diff --git a/docs/master/_modules/torch/nn/modules/adaptive.html b/docs/master/_modules/torch/nn/modules/adaptive.html
index ee18b38d8639..58fbe162e673 100644
--- a/docs/master/_modules/torch/nn/modules/adaptive.html
+++ b/docs/master/_modules/torch/nn/modules/adaptive.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/batchnorm.html b/docs/master/_modules/torch/nn/modules/batchnorm.html
index 4eaea1f9f44c..8c348df1301b 100644
--- a/docs/master/_modules/torch/nn/modules/batchnorm.html
+++ b/docs/master/_modules/torch/nn/modules/batchnorm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/container.html b/docs/master/_modules/torch/nn/modules/container.html
index e528c6b105f1..0961486db29c 100644
--- a/docs/master/_modules/torch/nn/modules/container.html
+++ b/docs/master/_modules/torch/nn/modules/container.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/conv.html b/docs/master/_modules/torch/nn/modules/conv.html
index bd52fd244587..cf23c26c30b0 100644
--- a/docs/master/_modules/torch/nn/modules/conv.html
+++ b/docs/master/_modules/torch/nn/modules/conv.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/distance.html b/docs/master/_modules/torch/nn/modules/distance.html
index 786f0b409c7f..cecd7cb417a8 100644
--- a/docs/master/_modules/torch/nn/modules/distance.html
+++ b/docs/master/_modules/torch/nn/modules/distance.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/dropout.html b/docs/master/_modules/torch/nn/modules/dropout.html
index 72f42054b1bf..ef97461ce1db 100644
--- a/docs/master/_modules/torch/nn/modules/dropout.html
+++ b/docs/master/_modules/torch/nn/modules/dropout.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/flatten.html b/docs/master/_modules/torch/nn/modules/flatten.html
index 8b3f4cce10ef..6462f887f5e0 100644
--- a/docs/master/_modules/torch/nn/modules/flatten.html
+++ b/docs/master/_modules/torch/nn/modules/flatten.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/fold.html b/docs/master/_modules/torch/nn/modules/fold.html
index c18f1cdd552f..2e5b6369a273 100644
--- a/docs/master/_modules/torch/nn/modules/fold.html
+++ b/docs/master/_modules/torch/nn/modules/fold.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/instancenorm.html b/docs/master/_modules/torch/nn/modules/instancenorm.html
index 614b012335ef..e53bd053f0cf 100644
--- a/docs/master/_modules/torch/nn/modules/instancenorm.html
+++ b/docs/master/_modules/torch/nn/modules/instancenorm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/linear.html b/docs/master/_modules/torch/nn/modules/linear.html
index 71b5ae011a78..bfac7e543219 100644
--- a/docs/master/_modules/torch/nn/modules/linear.html
+++ b/docs/master/_modules/torch/nn/modules/linear.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/loss.html b/docs/master/_modules/torch/nn/modules/loss.html
index a1e80a4e334b..f1d916de9233 100644
--- a/docs/master/_modules/torch/nn/modules/loss.html
+++ b/docs/master/_modules/torch/nn/modules/loss.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/module.html b/docs/master/_modules/torch/nn/modules/module.html
index c21745ee83a5..9bcd8f1caf06 100644
--- a/docs/master/_modules/torch/nn/modules/module.html
+++ b/docs/master/_modules/torch/nn/modules/module.html
@@ -158,7 +158,7 @@
return handle
-[docs] class Module :
+
class Module :
r """Base class for all neural network modules.
Your models should also subclass this class.
@@ -580,7 +580,7 @@
Source code for torch.nn.modules.module
"""
forward : Callable [ ... , Any ] = _forward_unimplemented
-[docs] def register_buffer ( self , name : str , tensor : Optional [ Tensor ], persistent : bool = True ) -> None :
+
def register_buffer ( self , name : str , tensor : Optional [ Tensor ], persistent : bool = True ) -> None :
r """Adds a buffer to the module.
This is typically used to register a buffer that should not to be
@@ -630,9 +630,9 @@
Source code for torch.nn.modules.module
if persistent :
self . _non_persistent_buffers_set . discard ( name )
else :
- self . _non_persistent_buffers_set . add ( name )
+
self . _non_persistent_buffers_set . add ( name )
-
[docs] def register_parameter ( self , name : str , param : Optional [ Parameter ]) -> None :
+
def register_parameter ( self , name : str , param : Optional [ Parameter ]) -> None :
r """Adds a parameter to the module.
The parameter can be accessed as an attribute using given name.
@@ -669,9 +669,9 @@
Source code for torch.nn.modules.module
"as a function of another Tensor, compute the value in "
"the forward() method." . format ( name ))
else :
- self . _parameters [ name ] = param
+
self . _parameters [ name ] = param
-
[docs] def add_module ( self , name : str , module : Optional [ 'Module' ]) -> None :
+
def add_module ( self , name : str , module : Optional [ 'Module' ]) -> None :
r """Adds a child module to the current module.
The module can be accessed as an attribute using the given name.
@@ -693,7 +693,7 @@
Source code for torch.nn.modules.module
raise KeyError ( "module name can't contain \" . \" " )
elif name == '' :
raise KeyError ( "module name can't be empty string \"\" " )
- self . _modules [ name ] = module
+
self . _modules [ name ] = module
def _apply ( self , fn ):
for module in self . children ():
@@ -744,7 +744,7 @@
Source code for torch.nn.modules.module
return self
-[docs] def apply ( self : T , fn : Callable [[ 'Module' ], None ]) -> T :
+
def apply ( self : T , fn : Callable [[ 'Module' ], None ]) -> T :
r """Applies ``fn`` recursively to every submodule (as returned by ``.children()``)
as well as self. Typical use includes initializing the parameters of a model
(see also :ref:`nn-init-doc`).
@@ -785,9 +785,9 @@
Source code for torch.nn.modules.module
for module in self . children ():
module . apply ( fn )
fn ( self )
- return self
+
return self
-
[docs] def cuda ( self : T , device : Optional [ Union [ int , device ]] = None ) -> T :
+
def cuda ( self : T , device : Optional [ Union [ int , device ]] = None ) -> T :
r """Moves all model parameters and buffers to the GPU.
This also makes associated parameters and buffers different objects. So
@@ -801,17 +801,17 @@
Source code for torch.nn.modules.module
Returns:
Module: self
"""
- return self . _apply ( lambda t : t . cuda ( device ))
+
return self . _apply ( lambda t : t . cuda ( device ))
-
[docs] def cpu ( self : T ) -> T :
+
def cpu ( self : T ) -> T :
r """Moves all model parameters and buffers to the CPU.
Returns:
Module: self
"""
-
return self . _apply ( lambda t : t . cpu ())
+
return self . _apply ( lambda t : t . cpu ())
-
[docs] def type ( self : T , dst_type : Union [ dtype , str ]) -> T :
+
def type ( self : T , dst_type : Union [ dtype , str ]) -> T :
r """Casts all parameters and buffers to :attr:`dst_type`.
Arguments:
@@ -820,39 +820,39 @@
Source code for torch.nn.modules.module
Returns:
Module: self
"""
- return self . _apply ( lambda t : t . type ( dst_type ))
+
return self . _apply ( lambda t : t . type ( dst_type ))
-
[docs] def float ( self : T ) -> T :
+
def float ( self : T ) -> T :
r """Casts all floating point parameters and buffers to float datatype.
Returns:
Module: self
"""
-
return self . _apply ( lambda t : t . float () if t . is_floating_point () else t )
+
return self . _apply ( lambda t : t . float () if t . is_floating_point () else t )
-
[docs] def double ( self : T ) -> T :
+
def double ( self : T ) -> T :
r """Casts all floating point parameters and buffers to ``double`` datatype.
Returns:
Module: self
"""
-
return self . _apply ( lambda t : t . double () if t . is_floating_point () else t )
+
return self . _apply ( lambda t : t . double () if t . is_floating_point () else t )
-
[docs] def half ( self : T ) -> T :
+
def half ( self : T ) -> T :
r """Casts all floating point parameters and buffers to ``half`` datatype.
Returns:
Module: self
"""
-
return self . _apply ( lambda t : t . half () if t . is_floating_point () else t )
+
return self . _apply ( lambda t : t . half () if t . is_floating_point () else t )
-
[docs] def bfloat16 ( self : T ) -> T :
+
def bfloat16 ( self : T ) -> T :
r """Casts all floating point parameters and buffers to ``bfloat16`` datatype.
Returns:
Module: self
"""
-
return self . _apply ( lambda t : t . bfloat16 () if t . is_floating_point () else t )
+
return self . _apply ( lambda t : t . bfloat16 () if t . is_floating_point () else t )
@overload
def to ( self : T , device : Optional [ Union [ int , device ]] = ... , dtype : Optional [ Union [ dtype , str ]] = ... ,
@@ -867,7 +867,7 @@
Source code for torch.nn.modules.module
def to ( self : T , tensor : Tensor , non_blocking : bool = ... ) -> T :
...
-[docs] def to ( self , * args , ** kwargs ):
+
def to ( self , * args , ** kwargs ):
r """Moves and/or casts the parameters and buffers.
This can be called as
@@ -950,9 +950,9 @@
Source code for torch.nn.modules.module
return t . to ( device , dtype if t . is_floating_point () else None , non_blocking , memory_format = convert_to_format )
return t . to ( device , dtype if t . is_floating_point () else None , non_blocking )
- return self . _apply ( convert )
+
return self . _apply ( convert )
-
[docs] def register_backward_hook (
+
def register_backward_hook (
self , hook : Callable [[ 'Module' , _grad_t , _grad_t ], Union [ None , Tensor ]]
) -> RemovableHandle :
r """Registers a backward hook on the module.
