mlserver_onnx

The mlserver extension is used to run model in onnx format. mlserver doc

struct

onnx/mlseerver_onnx/...
 |
  __init__.py
  onnx.py

place to

how to build

Dockerfile

1. copy directory

$ cp /onnx mlserver/runtimes

2. update Dockerfile

COPY ../onnx ./runtimes/onnx

3. insert in pyproject.toml

mlserver-onnx = {path = "./runtimes/onnx", develop = true}

4. update poetry https://github.com/SeldonIO/MLServer/blob/master/poetry.lock