An MLModel encapsulates a Core ML model's prediction methods, configuration, and model description. You can use the coremltools package to convert trained models from a variety of training tools into Core ML models. For the full list of model types, see Model.
The converters in coremltools return a converted model as an
MLModel object. You can then save the
MLModel as an
.mlmodel file, use it to make predictions, modify the input and output descriptions, and set the model's metadata. The MLModel class defines the minimal interface to a Core ML model in Python.
A key component of Core ML is the public specification for representing machine learning models. This specification is defined in the Model.proto protobuf (for a more readable version, see Model), which can be created using any language supported by protobuf (such as Python, C++, Java, C#, and Perl).
At a high level, the protobuf specification consists of the following:
- Model description: The model name and and the types of input and output.
- Model parameters: The set of parameters required to represent a specific instance of the model.
- Metadata: Information about the origin, license, and author of the model.
The following code snippets demonstrate how to use an MLModel. To use these code snippets, first import
coremltools as follows:
import coremltools as ct
# Load the MLModel mlmodel = ct.models.MLModel('path/to/the/model.mlmodel') # Save the MLModel mlmodel.save('path/to/the/saved/model.mlmodel')
# Use the MLModel for prediction mlmodel.predict(...)
spec object is the parsed protobuf object of the Core ML model. The following code snippets show how you can get the spec from the MLModel, print its description, get the type of MLModel, and save the MLModel directly from the
spec. It also shows how to convert the
spec to MLModel and compile the model in one step:
# Get the spec from the MLModel spec = mlmodel.get_spec() # Print the input/output description for the MLModel print(spec.description) # Get the type of MLModel (NeuralNetwork, SupportVectorRegressor, Pipeline etc) print(spec.WhichOneof('Type')) # Save out the MLModel directly from the spec ct.models.utils.save_spec(spec, 'path/to/the/saved/model.mlmodel') # Convert spec to MLModel. This step also compiles the model. mlmodel = ct.models.MLModel(spec) # Load the spec from the saved .mlmodel file directly spec = ct.models.utils.load_spec('path/to/the/model.mlmodel')
The following example converts a model from Scikit-learn. The model predicts the price of a house based on three features (bedroom, bath, size).
The example shows how you can use the MLModel object returned by the conversion to update the metadata and input/output descriptions that are displayed in the Xcode preview. (For details about the Xcode preview, see Xcode Model Preview Types.
from sklearn.linear_model import LinearRegression import pandas as pd # Load data data = pd.read_csv('houses.csv') # Train a model model = LinearRegression() model.fit(data[["bedroom", "bath", "size"]], data["price"]) # Convert and save the scikit-learn model import coremltools as ct model = ct.converters.sklearn.convert(model, ["bedroom", "bath", "size"], "price") # Set model metadata model.author = 'John Smith' model.license = 'BSD' model.short_description = 'Predicts the price of a house in the Seattle area.' model.version = '1' # Set feature descriptions manually model.input_description['bedroom'] = 'Number of bedrooms' model.input_description['bathrooms'] = 'Number of bathrooms' model.input_description['size'] = 'Size (in square feet)' # Set the output descriptions model.output_description['price'] = 'Price of the house' # Save the model model.save('HousePricer.mlmodel')
Updated about a month ago