@@ -986,9 +986,9 @@
Source code for torch.nn.modules.module
"""
handle = hooks . RemovableHandle ( self . _backward_hooks )
self . _backward_hooks [ handle . id ] = hook
- return handle
+
return handle
-
[docs] def register_forward_pre_hook ( self , hook : Callable [ ... , None ]) -> RemovableHandle :
+
def register_forward_pre_hook ( self , hook : Callable [ ... , None ]) -> RemovableHandle :
r """Registers a forward pre-hook on the module.
The hook will be called every time before :func:`forward` is invoked.
@@ -1009,9 +1009,9 @@
Source code for torch.nn.modules.module
"""
handle = hooks . RemovableHandle ( self . _forward_pre_hooks )
self . _forward_pre_hooks [ handle . id ] = hook
- return handle
+
return handle
-
[docs] def register_forward_hook ( self , hook : Callable [ ... , None ]) -> RemovableHandle :
+
def register_forward_hook ( self , hook : Callable [ ... , None ]) -> RemovableHandle :
r """Registers a forward hook on the module.
The hook will be called every time after :func:`forward` has computed an output.
@@ -1032,7 +1032,7 @@
Source code for torch.nn.modules.module
"""
handle = hooks . RemovableHandle ( self . _forward_hooks )
self . _forward_hooks [ handle . id ] = hook
- return handle
+
return handle
def _slow_forward ( self , * input , ** kwargs ):
tracing_state = torch . _C . _get_tracing_state ()
@@ -1220,7 +1220,7 @@
Source code for torch.nn.modules.module
def state_dict ( self , prefix : str = ... , keep_vars : bool = ... ) -> Dict [ str , Tensor ]:
...
-[docs] def state_dict ( self , destination = None , prefix = '' , keep_vars = False ):
+
def state_dict ( self , destination = None , prefix = '' , keep_vars = False ):
r """Returns a dictionary containing a whole state of the module.
Both parameters and persistent buffers (e.g. running averages) are
@@ -1248,7 +1248,7 @@
Source code for torch.nn.modules.module
hook_result = hook ( self , destination , prefix , local_metadata )
if hook_result is not None :
destination = hook_result
- return destination
+
return destination
def _register_load_state_dict_pre_hook ( self , hook ):
r """These hooks will be called with arguments: `state_dict`, `prefix`,
@@ -1336,7 +1336,7 @@
Source code for torch.nn.modules.module
if input_name not in self . _modules and input_name not in local_state :
unexpected_keys . append ( key )
-[docs] def load_state_dict ( self , state_dict : Union [ Dict [ str , Tensor ], Dict [ str , Tensor ]],
+
def load_state_dict ( self , state_dict : Union [ Dict [ str , Tensor ], Dict [ str , Tensor ]],
strict : bool = True ):
r """Copies parameters and buffers from :attr:`state_dict` into
this module and its descendants. If :attr:`strict` is ``True``, then
@@ -1389,7 +1389,7 @@
Source code for torch.nn.modules.module
if len ( error_msgs ) > 0 :
raise RuntimeError ( 'Error(s) in loading state_dict for {} : \n\t {} ' . format (
self . __class__ . __name__ , " \n\t " . join ( error_msgs )))
- return _IncompatibleKeys ( missing_keys , unexpected_keys )
+
return _IncompatibleKeys ( missing_keys , unexpected_keys )
def _named_members ( self , get_members_fn , prefix = '' , recurse = True ):
r """Helper method for yielding various names + members of modules."""
@@ -1404,7 +1404,7 @@
Source code for torch.nn.modules.module
name = module_prefix + ( '.' if module_prefix else '' ) + k
yield name , v
-[docs] def parameters ( self , recurse : bool = True ) -> Iterator [ Parameter ]:
+
def parameters ( self , recurse : bool = True ) -> Iterator [ Parameter ]:
r """Returns an iterator over module parameters.
This is typically passed to an optimizer.
@@ -1426,9 +1426,9 @@
Source code for torch.nn.modules.module
"""
for name , param in self . named_parameters ( recurse = recurse ):
- yield param
+
yield param
-
[docs] def named_parameters ( self , prefix : str = '' , recurse : bool = True ) -> Iterator [ Tuple [ str , Tensor ]]:
+
def named_parameters ( self , prefix : str = '' , recurse : bool = True ) -> Iterator [ Tuple [ str , Tensor ]]:
r """Returns an iterator over module parameters, yielding both the
name of the parameter as well as the parameter itself.
@@ -1452,9 +1452,9 @@
Source code for torch.nn.modules.module
lambda module : module . _parameters . items (),
prefix = prefix , recurse = recurse )
for elem in gen :
- yield elem
+
yield elem
-
[docs] def buffers ( self , recurse : bool = True ) -> Iterator [ Tensor ]:
+
def buffers ( self , recurse : bool = True ) -> Iterator [ Tensor ]:
r """Returns an iterator over module buffers.
Args:
@@ -1474,9 +1474,9 @@
Source code for torch.nn.modules.module
"""
for name , buf in self . named_buffers ( recurse = recurse ):
- yield buf
+
yield buf
-
[docs] def named_buffers ( self , prefix : str = '' , recurse : bool = True ) -> Iterator [ Tuple [ str , Tensor ]]:
+
def named_buffers ( self , prefix : str = '' , recurse : bool = True ) -> Iterator [ Tuple [ str , Tensor ]]:
r """Returns an iterator over module buffers, yielding both the
name of the buffer as well as the buffer itself.
@@ -1500,18 +1500,18 @@
Source code for torch.nn.modules.module
lambda module : module . _buffers . items (),
prefix = prefix , recurse = recurse )
for elem in gen :
- yield elem
+
yield elem
-
[docs] def children ( self ) -> Iterator [ 'Module' ]:
+
def children ( self ) -> Iterator [ 'Module' ]:
r """Returns an iterator over immediate children modules.
Yields:
Module: a child module
"""
for name , module in self . named_children ():
-
yield module
+
yield module
-
[docs] def named_children ( self ) -> Iterator [ Tuple [ str , 'Module' ]]:
+
def named_children ( self ) -> Iterator [ Tuple [ str , 'Module' ]]:
r """Returns an iterator over immediate children modules, yielding both
the name of the module as well as the module itself.
@@ -1529,9 +1529,9 @@
Source code for torch.nn.modules.module
for name , module in self . _modules . items ():
if module is not None and module not in memo :
memo . add ( module )
- yield name , module
+
yield name , module
-
[docs] def modules ( self ) -> Iterator [ 'Module' ]:
+
def modules ( self ) -> Iterator [ 'Module' ]:
r """Returns an iterator over all modules in the network.
Yields:
@@ -1556,9 +1556,9 @@
Source code for torch.nn.modules.module
"""
for name , module in self . named_modules ():
- yield module
+
yield module
-
[docs] def named_modules ( self , memo : Optional [ Set [ 'Module' ]] = None , prefix : str = '' ):
+
def named_modules ( self , memo : Optional [ Set [ 'Module' ]] = None , prefix : str = '' ):
r """Returns an iterator over all modules in the network, yielding
both the name of the module as well as the module itself.
@@ -1594,9 +1594,9 @@
Source code for torch.nn.modules.module
continue
submodule_prefix = prefix + ( '.' if prefix else '' ) + name
for m in module . named_modules ( memo , submodule_prefix ):
- yield m
+
yield m
-
[docs] def train ( self : T , mode : bool = True ) -> T :
+
def train ( self : T , mode : bool = True ) -> T :
r """Sets the module in training mode.
This has any effect only on certain modules. See documentations of
@@ -1614,9 +1614,9 @@
Source code for torch.nn.modules.module
self . training = mode
for module in self . children ():
module . train ( mode )
- return self
+
return self
-
[docs] def eval ( self : T ) -> T :
+
def eval ( self : T ) -> T :
r """Sets the module in evaluation mode.
This has any effect only on certain modules. See documentations of
@@ -1629,9 +1629,9 @@
Source code for torch.nn.modules.module
Returns:
Module: self
"""
- return self . train ( False )
+
return self . train ( False )
-
[docs] def requires_grad_ ( self : T , requires_grad : bool = True ) -> T :
+
def requires_grad_ ( self : T , requires_grad : bool = True ) -> T :
r """Change if autograd should record operations on parameters in this
module.
@@ -1650,9 +1650,9 @@
Source code for torch.nn.modules.module
"""
for p in self . parameters ():
p . requires_grad_ ( requires_grad )
- return self
+
return self
-
[docs] def zero_grad ( self ) -> None :
+
def zero_grad ( self ) -> None :
r """Sets gradients of all model parameters to zero."""
if getattr ( self , '_is_replica' , False ):
warnings . warn (
@@ -1664,7 +1664,7 @@
Source code for torch.nn.modules.module
for p in self . parameters ():
if p . grad is not None :
p . grad . detach_ ()
- p . grad . zero_ ()
+
p . grad . zero_ ()
def share_memory ( self : T ) -> T :
return self . _apply ( lambda t : t . share_memory_ ())
@@ -1672,14 +1672,14 @@
Source code for torch.nn.modules.module
def _get_name ( self ):
return self . __class__ . __name__
-
+ return ''
def __repr__ ( self ):
# We treat the extra repr like the sub-module, one item per line
@@ -1730,7 +1730,7 @@ Source code for torch.nn.modules.module
replica . _modules = replica . _modules . copy ()
replica . _is_replica = True
- return replica return replica
diff --git a/docs/master/_modules/torch/nn/modules/normalization.html b/docs/master/_modules/torch/nn/modules/normalization.html
index 9f51992e177b..1c2e8a5e767a 100644
--- a/docs/master/_modules/torch/nn/modules/normalization.html
+++ b/docs/master/_modules/torch/nn/modules/normalization.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/padding.html b/docs/master/_modules/torch/nn/modules/padding.html
index 3ebd0e4a4af0..8a6be1f8dadb 100644
--- a/docs/master/_modules/torch/nn/modules/padding.html
+++ b/docs/master/_modules/torch/nn/modules/padding.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/pixelshuffle.html b/docs/master/_modules/torch/nn/modules/pixelshuffle.html
index 172b78974753..837e032e2e67 100644
--- a/docs/master/_modules/torch/nn/modules/pixelshuffle.html
+++ b/docs/master/_modules/torch/nn/modules/pixelshuffle.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/pooling.html b/docs/master/_modules/torch/nn/modules/pooling.html
index 085001177b00..9157849a9019 100644
--- a/docs/master/_modules/torch/nn/modules/pooling.html
+++ b/docs/master/_modules/torch/nn/modules/pooling.html
@@ -158,7 +158,7 @@
@@ -729,7 +729,7 @@
Source code for torch.nn.modules.pooling
self . padding , output_size )
-
[docs] class MaxUnpool3d ( _MaxUnpoolNd ):
+
class MaxUnpool3d ( _MaxUnpoolNd ):
r """Computes a partial inverse of :class:`MaxPool3d`.
