2. Create Model Inference
ONNX is used in the core of Modelify. All framework models are converted to ONNX format. Modelify provides Model Infrence class to do that. There are 3 required parameters;
- model: your model object
- framework: the framework name (check out here)
- inputs: inputs structure ( check out here)
Modelify has its own input structure and you need to create the inputs with this structure. Basically, it is expected column-based inputs as a sequence if your model input is a data frame. However, Modelify has a helper function that can generate an model schema from pandas dataframe. To get more information, check out the input structures here.
from modelify import ModelInference
from modelify.helpers import create_schema
model_input = create_schema(X_train)
inference = ModelInference(model=model, framework="LIGHTGBM", inputs=model_input)