MulAddConst¶

Description¶

Info

Parent class: Module

Derived classes: -

This module applies a linear operator to the tensor.

Initializing¶

def __init__(self, a=1.0, b=0.0, inplace=False, name=None):

Parameters

Parameter	Allowed types	Description	Default
a	float	Coefficient $a$ of the linear operator $ax+b$ .	1.0
b	float	Coefficient $b$ of the linear operator $ax+b$ .	0.0
inplace	bool	If True, then the output tensor will be written in the place of the input tensor in memory	False
name	str	Layer name	None

Explanations

inplace - flag showing whether additional memory resources should be allocated for the result. If True, then the output tensor will be written in the place of the input tensor in memory, which can negatively affect the network, if the input tensor takes part in calculations on other branches of the graph.

Examples¶

Necessary imports.

Info

gpuarray is required to properly place the tensor in the GPU

import numpy as np
from PuzzleLib.Backend import gpuarray
from PuzzleLib.Modules import MulAddConst

Info

gpuarray необходим для правильного размещения тензора на GPU

batchsize, maps, h, w = 1, 1, 3, 3
data = gpuarray.to_gpu(np.arange(batchsize * maps * h * w).reshape((batchsize, maps, h, w)).astype(np.float32))
print(data)

[[[[0. 1. 2.]
   [3. 4. 5.]
   [6. 7. 8.]]]]

mulAdd = MulAddConst(a=3.14, b=42.0)
print(mulAdd(data))

[[[[42.       45.14     48.28    ]
   [51.420002 54.56     57.7     ]
   [60.84     63.98     67.12    ]]]]