:class:`MaxPool3d` is not fully invertible, since the non-maximal values are lost.
@@ -792,7 +792,7 @@
Source code for torch.nn.modules.pooling
def forward ( self , input : Tensor , indices : Tensor , output_size : Optional [ List [ int ]] = None ) -> Tensor :
return F . max_unpool3d ( input , indices , self . kernel_size , self . stride ,
- self . padding , output_size )
+
self . padding , output_size )
class _AvgPoolNd ( Module ):
diff --git a/docs/master/_modules/torch/nn/modules/rnn.html b/docs/master/_modules/torch/nn/modules/rnn.html
index 1654f2fd4987..49866967db59 100644
--- a/docs/master/_modules/torch/nn/modules/rnn.html
+++ b/docs/master/_modules/torch/nn/modules/rnn.html
@@ -158,7 +158,7 @@
@@ -722,6 +722,8 @@
Source code for torch.nn.modules.rnn
All the weights and biases are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`
where :math:`k = \frac{1}{\text{hidden\_size}}`
+ .. include:: ../cudnn_rnn_determinism.rst
+
.. include:: ../cudnn_persistent_rnn.rst
Examples::
@@ -849,6 +851,8 @@ Source code for torch.nn.modules.rnn
All the weights and biases are initialized from :math:`\mathcal{U}(-\sqrt{k}, \sqrt{k})`
where :math:`k = \frac{1}{\text{hidden\_size}}`
+ .. include:: ../cudnn_rnn_determinism.rst
+
.. include:: ../cudnn_persistent_rnn.rst
Examples::
diff --git a/docs/master/_modules/torch/nn/modules/sparse.html b/docs/master/_modules/torch/nn/modules/sparse.html
index aec2078feb2d..6e78023fe202 100644
--- a/docs/master/_modules/torch/nn/modules/sparse.html
+++ b/docs/master/_modules/torch/nn/modules/sparse.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/transformer.html b/docs/master/_modules/torch/nn/modules/transformer.html
index 0e4915f6e0a6..b6c88d16def6 100644
--- a/docs/master/_modules/torch/nn/modules/transformer.html
+++ b/docs/master/_modules/torch/nn/modules/transformer.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/modules/upsampling.html b/docs/master/_modules/torch/nn/modules/upsampling.html
index 839456ce975a..5f4c79d14a87 100644
--- a/docs/master/_modules/torch/nn/modules/upsampling.html
+++ b/docs/master/_modules/torch/nn/modules/upsampling.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/parallel/comm.html b/docs/master/_modules/torch/nn/parallel/comm.html
index 8e8adb80aff8..de515f903e00 100644
--- a/docs/master/_modules/torch/nn/parallel/comm.html
+++ b/docs/master/_modules/torch/nn/parallel/comm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/parallel/data_parallel.html b/docs/master/_modules/torch/nn/parallel/data_parallel.html
index 59a27e1bf8a1..34e664e86698 100644
--- a/docs/master/_modules/torch/nn/parallel/data_parallel.html
+++ b/docs/master/_modules/torch/nn/parallel/data_parallel.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/parallel/distributed.html b/docs/master/_modules/torch/nn/parallel/distributed.html
index c881e80e8bfc..1fd2847d5bbc 100644
--- a/docs/master/_modules/torch/nn/parallel/distributed.html
+++ b/docs/master/_modules/torch/nn/parallel/distributed.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/parameter.html b/docs/master/_modules/torch/nn/parameter.html
index f3d30c3ed1f5..bd0c8bb72071 100644
--- a/docs/master/_modules/torch/nn/parameter.html
+++ b/docs/master/_modules/torch/nn/parameter.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/qat/modules/conv.html b/docs/master/_modules/torch/nn/qat/modules/conv.html
index 441425e4db24..aea2c4779204 100644
--- a/docs/master/_modules/torch/nn/qat/modules/conv.html
+++ b/docs/master/_modules/torch/nn/qat/modules/conv.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/qat/modules/linear.html b/docs/master/_modules/torch/nn/qat/modules/linear.html
index cef6049765d8..367dbe4f3adc 100644
--- a/docs/master/_modules/torch/nn/qat/modules/linear.html
+++ b/docs/master/_modules/torch/nn/qat/modules/linear.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/dynamic/modules/linear.html b/docs/master/_modules/torch/nn/quantized/dynamic/modules/linear.html
index 2eafd3895abf..7424a316cc2f 100644
--- a/docs/master/_modules/torch/nn/quantized/dynamic/modules/linear.html
+++ b/docs/master/_modules/torch/nn/quantized/dynamic/modules/linear.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/dynamic/modules/rnn.html b/docs/master/_modules/torch/nn/quantized/dynamic/modules/rnn.html
index 58abfb3b787d..b2381afa65b6 100644
--- a/docs/master/_modules/torch/nn/quantized/dynamic/modules/rnn.html
+++ b/docs/master/_modules/torch/nn/quantized/dynamic/modules/rnn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/functional.html b/docs/master/_modules/torch/nn/quantized/functional.html
index 9a29e3b4264b..eed88a66857e 100644
--- a/docs/master/_modules/torch/nn/quantized/functional.html
+++ b/docs/master/_modules/torch/nn/quantized/functional.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules.html b/docs/master/_modules/torch/nn/quantized/modules.html
index ab9397ee174e..56b9ab4ca800 100644
--- a/docs/master/_modules/torch/nn/quantized/modules.html
+++ b/docs/master/_modules/torch/nn/quantized/modules.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/activation.html b/docs/master/_modules/torch/nn/quantized/modules/activation.html
index 1b577ecbee6e..a468345c319d 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/activation.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/activation.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/batchnorm.html b/docs/master/_modules/torch/nn/quantized/modules/batchnorm.html
index 2b0c02f38e20..de5f553c57ad 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/batchnorm.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/batchnorm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/conv.html b/docs/master/_modules/torch/nn/quantized/modules/conv.html
index dbe4750344c4..7535eaf85a55 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/conv.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/conv.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/functional_modules.html b/docs/master/_modules/torch/nn/quantized/modules/functional_modules.html
index ed64305bce2a..b0ca8fd23736 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/functional_modules.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/functional_modules.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/linear.html b/docs/master/_modules/torch/nn/quantized/modules/linear.html
index 11485cec522e..bdabdf895aea 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/linear.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/linear.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/quantized/modules/normalization.html b/docs/master/_modules/torch/nn/quantized/modules/normalization.html
index f56c339f85d5..16ba80a45143 100644
--- a/docs/master/_modules/torch/nn/quantized/modules/normalization.html
+++ b/docs/master/_modules/torch/nn/quantized/modules/normalization.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/clip_grad.html b/docs/master/_modules/torch/nn/utils/clip_grad.html
index 93af006fa9c1..3b96a11f7934 100644
--- a/docs/master/_modules/torch/nn/utils/clip_grad.html
+++ b/docs/master/_modules/torch/nn/utils/clip_grad.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/convert_parameters.html b/docs/master/_modules/torch/nn/utils/convert_parameters.html
index 1ad9355f8d11..dea2f1575ee5 100644
--- a/docs/master/_modules/torch/nn/utils/convert_parameters.html
+++ b/docs/master/_modules/torch/nn/utils/convert_parameters.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/prune.html b/docs/master/_modules/torch/nn/utils/prune.html
index 0b069d3eef51..99b42f5d030b 100644
--- a/docs/master/_modules/torch/nn/utils/prune.html
+++ b/docs/master/_modules/torch/nn/utils/prune.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/rnn.html b/docs/master/_modules/torch/nn/utils/rnn.html
index f3c287fce572..290ec5feecbd 100644
--- a/docs/master/_modules/torch/nn/utils/rnn.html
+++ b/docs/master/_modules/torch/nn/utils/rnn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/spectral_norm.html b/docs/master/_modules/torch/nn/utils/spectral_norm.html
index 8328cce42b0c..1bd146f32cdb 100644
--- a/docs/master/_modules/torch/nn/utils/spectral_norm.html
+++ b/docs/master/_modules/torch/nn/utils/spectral_norm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/nn/utils/weight_norm.html b/docs/master/_modules/torch/nn/utils/weight_norm.html
index 7158fbb3ddcf..726c5cd0be3f 100644
--- a/docs/master/_modules/torch/nn/utils/weight_norm.html
+++ b/docs/master/_modules/torch/nn/utils/weight_norm.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/onnx.html b/docs/master/_modules/torch/onnx.html
index 5ad48ad7928d..a7fb6654632b 100644
