
Enable WhiteBox explainer v2 (e.g. IntegratedGradients)
In the current design of SC we have the assumption that a model has one ModelUri however in the case of white box explainers we can have upto 2 artifacts (the explainer artifact and the underlying inference model artifact) to load.
I am not sure if there is an easy way to fully support this use case, some alternatives are:
follow explainer server v1patternof assuming that we can only load the inference model) and we always create the explainer on the fly
assume that ModelUri is a directory that would contain both the explainer and the model artifacts.