This package provides a MLServer runtime compatible with alibi-detect models.
You can install the mlserver-alibi-detect
runtime, alongside mlserver
, as:
pip install mlserver mlserver-alibi-detect
For further information on how to use MLServer with Alibi-Detect, you can check out this worked out example.
If no content type is present on the request or metadata, the Alibi-Detect 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.
The Alibi Detect runtime exposes a couple setting flags which can be used to
customise how the runtime behaves.
These settings can be added under the parameters.extra
section of your
model-settings.json
file, e.g.
---
emphasize-lines: 6-8
---
{
"name": "drift-detector",
"implementation": "mlserver_alibi_detect.AlibiDetectRuntime",
"parameters": {
"uri": "./alibi-detect-artifact/",
"extra": {
"batch_size": 5
}
}
}
You can find the full reference of the accepted extra settings for the Alibi Detect runtime below:
.. autopydantic_settings:: mlserver_alibi_detect.runtime.AlibiDetectSettings