--- a/docs/master/_modules/torch/onnx.html
+++ b/docs/master/_modules/torch/onnx.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/onnx/operators.html b/docs/master/_modules/torch/onnx/operators.html
index f4c14aebbc04..a1b9c5908b89 100644
--- a/docs/master/_modules/torch/onnx/operators.html
+++ b/docs/master/_modules/torch/onnx/operators.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/adadelta.html b/docs/master/_modules/torch/optim/adadelta.html
index 6042f29dd750..0910a6f85a5e 100644
--- a/docs/master/_modules/torch/optim/adadelta.html
+++ b/docs/master/_modules/torch/optim/adadelta.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/adagrad.html b/docs/master/_modules/torch/optim/adagrad.html
index 75140e0e5bf3..bbf55b60db6d 100644
--- a/docs/master/_modules/torch/optim/adagrad.html
+++ b/docs/master/_modules/torch/optim/adagrad.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/adam.html b/docs/master/_modules/torch/optim/adam.html
index 0ec5fd95b8bb..212a36fa54b0 100644
--- a/docs/master/_modules/torch/optim/adam.html
+++ b/docs/master/_modules/torch/optim/adam.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/adamax.html b/docs/master/_modules/torch/optim/adamax.html
index fba280d37c32..6f4ffd91e136 100644
--- a/docs/master/_modules/torch/optim/adamax.html
+++ b/docs/master/_modules/torch/optim/adamax.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/adamw.html b/docs/master/_modules/torch/optim/adamw.html
index e757baac9a82..6cc9d546cd74 100644
--- a/docs/master/_modules/torch/optim/adamw.html
+++ b/docs/master/_modules/torch/optim/adamw.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/asgd.html b/docs/master/_modules/torch/optim/asgd.html
index 7ac6f04a46da..e6e581db18ce 100644
--- a/docs/master/_modules/torch/optim/asgd.html
+++ b/docs/master/_modules/torch/optim/asgd.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/lbfgs.html b/docs/master/_modules/torch/optim/lbfgs.html
index ddbdf3217eba..7eb434d2d288 100644
--- a/docs/master/_modules/torch/optim/lbfgs.html
+++ b/docs/master/_modules/torch/optim/lbfgs.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/lr_scheduler.html b/docs/master/_modules/torch/optim/lr_scheduler.html
index 9d6af0a12e9f..d9aaf8cf7dec 100644
--- a/docs/master/_modules/torch/optim/lr_scheduler.html
+++ b/docs/master/_modules/torch/optim/lr_scheduler.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/optimizer.html b/docs/master/_modules/torch/optim/optimizer.html
index 6db02723e9ee..b5589478d395 100644
--- a/docs/master/_modules/torch/optim/optimizer.html
+++ b/docs/master/_modules/torch/optim/optimizer.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/rmsprop.html b/docs/master/_modules/torch/optim/rmsprop.html
index bbb7cac1782c..ce9863bdad42 100644
--- a/docs/master/_modules/torch/optim/rmsprop.html
+++ b/docs/master/_modules/torch/optim/rmsprop.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/rprop.html b/docs/master/_modules/torch/optim/rprop.html
index bee1d7fe5159..459f301edc35 100644
--- a/docs/master/_modules/torch/optim/rprop.html
+++ b/docs/master/_modules/torch/optim/rprop.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/sgd.html b/docs/master/_modules/torch/optim/sgd.html
index 2cc7e3e74711..cad2530fc15a 100644
--- a/docs/master/_modules/torch/optim/sgd.html
+++ b/docs/master/_modules/torch/optim/sgd.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/optim/sparse_adam.html b/docs/master/_modules/torch/optim/sparse_adam.html
index bccc7a8b4541..c12659181712 100644
--- a/docs/master/_modules/torch/optim/sparse_adam.html
+++ b/docs/master/_modules/torch/optim/sparse_adam.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization.html b/docs/master/_modules/torch/quantization.html
index 66f76db88274..b897951f7521 100644
--- a/docs/master/_modules/torch/quantization.html
+++ b/docs/master/_modules/torch/quantization.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization/fake_quantize.html b/docs/master/_modules/torch/quantization/fake_quantize.html
index b10924f11109..f77c3d24e9a8 100644
--- a/docs/master/_modules/torch/quantization/fake_quantize.html
+++ b/docs/master/_modules/torch/quantization/fake_quantize.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization/fuse_modules.html b/docs/master/_modules/torch/quantization/fuse_modules.html
index c96bbe03c11c..037681e9291b 100644
--- a/docs/master/_modules/torch/quantization/fuse_modules.html
+++ b/docs/master/_modules/torch/quantization/fuse_modules.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization/observer.html b/docs/master/_modules/torch/quantization/observer.html
index 5d28bf075cc6..aeec77c32d9d 100644
--- a/docs/master/_modules/torch/quantization/observer.html
+++ b/docs/master/_modules/torch/quantization/observer.html
@@ -158,7 +158,7 @@
@@ -607,6 +607,9 @@ Source code for torch.quantization.observer
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
+ initial_dynamic_qrange: Nullable tuple containing the initial qmin and qmax values
+ to support dynamic quantization range. If unspecified,
+ the initialization of qmin and qmax follows the 8-bit setup.
Given running min/max as :math:`x_\text{min}` and :math:`x_\text{max}`,
scale :math:`s` and zero point :math:`z` are computed as:
@@ -658,7 +661,7 @@ Source code for torch.quantization.observer
"""
def __init__ ( self , dtype = torch . quint8 , qscheme = torch . per_tensor_affine ,
- reduce_range = False ):
+ reduce_range = False , initial_dynamic_qrange = None ):
# For x86 quantized kernels, we need to ensure that the vpmaddubsw
# instruction does not overflow. We allow for a reduce_range argument to
# observers that reduces the quantized range to (0,127) or (-64, 63).
@@ -668,7 +671,8 @@ Source code for torch.quantization.observer
super ( MinMaxObserver , self ) . __init__ ( dtype = dtype ,
qscheme = qscheme ,
- reduce_range = reduce_range )
+ reduce_range = reduce_range ,
+ initial_dynamic_qrange = initial_dynamic_qrange )
self . register_buffer ( 'min_val' , torch . tensor ([]))
self . register_buffer ( 'max_val' , torch . tensor ([]))
if self . qscheme == torch . per_tensor_symmetric and \
@@ -738,6 +742,9 @@ Source code for torch.quantization.observer
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
+ initial_dynamic_qrange: Nullable tuple containing the initial qmin and qmax values
+ to support dynamic quantization range. If unspecified,
+ the initialization of qmin and qmax follows the 8-bit setup.
The moving average min/max is computed as follows
@@ -766,11 +773,13 @@ Source code for torch.quantization.observer
and zero_point are set to 1.0 and 0.
"""
def __init__ ( self , averaging_constant = 0.01 , dtype = torch . quint8 ,
- qscheme = torch . per_tensor_affine , reduce_range = False ):
+ qscheme = torch . per_tensor_affine , reduce_range = False ,
+ initial_dynamic_qrange = None ):
self . averaging_constant = averaging_constant
super ( MovingAverageMinMaxObserver , self ) . __init__ ( dtype = dtype ,
qscheme = qscheme ,
- reduce_range = reduce_range )
+ reduce_range = reduce_range ,
+ initial_dynamic_qrange = initial_dynamic_qrange )
def forward ( self , x_orig ):
x = x_orig . detach () # avoid keeping autograd tape
@@ -879,6 +888,9 @@ Source code for torch.quantization.observer
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
+ initial_dynamic_qrange: Nullable tuple containing the initial qmin and qmax values
+ to support dynamic quantization range. If unspecified,
+ the initialization of qmin and qmax follows the 8-bit setup.
The quantization parameters are computed the same way as in
:class:`~torch.quantization.observer.MinMaxObserver`, with the difference
@@ -890,10 +902,12 @@ Source code for torch.quantization.observer
"""
def __init__ ( self , ch_axis = 0 , dtype = torch . quint8 ,
- qscheme = torch . per_channel_affine , reduce_range = False ):
+ qscheme = torch . per_channel_affine , reduce_range = False ,
+ initial_dyanmic_qrange = None ):
super ( PerChannelMinMaxObserver , self ) . __init__ ( dtype = dtype ,
qscheme = qscheme ,
- reduce_range = reduce_range )
+ reduce_range = reduce_range ,
+ initial_dynamic_qrange = initial_dyanmic_qrange )
self . ch_axis = ch_axis
self . register_buffer ( 'min_vals' , torch . tensor ([]))
self . register_buffer ( 'max_vals' , torch . tensor ([]))
@@ -976,6 +990,9 @@ Source code for torch.quantization.observer
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
+ initial_dynamic_qrange: Nullable tuple containing the initial qmin and qmax values
+ to support dynamic quantization range. If unspecified,
+ the initialization of qmin and qmax follows the 8-bit setup.
