bn Class Reference

Inheritance diagram for bn:

Collaboration diagram for bn:

Public Member Functions
def	__init__ (self, output_size, eps=1e-5, momentum=0.1, cross_replica=False, mybn=False)
	Constructor.
def	forward (self, x)
	forward

Public Attributes
	output_size
	output size
	gain
	Prepare gain and bias layers.
	bias
	bias
	eps
	epsilon to avoid dividing by 0
	momentum
	Momentum.
	cross_replica
	Use cross-replica batchnorm?
	mybn
	Use my batchnorm?

Detailed Description

Normal, non-class-conditional BN

Definition at line 460 of file ieagan.py.

Constructor & Destructor Documentation

__init__()

def __init__	(		self,
			output_size,
			eps = 1e-5,
			momentum = 0.1,
			cross_replica = False,
			mybn = False
	)

Constructor.

Definition at line 466 of file ieagan.py.

    ):
        super(bn, self).__init__()
        
        self.output_size = output_size
        
        self.gain = P(torch.ones(output_size), requires_grad=True)
        
        self.bias = P(torch.zeros(output_size), requires_grad=True)
        
        self.eps = eps
        
        self.momentum = momentum
        
        self.cross_replica = cross_replica
        
        self.mybn = mybn
 
        if mybn:
            self.bn = myBN(output_size, self.eps, self.momentum)
        # Register buffers if neither of the above
        else:
            self.register_buffer("stored_mean", torch.zeros(output_size))
            self.register_buffer("stored_var", torch.ones(output_size))
 

Member Function Documentation

forward()

def forward	(		self,
			x
	)

forward

Definition at line 498 of file ieagan.py.

    def forward(self, x):
        if self.mybn:
            gain = self.gain.view(1, -1, 1, 1)
            bias = self.bias.view(1, -1, 1, 1)
            return self.bn(x, gain=gain, bias=bias)
        else:
            return F.batch_norm(
                x,
                self.stored_mean,
                self.stored_var,
                self.gain,
                self.bias,
                self.training,
                self.momentum,
                self.eps,
            )
 
 

Member Data Documentation

bias

bias

bias

Definition at line 480 of file ieagan.py.

bn

bn

Definition at line 491 of file ieagan.py.

cross_replica

cross_replica

Use cross-replica batchnorm?

Definition at line 486 of file ieagan.py.

eps

eps

epsilon to avoid dividing by 0

Definition at line 482 of file ieagan.py.

gain

gain

Prepare gain and bias layers.

Definition at line 478 of file ieagan.py.

momentum

momentum

Momentum.

Definition at line 484 of file ieagan.py.

mybn

mybn

Use my batchnorm?

Definition at line 488 of file ieagan.py.

output_size

output_size

output size

Definition at line 476 of file ieagan.py.

---

The documentation for this class was generated from the following file:

• pxd/scripts/pxd/background_generator/models/ieagan.py