Constructor Detail

• LREMetaData

  LREMetaData()

Method Detail

• getDeviceID

   final static Integer getDeviceID()

  Device ID running Model

• getModelID

   final static String getModelID()

  Model's UUID

• getModelPrecision

   final static String getModelPrecision()

  Model-supported numerical precision used at compile time, e.g., float32, float16, or int8.

  Other data types are possible depending on the target hardware.

• getForgeVersion

   final static String getForgeVersion()

  LEIP SDK version used at compile time.

• getNumberOfOutputs

   final static Integer getNumberOfOutputs()

  Model's number of output tensors.

• getNumberOfInputs

   final static Integer getNumberOfInputs()

  Model's number of input tensors.

• getDeviceType

   final static String getDeviceType()

  Model-supported device type.

• getTarget

   final static String getTarget()

  Model's target compiling info.

  Returns:

  the collection of flags used to specify the target at compile time

• getHost

   final static String getHost()

  Model's host compiling info. The host is the device used to execute the LRE; if no hardware acceleration is present in the system, the host and target may be the same. Typically, the target is the GPU/FPGA/other and the host is the CPU on the system.

• getInputLayouts

   final static String getInputLayouts()

  Model's layout of input tensors. This is meaningful for computer vision models. For these cases, the possible values are NCHW (channel first) or NHWC (channel last).

• getInputShapes

   final static String getInputShapes()

  Model's input tensor shapes. For example: "shape": { "batch": 1, "channel": 3, "width": 640, "height": 640 }

• getOutputShapes

   final static String getOutputShapes()

  Model's output tensor shapes. For example: "shape": { "batch": 1, "candidates": 6300, "classes": 26 }

• getInputDTypes

   final static String getInputDTypes()

  Model's input data types.

• getOutputDTypes

   final static String getOutputDTypes()

  Model's output data type.

• getTRTCalibration

   final static String getTRTCalibration()

  Model's TensorRT calibration tables.

• getOptimization

   final static String getOptimization()

  Model's optimization parameters used at compile time.

• getQuantization

   final static String getQuantization()

  Model's quantization parameters used at compile time.

• getInputWidth

   final static Integer getInputWidth()

  Model's input shape width.

• getInputHeight

   final static Integer getInputHeight()

  Model's input shape height.