The quantization parameters are computed the same way as in
:class:`~torch.quantization.observer.MovingAverageMinMaxObserver`, with the
@@ -987,10 +1004,11 @@ Source code for torch.quantization.observer
"""
def __init__ ( self , averaging_constant = 0.01 , ch_axis = 0 , dtype = torch . quint8 ,
- qscheme = torch . per_channel_affine , reduce_range = False ):
+ qscheme = torch . per_channel_affine , reduce_range = False ,
+ initial_dynamic_qrange = None ):
super ( MovingAveragePerChannelMinMaxObserver , self ) . __init__ (
ch_axis = ch_axis , dtype = dtype , qscheme = qscheme ,
- reduce_range = reduce_range )
+ reduce_range = reduce_range , initial_dyanmic_qrange = initial_dynamic_qrange )
self . averaging_constant = averaging_constant
def forward ( self , x_orig ):
diff --git a/docs/master/_modules/torch/quantization/qconfig.html b/docs/master/_modules/torch/quantization/qconfig.html
index 0a5bbdc92ea1..22abdf0a0e1d 100644
--- a/docs/master/_modules/torch/quantization/qconfig.html
+++ b/docs/master/_modules/torch/quantization/qconfig.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization/quantize.html b/docs/master/_modules/torch/quantization/quantize.html
index a08e0fb10a44..5e3f6ea6818d 100644
--- a/docs/master/_modules/torch/quantization/quantize.html
+++ b/docs/master/_modules/torch/quantization/quantize.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quantization/stubs.html b/docs/master/_modules/torch/quantization/stubs.html
index 4a715fbd0933..59af1a67edd7 100644
--- a/docs/master/_modules/torch/quantization/stubs.html
+++ b/docs/master/_modules/torch/quantization/stubs.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/quasirandom.html b/docs/master/_modules/torch/quasirandom.html
index 8ebd4d94f4fa..08a4d34cbde1 100644
--- a/docs/master/_modules/torch/quasirandom.html
+++ b/docs/master/_modules/torch/quasirandom.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/random.html b/docs/master/_modules/torch/random.html
index f3a974e011cd..5615e4ad011d 100644
--- a/docs/master/_modules/torch/random.html
+++ b/docs/master/_modules/torch/random.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/serialization.html b/docs/master/_modules/torch/serialization.html
index a6b997dd3dd5..75bbd5a006e4 100644
--- a/docs/master/_modules/torch/serialization.html
+++ b/docs/master/_modules/torch/serialization.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/sparse.html b/docs/master/_modules/torch/sparse.html
index cb81f6cba128..ee1f84c350f4 100644
--- a/docs/master/_modules/torch/sparse.html
+++ b/docs/master/_modules/torch/sparse.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/storage.html b/docs/master/_modules/torch/storage.html
index fd9e398fddf0..ad0318f9d184 100644
--- a/docs/master/_modules/torch/storage.html
+++ b/docs/master/_modules/torch/storage.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/tensor.html b/docs/master/_modules/torch/tensor.html
index 8954a0d105b9..5b187586a239 100644
--- a/docs/master/_modules/torch/tensor.html
+++ b/docs/master/_modules/torch/tensor.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/checkpoint.html b/docs/master/_modules/torch/utils/checkpoint.html
index 432f33a1c98b..9a0ab92a1451 100644
--- a/docs/master/_modules/torch/utils/checkpoint.html
+++ b/docs/master/_modules/torch/utils/checkpoint.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/cpp_extension.html b/docs/master/_modules/torch/utils/cpp_extension.html
index 39da3c4e355c..6b989e0984c6 100644
--- a/docs/master/_modules/torch/utils/cpp_extension.html
+++ b/docs/master/_modules/torch/utils/cpp_extension.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/data/_utils/worker.html b/docs/master/_modules/torch/utils/data/_utils/worker.html
index 478ed791e28b..a0811fbc8f75 100644
--- a/docs/master/_modules/torch/utils/data/_utils/worker.html
+++ b/docs/master/_modules/torch/utils/data/_utils/worker.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/data/dataloader.html b/docs/master/_modules/torch/utils/data/dataloader.html
index becf71fd323b..4c1268579a7e 100644
--- a/docs/master/_modules/torch/utils/data/dataloader.html
+++ b/docs/master/_modules/torch/utils/data/dataloader.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/data/dataset.html b/docs/master/_modules/torch/utils/data/dataset.html
index 22ff8ef376a9..46045e3d39f3 100644
--- a/docs/master/_modules/torch/utils/data/dataset.html
+++ b/docs/master/_modules/torch/utils/data/dataset.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/data/distributed.html b/docs/master/_modules/torch/utils/data/distributed.html
index 347448d9a1d8..88d5187cf115 100644
--- a/docs/master/_modules/torch/utils/data/distributed.html
+++ b/docs/master/_modules/torch/utils/data/distributed.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/data/sampler.html b/docs/master/_modules/torch/utils/data/sampler.html
index 1ee15eed6801..f0583f8064c2 100644
--- a/docs/master/_modules/torch/utils/data/sampler.html
+++ b/docs/master/_modules/torch/utils/data/sampler.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torch/utils/mobile_optimizer.html b/docs/master/_modules/torch/utils/mobile_optimizer.html
index 988232f255c3..01c092cd695f 100644
--- a/docs/master/_modules/torch/utils/mobile_optimizer.html
+++ b/docs/master/_modules/torch/utils/mobile_optimizer.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision.html b/docs/master/_modules/torchvision.html
index c6abb1e4738e..3dccd1f98641 100644
--- a/docs/master/_modules/torchvision.html
+++ b/docs/master/_modules/torchvision.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/celeba.html b/docs/master/_modules/torchvision/datasets/celeba.html
index 89754a22d64c..4a6bdaa83988 100644
--- a/docs/master/_modules/torchvision/datasets/celeba.html
+++ b/docs/master/_modules/torchvision/datasets/celeba.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/cifar.html b/docs/master/_modules/torchvision/datasets/cifar.html
index 51b8a085896f..7df00ad0809b 100644
--- a/docs/master/_modules/torchvision/datasets/cifar.html
+++ b/docs/master/_modules/torchvision/datasets/cifar.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/cityscapes.html b/docs/master/_modules/torchvision/datasets/cityscapes.html
index 9ae1c1063da5..da92ccb64c36 100644
--- a/docs/master/_modules/torchvision/datasets/cityscapes.html
+++ b/docs/master/_modules/torchvision/datasets/cityscapes.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/coco.html b/docs/master/_modules/torchvision/datasets/coco.html
index e607f560ebad..95e10aece1d6 100644
--- a/docs/master/_modules/torchvision/datasets/coco.html
+++ b/docs/master/_modules/torchvision/datasets/coco.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/fakedata.html b/docs/master/_modules/torchvision/datasets/fakedata.html
index 71a1c0291697..c12f5e2e3596 100644
--- a/docs/master/_modules/torchvision/datasets/fakedata.html
+++ b/docs/master/_modules/torchvision/datasets/fakedata.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/flickr.html b/docs/master/_modules/torchvision/datasets/flickr.html
index 33be1f54417c..0d910b0d9185 100644
--- a/docs/master/_modules/torchvision/datasets/flickr.html
+++ b/docs/master/_modules/torchvision/datasets/flickr.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/folder.html b/docs/master/_modules/torchvision/datasets/folder.html
index 0ab06d00238d..20e565c009d0 100644
--- a/docs/master/_modules/torchvision/datasets/folder.html
+++ b/docs/master/_modules/torchvision/datasets/folder.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/hmdb51.html b/docs/master/_modules/torchvision/datasets/hmdb51.html
index 0bdfd589ded4..93762ee2ac63 100644
--- a/docs/master/_modules/torchvision/datasets/hmdb51.html
+++ b/docs/master/_modules/torchvision/datasets/hmdb51.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/imagenet.html b/docs/master/_modules/torchvision/datasets/imagenet.html
index ad10c66763a7..36863a807b1f 100644
--- a/docs/master/_modules/torchvision/datasets/imagenet.html
+++ b/docs/master/_modules/torchvision/datasets/imagenet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/kinetics.html b/docs/master/_modules/torchvision/datasets/kinetics.html
index 7e22573d685b..66babb7e2fa9 100644
--- a/docs/master/_modules/torchvision/datasets/kinetics.html
+++ b/docs/master/_modules/torchvision/datasets/kinetics.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/lsun.html b/docs/master/_modules/torchvision/datasets/lsun.html
index 1fa62f49fd72..5debb09fb363 100644
--- a/docs/master/_modules/torchvision/datasets/lsun.html
+++ b/docs/master/_modules/torchvision/datasets/lsun.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/mnist.html b/docs/master/_modules/torchvision/datasets/mnist.html
index b0d1092b9fe6..f93f2bfa665b 100644
--- a/docs/master/_modules/torchvision/datasets/mnist.html
+++ b/docs/master/_modules/torchvision/datasets/mnist.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/phototour.html b/docs/master/_modules/torchvision/datasets/phototour.html
index b85070b2b244..549b94594ca9 100644
--- a/docs/master/_modules/torchvision/datasets/phototour.html
+++ b/docs/master/_modules/torchvision/datasets/phototour.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/sbd.html b/docs/master/_modules/torchvision/datasets/sbd.html
index bff3a5ff4f89..3452a47e171a 100644
--- a/docs/master/_modules/torchvision/datasets/sbd.html
+++ b/docs/master/_modules/torchvision/datasets/sbd.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/sbu.html b/docs/master/_modules/torchvision/datasets/sbu.html
index 4ff9accde384..c711fb6eceda 100644
--- a/docs/master/_modules/torchvision/datasets/sbu.html
+++ b/docs/master/_modules/torchvision/datasets/sbu.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/stl10.html b/docs/master/_modules/torchvision/datasets/stl10.html
index 74031ce4ced0..28701bf47d2c 100644
--- a/docs/master/_modules/torchvision/datasets/stl10.html
