altametris.sara.yolo.exporter
=============================

.. py:module:: altametris.sara.yolo.exporter

.. autoapi-nested-parse::

   YOLO exporter implementation.

   Provides export functionality for YOLO models to various formats
   (ONNX, TensorRT, CoreML, etc.) with validation.

   .. rubric:: Example

   >>> exporter = YoloExporter(model_path="weights/best.pt")
   >>> onnx_path = exporter.export(format="onnx", imgsz=640)
   >>> exporter.validate_export(onnx_path)



Attributes
----------

.. autoapisummary::

   altametris.sara.yolo.exporter.logger


Classes
-------

.. autoapisummary::

   altametris.sara.yolo.exporter.YoloExporter


Module Contents
---------------

.. py:data:: logger

.. py:class:: YoloExporter(model_path: Union[str, pathlib.Path], output_dir: Optional[Union[str, pathlib.Path]] = None)

   Bases: :py:obj:`altametris.sara.core.base_exporter.BaseExporter`


   YOLO exporter implementing BaseExporter interface.

   Supports export to multiple formats via Ultralytics backend.

   :param model_path: Path to YOLO model weights
   :param output_dir: Output directory for exports

   .. rubric:: Example

   >>> exporter = YoloExporter(model_path="yolo11x.pt")
   >>> onnx_path = exporter.export(format="onnx", imgsz=640)
   >>> trt_path = exporter.export(format="tensorrt", imgsz=640, half=True)


   .. py:attribute:: SUPPORTED_FORMATS
      :value: ['onnx', 'torchscript', 'tensorrt', 'coreml', 'tflite', 'engine']



   .. py:method:: export(format: str = 'onnx', imgsz: Union[int, tuple[int, int]] = 640, half: bool = False, dynamic: bool = False, simplify: bool = True, **kwargs: Any) -> pathlib.Path

      Export YOLO model to specified format.

      :param format: Export format
      :param imgsz: Input image size (int or (height, width))
      :param half: Use FP16 precision
      :param dynamic: Dynamic axes for ONNX
      :param simplify: Simplify ONNX model
      :param \*\*kwargs: Additional export arguments

      :returns: Path to exported model

      :raises ExportError: If export fails

      .. rubric:: Example

      >>> # Export to ONNX
      >>> exporter.export(format="onnx", imgsz=640, half=False)
      >>>
      >>> # Export to TensorRT
      >>> exporter.export(format="tensorrt", imgsz=640, half=True)



   .. py:method:: validate_export(export_path: pathlib.Path, test_image: Optional[Union[str, pathlib.Path]] = None, **kwargs: Any) -> bool

      Validate exported model.

      :param export_path: Path to exported model
      :param test_image: Optional test image for validation
      :param \*\*kwargs: Additional validation arguments

      :returns: True if validation passes

      :raises ExportError: If validation fails

      .. rubric:: Example

      >>> onnx_path = exporter.export(format="onnx")
      >>> exporter.validate_export(onnx_path, test_image="test.jpg")



   .. py:method:: export_to_tensorrt(imgsz: Union[int, tuple[int, int]] = 640, half: bool = True, workspace: float = 4.0, **kwargs: Any) -> pathlib.Path

      Export to TensorRT with optimized settings.

      :param imgsz: Input image size
      :param half: Use FP16 precision
      :param workspace: TensorRT workspace size (GB)
      :param \*\*kwargs: Additional TensorRT arguments

      :returns: Path to TensorRT engine

      .. rubric:: Example

      >>> trt_path = exporter.export_to_tensorrt(imgsz=640, half=True)



   .. py:method:: export_to_onnx(imgsz: Union[int, tuple[int, int]] = 640, dynamic: bool = True, simplify: bool = True, **kwargs: Any) -> pathlib.Path

      Export to ONNX with optimized settings.

      :param imgsz: Input image size
      :param dynamic: Dynamic axes for variable input size
      :param simplify: Simplify ONNX graph
      :param \*\*kwargs: Additional ONNX arguments

      :returns: Path to ONNX model

      .. rubric:: Example

      >>> onnx_path = exporter.export_to_onnx(imgsz=640, dynamic=True)



   .. py:method:: __repr__() -> str

      String representation.