uffebjorklund / OnnxModelDemo

ML.NET load ONNX model an test from WebAPI

ONNX C# Test

A tiny proof of concept with ML.NET

Idea

To load a trained ONNX model into ML.NET and then test the prediction engine by passing in each row of the inferenceData.csv file to the Predition endpoint.

Instructions

Build & Run

dotnet build
dotnet run

Test

Take a line from the inferenceData.csv file and send as an array to the endpoint. The model requires an array of floats with 20 items, more or less will cause an error

Example...

curl -X 'POST' \
  'https://localhost:5001/Prediction' \
  -H 'accept: text/plain' \
  -H 'Content-Type: application/json' \
  -d '[
  -0.011430789,-0.072804466,-0.072202496,0.010752341,10.005615,9.971158,10.071351,9.941953,-0.095031515,-0.082216434,-0.06416558,-0.060096428,10.077214,9.920494,9.9190235,-0.035336476,-0.08147322,0.043970365,-0.02370813,9.980851
]'