+++ b/docs/master/_modules/torchvision/datasets/stl10.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/svhn.html b/docs/master/_modules/torchvision/datasets/svhn.html
index 4af3d36fa7c8..807e938e109e 100644
--- a/docs/master/_modules/torchvision/datasets/svhn.html
+++ b/docs/master/_modules/torchvision/datasets/svhn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/ucf101.html b/docs/master/_modules/torchvision/datasets/ucf101.html
index 05198671fa27..14d3decee7d4 100644
--- a/docs/master/_modules/torchvision/datasets/ucf101.html
+++ b/docs/master/_modules/torchvision/datasets/ucf101.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/usps.html b/docs/master/_modules/torchvision/datasets/usps.html
index f18a6aab8329..4b5d3ba99a18 100644
--- a/docs/master/_modules/torchvision/datasets/usps.html
+++ b/docs/master/_modules/torchvision/datasets/usps.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/datasets/voc.html b/docs/master/_modules/torchvision/datasets/voc.html
index dec27d91864d..66b65c7e840d 100644
--- a/docs/master/_modules/torchvision/datasets/voc.html
+++ b/docs/master/_modules/torchvision/datasets/voc.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/io/video.html b/docs/master/_modules/torchvision/io/video.html
index e38f893afe26..0f777769bc68 100644
--- a/docs/master/_modules/torchvision/io/video.html
+++ b/docs/master/_modules/torchvision/io/video.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/alexnet.html b/docs/master/_modules/torchvision/models/alexnet.html
index 85885afa50a0..bd113250e4ce 100644
--- a/docs/master/_modules/torchvision/models/alexnet.html
+++ b/docs/master/_modules/torchvision/models/alexnet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/densenet.html b/docs/master/_modules/torchvision/models/densenet.html
index bdab04727ec3..7a94eca1c2bb 100644
--- a/docs/master/_modules/torchvision/models/densenet.html
+++ b/docs/master/_modules/torchvision/models/densenet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/detection/faster_rcnn.html b/docs/master/_modules/torchvision/models/detection/faster_rcnn.html
index a83fc5f0eb99..1fd510717740 100644
--- a/docs/master/_modules/torchvision/models/detection/faster_rcnn.html
+++ b/docs/master/_modules/torchvision/models/detection/faster_rcnn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/detection/keypoint_rcnn.html b/docs/master/_modules/torchvision/models/detection/keypoint_rcnn.html
index 688d8f0cbb28..3b6f8fe36419 100644
--- a/docs/master/_modules/torchvision/models/detection/keypoint_rcnn.html
+++ b/docs/master/_modules/torchvision/models/detection/keypoint_rcnn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/detection/mask_rcnn.html b/docs/master/_modules/torchvision/models/detection/mask_rcnn.html
index 3f34cb1716fb..cb0cbdf81f72 100644
--- a/docs/master/_modules/torchvision/models/detection/mask_rcnn.html
+++ b/docs/master/_modules/torchvision/models/detection/mask_rcnn.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/googlenet.html b/docs/master/_modules/torchvision/models/googlenet.html
index b22cce75bd54..0bc511c1bee8 100644
--- a/docs/master/_modules/torchvision/models/googlenet.html
+++ b/docs/master/_modules/torchvision/models/googlenet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/inception.html b/docs/master/_modules/torchvision/models/inception.html
index 48a465a3097b..cd6f88283828 100644
--- a/docs/master/_modules/torchvision/models/inception.html
+++ b/docs/master/_modules/torchvision/models/inception.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/mnasnet.html b/docs/master/_modules/torchvision/models/mnasnet.html
index fcc608c3e302..8f28f063befc 100644
--- a/docs/master/_modules/torchvision/models/mnasnet.html
+++ b/docs/master/_modules/torchvision/models/mnasnet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/mobilenet.html b/docs/master/_modules/torchvision/models/mobilenet.html
index 87c98917b775..f026f4ff7740 100644
--- a/docs/master/_modules/torchvision/models/mobilenet.html
+++ b/docs/master/_modules/torchvision/models/mobilenet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/resnet.html b/docs/master/_modules/torchvision/models/resnet.html
index 4b7f9622225e..91aeb18e97a4 100644
--- a/docs/master/_modules/torchvision/models/resnet.html
+++ b/docs/master/_modules/torchvision/models/resnet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/segmentation/segmentation.html b/docs/master/_modules/torchvision/models/segmentation/segmentation.html
index 39ce4b78ed19..f90cddab4409 100644
--- a/docs/master/_modules/torchvision/models/segmentation/segmentation.html
+++ b/docs/master/_modules/torchvision/models/segmentation/segmentation.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/shufflenetv2.html b/docs/master/_modules/torchvision/models/shufflenetv2.html
index 2d4f9b198ca4..d05b5fef0754 100644
--- a/docs/master/_modules/torchvision/models/shufflenetv2.html
+++ b/docs/master/_modules/torchvision/models/shufflenetv2.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/squeezenet.html b/docs/master/_modules/torchvision/models/squeezenet.html
index 076cff491531..91f07cb94bb2 100644
--- a/docs/master/_modules/torchvision/models/squeezenet.html
+++ b/docs/master/_modules/torchvision/models/squeezenet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/vgg.html b/docs/master/_modules/torchvision/models/vgg.html
index 6e1ca5eb52c7..3874f39968bf 100644
--- a/docs/master/_modules/torchvision/models/vgg.html
+++ b/docs/master/_modules/torchvision/models/vgg.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/models/video/resnet.html b/docs/master/_modules/torchvision/models/video/resnet.html
index a61088a343f4..4890379a9f43 100644
--- a/docs/master/_modules/torchvision/models/video/resnet.html
+++ b/docs/master/_modules/torchvision/models/video/resnet.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/ops/roi_align.html b/docs/master/_modules/torchvision/ops/roi_align.html
index 7ba08ff67d84..2b0d81e96ec1 100644
--- a/docs/master/_modules/torchvision/ops/roi_align.html
+++ b/docs/master/_modules/torchvision/ops/roi_align.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/ops/roi_pool.html b/docs/master/_modules/torchvision/ops/roi_pool.html
index 116b6cab65cd..115f83bbe900 100644
--- a/docs/master/_modules/torchvision/ops/roi_pool.html
+++ b/docs/master/_modules/torchvision/ops/roi_pool.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/transforms/functional.html b/docs/master/_modules/torchvision/transforms/functional.html
index be5da6356dba..1f63b32c48f0 100644
--- a/docs/master/_modules/torchvision/transforms/functional.html
+++ b/docs/master/_modules/torchvision/transforms/functional.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/transforms/transforms.html b/docs/master/_modules/torchvision/transforms/transforms.html
index e74ee8dba43f..17bf58983221 100644
--- a/docs/master/_modules/torchvision/transforms/transforms.html
+++ b/docs/master/_modules/torchvision/transforms/transforms.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_modules/torchvision/utils.html b/docs/master/_modules/torchvision/utils.html
index 508f84fdf38a..a158a9534ffa 100644
--- a/docs/master/_modules/torchvision/utils.html
+++ b/docs/master/_modules/torchvision/utils.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/_sources/cudnn_rnn_determinism.rst.txt b/docs/master/_sources/cudnn_rnn_determinism.rst.txt
new file mode 100644
index 000000000000..6568b73ef097
--- /dev/null
+++ b/docs/master/_sources/cudnn_rnn_determinism.rst.txt
@@ -0,0 +1,16 @@
+.. warning::
+ There are known non-determinism issues for RNN functions on some versions of cuDNN and CUDA.
+ You can enforce deterministic behavior by setting the following environment variables:
+
+ On CUDA 10.1, set environment variable ``CUDA_LAUNCH_BLOCKING=1``.
+ This may affect performance.
+
+ On CUDA 10.2 or later, set environment variable
+ (note the leading colon symbol)
+ ``CUBLAS_WORKSPACE_CONFIG=:16:8``
+ or
+ ``CUBLAS_WORKSPACE_CONFIG=:4096:2``.
+
+ See the `cuDNN 8 Release Notes`_ for more information.
+
+.. _cuDNN 8 Release Notes: https://docs.nvidia.com/deeplearning/sdk/cudnn-release-notes/rel_8.html
diff --git a/docs/master/_sources/generated/torch.nn.Unflatten.rst.txt b/docs/master/_sources/generated/torch.nn.Unflatten.rst.txt
new file mode 100644
index 000000000000..e593b22fc064
--- /dev/null
+++ b/docs/master/_sources/generated/torch.nn.Unflatten.rst.txt
@@ -0,0 +1,13 @@
+.. role:: hidden
+ :class: hidden-section
+.. currentmodule:: torch.nn
+
+
+Unflatten
+=========
+
+.. autoclass:: Unflatten
+ :inherited-members:
+ :members:
+
+.. autogenerated from source/_templates/autosummary/class.rst
\ No newline at end of file
diff --git a/docs/master/_sources/nn.rst.txt b/docs/master/_sources/nn.rst.txt
index 84535d16cc82..3a6cb7e19316 100644
--- a/docs/master/_sources/nn.rst.txt
+++ b/docs/master/_sources/nn.rst.txt
@@ -355,6 +355,7 @@ Utility functions in other modules
nn.utils.rnn.pack_sequence
nn.Flatten
+ nn.Unflatten
Quantized Functions
--------------------
diff --git a/docs/master/_sources/notes/randomness.rst.txt b/docs/master/_sources/notes/randomness.rst.txt
index bddc57e993c1..de2a2039efe0 100644
--- a/docs/master/_sources/notes/randomness.rst.txt
+++ b/docs/master/_sources/notes/randomness.rst.txt
@@ -47,13 +47,16 @@ is implemented using :meth:`index_select`, the backward path for
which is implemented using :meth:`index_add_`, which is known to operate
nondeterministically (in the forward direction) on the CUDA backend (see above).
-CuDNN
+cuDNN
.....
-When running on the CuDNN backend, two further options must be set::
+When running on the cuDNN backend, two further options must be set::
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
+On some versions of cuDNN and CUDA, RNN and LSTM may have non-deterministic behavior.
+See :meth:`torch.nn.RNN` and :meth:`torch.nn.LSTM` for details and workarounds.
