module `onnxrt.ops_cpu.op_max_pool_`#

Short summary#

module mlprodict.onnxrt.ops_cpu.op_max_pool_

Implements runtime for operator MaxPool. The code is inspired from pool.cc in onnxruntime.

class	truncated documentation
`MaxPoolDouble`	Implements float runtime for operator Conv. The code is inspired from pool.cc …
`MaxPoolFloat`	Implements float runtime for operator Conv. The code is inspired from pool.cc …

Implements runtime for operator MaxPool. The code is inspired from pool.cc in onnxruntime.

class mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolDouble(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolDouble)#

Bases: pybind11_object

Implements float runtime for operator Conv. The code is inspired from pool.cc in onnxruntime. Supports double only.

__init__(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolDouble) → None#

compute(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolDouble, arg0: numpy.ndarray[numpy.float64]) → tuple#: Computes the output for operator MaxPool.

init(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolDouble, arg0: str, arg1: numpy.ndarray[numpy.int64], arg2: int, arg3: int, arg4: numpy.ndarray[numpy.int64], arg5: numpy.ndarray[numpy.int64], arg6: numpy.ndarray[numpy.int64]) → None#: Initializes the runtime with the ONNX attributes.

class mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolFloat(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolFloat)#

Bases: pybind11_object

Implements float runtime for operator Conv. The code is inspired from pool.cc in onnxruntime. Supports float only.

__init__(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolFloat) → None#

compute(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolFloat, arg0: numpy.ndarray[numpy.float32]) → tuple#: Computes the output for operator MaxPool.

init(self: mlprodict.onnxrt.ops_cpu.op_max_pool_.MaxPoolFloat, arg0: str, arg1: numpy.ndarray[numpy.int64], arg2: int, arg3: int, arg4: numpy.ndarray[numpy.int64], arg5: numpy.ndarray[numpy.int64], arg6: numpy.ndarray[numpy.int64]) → None#: Initializes the runtime with the ONNX attributes.