core_ml 0.1.0

Flutter bindings for Apple's Core ML framework: compile and load models, inspect their inputs and outputs, and run single, batch and stateful predictions.

0.1.0

• Initial release: Core ML bindings for iOS 18+ and macOS 15+ over Pigeon platform channels.
• MLModel: runtime compilation of .mlmodel / .mlpackage with an optional persistent destination, loading compiled .mlmodelc models and Flutter assets, MLModelAsset function names and descriptions.
• MLModelConfiguration: compute units, low-precision GPU accumulation, function name, display name and optimization hints.
• MLModelDescription: input, output, state and training features with multi-array, image, dictionary, sequence and state constraints, metadata, class labels, predicted feature names and isUpdatable.
• Predictions with every MLFeatureValue type: MLMultiArray (float16, float32, float64, int32, int8 as raw bytes with shape and strides), images from pixels, encoded bytes or files (resized to the input constraint), strings, int64s, doubles, dictionaries, sequences and undefined values.
• Async prediction, batch prediction and stateful prediction with MLState read/write access.
• CoreML.availableComputeDevices / allComputeDevices, PNG/JPEG export of image outputs, stable error codes for MLModelError.