+
.. warning::
Deterministic operation may have a negative single-run performance impact,
diff --git a/docs/master/_sources/tensor_view.rst.txt b/docs/master/_sources/tensor_view.rst.txt
index 502743dcbd4f..f310696f7113 100644
--- a/docs/master/_sources/tensor_view.rst.txt
+++ b/docs/master/_sources/tensor_view.rst.txt
@@ -70,6 +70,7 @@ For reference, here’s a full list of view ops in PyTorch:
- :meth:`~torch.Tensor.view_as`
- :meth:`~torch.Tensor.unbind`
- :meth:`~torch.Tensor.split`
+- :meth:`~torch.Tensor.split_with_sizes`
- :meth:`~torch.Tensor.chunk`
- :meth:`~torch.Tensor.indices` (sparse tensor only)
- :meth:`~torch.Tensor.values` (sparse tensor only)
diff --git a/docs/master/amp.html b/docs/master/amp.html
index 4bfe1a3f81f1..0b5813a77058 100644
--- a/docs/master/amp.html
+++ b/docs/master/amp.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/autograd.html b/docs/master/autograd.html
index 62603d00f5ef..30e867e85748 100644
--- a/docs/master/autograd.html
+++ b/docs/master/autograd.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/bottleneck.html b/docs/master/bottleneck.html
index 12f4e4ac46fa..102a669508e8 100644
--- a/docs/master/bottleneck.html
+++ b/docs/master/bottleneck.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/checkpoint.html b/docs/master/checkpoint.html
index 750410753d68..740a552af2fd 100644
--- a/docs/master/checkpoint.html
+++ b/docs/master/checkpoint.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/community/contribution_guide.html b/docs/master/community/contribution_guide.html
index 91e4a7816cb7..50eed11cb503 100644
--- a/docs/master/community/contribution_guide.html
+++ b/docs/master/community/contribution_guide.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/community/governance.html b/docs/master/community/governance.html
index ecfcd4cd7a51..aa90f91ca7ac 100644
--- a/docs/master/community/governance.html
+++ b/docs/master/community/governance.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/community/persons_of_interest.html b/docs/master/community/persons_of_interest.html
index 6e34b5aa9617..93e5ab50a8b1 100644
--- a/docs/master/community/persons_of_interest.html
+++ b/docs/master/community/persons_of_interest.html
@@ -158,7 +158,7 @@
diff --git a/docs/master/complex_numbers.html b/docs/master/complex_numbers.html
index 7a5b5eceb1a7..f7aebb90df01 100644
--- a/docs/master/complex_numbers.html
+++ b/docs/master/complex_numbers.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/cpp_extension.html b/docs/master/cpp_extension.html
index d4fd26522f85..f231815fc502 100644
--- a/docs/master/cpp_extension.html
+++ b/docs/master/cpp_extension.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/cpp_index.html b/docs/master/cpp_index.html
index 30508cfb4918..081ff0d9648f 100644
--- a/docs/master/cpp_index.html
+++ b/docs/master/cpp_index.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/cuda.html b/docs/master/cuda.html
index 1a575438fabe..38f565cf9f46 100644
--- a/docs/master/cuda.html
+++ b/docs/master/cuda.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/cudnn_persistent_rnn.html b/docs/master/cudnn_persistent_rnn.html
index a7f4592af4eb..398e0b556314 100644
--- a/docs/master/cudnn_persistent_rnn.html
+++ b/docs/master/cudnn_persistent_rnn.html
@@ -157,7 +157,7 @@
diff --git a/docs/master/cudnn_rnn_determinism.html b/docs/master/cudnn_rnn_determinism.html
new file mode 100644
index 000000000000..517abdf67da8
--- /dev/null
+++ b/docs/master/cudnn_rnn_determinism.html
@@ -0,0 +1,639 @@
+
+
+
+
+
+
+
+ <no title> — PyTorch master documentation
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ Docs
+
+ >
+
+
+
+ <no title>
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ Shortcuts
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Warning
+
There are known non-determinism issues for RNN functions on some versions of cuDNN and CUDA.
+You can enforce deterministic behavior by setting the following environment variables:
+
On CUDA 10.1, set environment variable CUDA_LAUNCH_BLOCKING=1
+
On CUDA 10.2 or later, set environment variable
+(note the leading colon symbol)
+CUBLAS_WORKSPACE_CONFIG=:16:8 CUBLAS_WORKSPACE_CONFIG=:4096:2
+
See the cuDNN 8 Release Notes for more information.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Docs
+
Access comprehensive developer documentation for PyTorch
+
View Docs
+
+
+
+
Tutorials
+
Get in-depth tutorials for beginners and advanced developers
+
View Tutorials
+
+
+
+
Resources
+
Find development resources and get your questions answered
+
View Resources
+
+
+
+
+
+
To analyze traffic and optimize your experience, we serve cookies on this site. By clicking or navigating, you agree to allow our usage of cookies. As the current maintainers of this site, Facebook’s Cookies Policy applies. Learn more, including about available controls: Cookies Policy .
+
+
+
Flatten
+
Warning
+
There are known non-determinism issues for RNN functions on some versions of cuDNN and CUDA.
+You can enforce deterministic behavior by setting the following environment variables:
+
On CUDA 10.1, set environment variable CUDA_LAUNCH_BLOCKING=1
+
On CUDA 10.2 or later, set environment variable
+(note the leading colon symbol)
+CUBLAS_WORKSPACE_CONFIG=:16:8 CUBLAS_WORKSPACE_CONFIG=:4096:2
+
See the cuDNN 8 Release Notes for more information.
+
Orphan
RNN
+
Warning
+
There are known non-determinism issues for RNN functions on some versions of cuDNN and CUDA.
+You can enforce deterministic behavior by setting the following environment variables:
+
On CUDA 10.1, set environment variable CUDA_LAUNCH_BLOCKING=1
+
On CUDA 10.2 or later, set environment variable
+(note the leading colon symbol)
+CUBLAS_WORKSPACE_CONFIG=:16:8 CUBLAS_WORKSPACE_CONFIG=:4096:2
+
See the cuDNN 8 Release Notes for more information.
+
Orphan
Unflatten — PyTorch master documentation
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+ Shortcuts
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Unflatten
+
+
+class torch.nn.Unflatten( dim: Union[int, str], unflattened_size: tuple ) [source]
+Unflattens a tensor into another tensor of a desired shape. For use with Sequential
+
+be either str or int when NamedTensor or Tensor is used, respectively.
+
+namedshape (tuple of tuples) if dim str or a tuple of ints as well as torch.Size if
+dim int .
+
+Shape:
+Input: ( N , ∗ d i m s ) (N, *dims) ( N , ∗ d i m s )
Output: ( N , C o u t , H o u t , W o u t ) (N, C_out, H_out, W_out) ( N , C o u t , H o u t , W o u t )
+
+
+Parameters
+
+dim (Union [ int , str ] ) – Dimension to be flattened
unflattened_size (Union [ tuple , torch.Size ] ) – Size of the output tensor
+
+Examples
+>>> input = torch . randn ( 2 , 50 )
+>>> # With tuple of ints
+>>> m = nn . Sequential (
+>>> nn . Linear ( 50 , 50 ),
+>>> nn . Unflatten ( 1 , ( 2 , 5 , 5 ))
+>>> )
+>>> output = m ( output )
+>>> output . size ()
+torch.Size([2, 2, 5, 5])
+>>> # With torch.Size
+>>> m = nn . Sequential (
+>>> nn . Linear ( 50 , 50 ),
+>>> nn . Unflatten ( 1 , torch . Size ([ 2 , 5 , 5 ]))
+>>> )
+>>> output = m ( output )
+>>> output . size ()
+torch.Size([2, 2, 5, 5])
+>>> # With namedshape (tuple of tuples)
+>>> m = nn . Sequential (
+>>> nn . Linear ( 50 , 50 ),
+>>> nn . Unflatten ( 'features' , (( 'C' , 2 ), ( 'H' , 50 ), ( 'W' , 50 )))
+>>> )
+>>> output = m ( output )
+>>> output . size ()
+torch.Size([2, 2, 5, 5])
+
+
+
+add_module( name: str, module: Optional[Module] ) → None
+Adds a child module to the current module.
+The module can be accessed as an attribute using the given name.
+
+Parameters
+
+
+
+
+
+
+
+apply( fn: Callable[Module, None] ) → T
+Applies fn .children() torch.nn.init
+
+Parameters
+fn (Module
+Returns
+self
+Return type
+Module
+
+Example:
+>>> @torch . no_grad ()
+>>> def init_weights ( m ):
+>>> print ( m )
+>>> if type ( m ) == nn . Linear :
+>>> m . weight . fill_ ( 1.0 )
+>>> print ( m . weight )
+>>> net = nn . Sequential ( nn . Linear ( 2 , 2 ), nn . Linear ( 2 , 2 ))
+>>> net . apply ( init_weights )
+Linear(in_features=2, out_features=2, bias=True)
+Parameter containing:
+tensor([[ 1., 1.],
+ [ 1., 1.]])
+Linear(in_features=2, out_features=2, bias=True)
+Parameter containing:
+tensor([[ 1., 1.],
+ [ 1., 1.]])
+Sequential(
+ (0): Linear(in_features=2, out_features=2, bias=True)
+ (1): Linear(in_features=2, out_features=2, bias=True)
+)
+Sequential(
+ (0): Linear(in_features=2, out_features=2, bias=True)
+ (1): Linear(in_features=2, out_features=2, bias=True)
+)
+
+
+
+
+
+bfloat16( ) → T
+Casts all floating point parameters and buffers to bfloat16
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+buffers( recurse: bool = True ) → Iterator[torch.Tensor]
+Returns an iterator over module buffers.
