This package provides a MLServer runtime compatible with CatBoost's CatboostClassifier
.
You can install the runtime, alongside mlserver
, as:
pip install mlserver mlserver-catboost
For further information on how to use MLServer with CatBoost, you can check out this worked out example.
If no content type is present on the request or metadata, the CatBoost runtime will try to decode the payload as a NumPy Array. To avoid this, either send a different content type explicitly, or define the correct one as part of your model's metadata.