torch.nn.modules._functions.SyncBatchNorm Class Reference

Inheritance diagram for torch.nn.modules._functions.SyncBatchNorm:

Static Public Member Functions
def	forward (self, input, weight, bias, running_mean, running_var, eps, momentum, process_group, world_size)
def	backward (self, grad_output)
Static Public Member Functions inherited from torch.autograd.function.Function
def	forward (ctx, args, kwargs)
def	backward (ctx, grad_outputs)

Public Attributes
	process_group
	world_size
Public Attributes inherited from torch.autograd.function._ContextMethodMixin
	to_save
	dirty_tensors
	non_differentiable

Additional Inherited Members
Public Member Functions inherited from torch.autograd.function.FunctionMeta
def	__init__ (cls, name, bases, attrs)
Public Member Functions inherited from torch.autograd.function._ContextMethodMixin
def	save_for_backward (self, tensors)
def	mark_dirty (self, args)
def	mark_shared_storage (self, pairs)
def	mark_non_differentiable (self, args)
Static Public Attributes inherited from torch.autograd.function.Function
	is_traceable

Detailed Description

Definition at line 5 of file _functions.py.

---

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

• torch/nn/modules/_functions.py