+
+Parameters
+recurse (bool
+Yields
+torch.Tensor – module buffer
+
+Example:
+>>> for buf in model . buffers ():
+>>> print ( type ( buf ), buf . size ())
+<class 'torch.Tensor'> (20L,)
+<class 'torch.Tensor'> (20L, 1L, 5L, 5L)
+
+
+
+
+
+children( ) → Iterator[torch.nn.modules.module.Module]
+Returns an iterator over immediate children modules.
+
+Yields
+Module – a child module
+
+
+
+
+
+cpu( ) → T
+Moves all model parameters and buffers to the CPU.
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+cuda( device: Union[int , torch.device , None] = None ) → T
+Moves all model parameters and buffers to the GPU.
+This also makes associated parameters and buffers different objects. So
+it should be called before constructing optimizer if the module will
+live on GPU while being optimized.
+
+Parameters
+device (int , optional ) – if specified, all parameters will be
+copied to that device
+Returns
+self
+Return type
+Module
+
+
+
+
+
+double( ) → T
+Casts all floating point parameters and buffers to double
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+eval( ) → T
+Sets the module in evaluation mode.
+This has any effect only on certain modules. See documentations of
+particular modules for details of their behaviors in training/evaluation
+mode, if they are affected, e.g. Dropout BatchNorm
+This is equivalent with self.train(False)
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+float( ) → T
+Casts all floating point parameters and buffers to float datatype.
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+half( ) → T
+Casts all floating point parameters and buffers to half
+
+Returns
+self
+Return type
+Module
+
+
+
+
+
+load_state_dict( state_dict: Dict[str, torch.Tensor], strict: bool = True )
+Copies parameters and buffers from state_dict strict True state_dict state_dict()
+
+Parameters
+
+state_dict (dict
strict (bool , optional ) – whether to strictly enforce that the keys
+in state_dict state_dict() True
+Returns
+
+
+Return type
+NamedTuple missing_keys unexpected_keys
+
+
+
+
+
+modules( ) → Iterator[torch.nn.modules.module.Module]
+Returns an iterator over all modules in the network.
+
+Yields
+Module – a module in the network
+
+
+
Note
+
Duplicate modules are returned only once. In the following
+example, l
+
Example:
+>>> l = nn . Linear ( 2 , 2 )
+>>> net = nn . Sequential ( l , l )
+>>> for idx , m in enumerate ( net . modules ()):
+ print(idx, '->', m)
+
+0 -> Sequential(
+ (0): Linear(in_features=2, out_features=2, bias=True)
+ (1): Linear(in_features=2, out_features=2, bias=True)
+)
+1 -> Linear(in_features=2, out_features=2, bias=True)
+
+
+
+
+
+named_buffers( prefix: str = '' , recurse: bool = True ) → Iterator[Tuple[str, torch.Tensor]]
+Returns an iterator over module buffers, yielding both the
+name of the buffer as well as the buffer itself.
+
+Parameters
+
+prefix (str
recurse (bool
+Yields
+(string, torch.Tensor) – Tuple containing the name and buffer
+
+Example:
+>>> for name , buf in self . named_buffers ():
+>>> if name in [ 'running_var' ]:
+>>> print ( buf . size ())
+
+
+
+
+
+named_children( ) → Iterator[Tuple[str, torch.nn.modules.module.Module]]
+Returns an iterator over immediate children modules, yielding both
+the name of the module as well as the module itself.
+
+Yields
+(string, Module) – Tuple containing a name and child module
+
+Example:
+>>> for name , module in model . named_children ():
+>>> if name in [ 'conv4' , 'conv5' ]:
+>>> print ( module )
+
+
+
+
+
+named_modules( memo: Optional[Set[Module]] = None , prefix: str = '' )
+Returns an iterator over all modules in the network, yielding
+both the name of the module as well as the module itself.
+
+Yields
+(string, Module) – Tuple of name and module
+
+
+
Note
+
Duplicate modules are returned only once. In the following
+example, l
+
Example:
+>>> l = nn . Linear ( 2 , 2 )
+>>> net = nn . Sequential ( l , l )
+>>> for idx , m in enumerate ( net . named_modules ()):
+ print(idx, '->', m)
+
+0 -> ('', Sequential(
+ (0): Linear(in_features=2, out_features=2, bias=True)
+ (1): Linear(in_features=2, out_features=2, bias=True)
+))
+1 -> ('0', Linear(in_features=2, out_features=2, bias=True))
+
+
+
+
+
+named_parameters( prefix: str = '' , recurse: bool = True ) → Iterator[Tuple[str, torch.Tensor]]
+Returns an iterator over module parameters, yielding both the
+name of the parameter as well as the parameter itself.
+
+Parameters
+
+prefix (str
recurse (bool
+Yields
+(string, Parameter) – Tuple containing the name and parameter
+
+Example:
+>>> for name , param in self . named_parameters ():
+>>> if name in [ 'bias' ]:
+>>> print ( param . size ())
+
+
+
+
+
+parameters( recurse: bool = True ) → Iterator[torch.nn.parameter.Parameter]
+Returns an iterator over module parameters.
+This is typically passed to an optimizer.
+
+Parameters
+recurse (bool
+Yields
+Parameter – module parameter
+
+Example:
+>>> for param in model . parameters ():
+>>> print ( type ( param ), param . size ())
+<class 'torch.Tensor'> (20L,)
+<class 'torch.Tensor'> (20L, 1L, 5L, 5L)
+
+
+
+
+
+register_backward_hook( hook: Callable[[Module, Union[Tuple[torch.Tensor, ...], torch.Tensor], Union[Tuple[torch.Tensor, ...], torch.Tensor]], Union[None, torch.Tensor]] ) → torch.utils.hooks.RemovableHandle
+Registers a backward hook on the module.
+
+
Warning
+
The current implementation will not have the presented behavior
+for complex Module grad_input grad_output Module torch.Tensor.register_hook()
+
The hook will be called every time the gradients with respect to module
+inputs are computed. The hook should have the following signature:
+hook ( module , grad_input , grad_output ) -> Tensor or None
+
+
The grad_input grad_output grad_input grad_input
+
+Returns
+a handle that can be used to remove the added hook by calling
+handle.remove()
+Return type
+torch.utils.hooks.RemovableHandle
+
+
+
+
+
+register_buffer( name: str, tensor: Optional[torch.Tensor], persistent: bool = True ) → None
+Adds a buffer to the module.
+This is typically used to register a buffer that should not to be
+considered a model parameter. For example, BatchNorm’s running_mean persistent False state_dict
+Buffers can be accessed as attributes using given names.
+
+Parameters
+
+name (string ) – name of the buffer. The buffer can be accessed
+from this module using the given name
tensor (Tensor
persistent (bool state_dict
+
+Example:
+>>> self . register_buffer ( 'running_mean' , torch . zeros ( num_features ))
+
+
+
+
+
+register_forward_hook( hook: Callable[..., None] ) → torch.utils.hooks.RemovableHandle
+Registers a forward hook on the module.
+The hook will be called every time after forward()
+hook ( module , input , output ) -> None or modified output
+
+
The input contains only the positional arguments given to the module.
+Keyword arguments won’t be passed to the hooks and only to the forward forward()
+
+Returns
+a handle that can be used to remove the added hook by calling
+handle.remove()
+Return type
+torch.utils.hooks.RemovableHandle
+
+
+
+
+
+register_forward_pre_hook( hook: Callable[..., None] ) → torch.utils.hooks.RemovableHandle
+Registers a forward pre-hook on the module.
+The hook will be called every time before forward()
+hook ( module , input ) -> None or modified input
+
+
The input contains only the positional arguments given to the module.
+Keyword arguments won’t be passed to the hooks and only to the forward
+
+Returns
+a handle that can be used to remove the added hook by calling
+handle.remove()
+Return type
+torch.utils.hooks.RemovableHandle
+
+
+
+
+
+register_parameter( name: str, param: Optional[torch.nn.parameter.Parameter] ) → None
+Adds a parameter to the module.
+The parameter can be accessed as an attribute using given name.
+
+Parameters
+
+
+
+
+
+
+
+requires_grad_( requires_grad: bool = True ) → T
+Change if autograd should record operations on parameters in this
+module.
+This method sets the parameters’ requires_grad
+This method is helpful for freezing part of the module for finetuning
+or training parts of a model individually (e.g., GAN training).
+
+Parameters
+requires_grad (bool True
+Returns
+self
+Return type
+Module
+
+
+
+
+
+state_dict( destination=None , prefix='' , keep_vars=False )
+Returns a dictionary containing a whole state of the module.
+Both parameters and persistent buffers (e.g. running averages) are
+included. Keys are corresponding parameter and buffer names.
+
+Returns
+a dictionary containing a whole state of the module
+Return type
+dict
+
+Example:
+>>> module . state_dict () . keys ()
+['bias', 'weight']
+
+
+
+
+
+to( *args , **kwargs )
+Moves and/or casts the parameters and buffers.
+This can be called as
+
+
+to( device=None , dtype=None , non_blocking=False )
+
+
+
+
+to( dtype , non_blocking=False )
+
+
+
+
+to( tensor , non_blocking=False )
+
+
+
+
+to( memory_format=torch.channels_last )
+
+
+Its signature is similar to torch.Tensor.to() dtype dtype device non_blocking
+See below for examples.
+
+
Note
+
This method modifies the module in-place.
+
+Parameters
+
+device (torch.device
dtype (torch.dtype
tensor (torch.Tensor
memory_format (torch.memory_format
+Returns
+self
+Return type
+Module
+
+Example:
+>>> linear = nn . Linear ( 2 , 2 )
+>>> linear . weight
+Parameter containing:
+tensor([[ 0.1913, -0.3420],
+ [-0.5113, -0.2325]])
+>>> linear . to ( torch . double )
+Linear(in_features=2, out_features=2, bias=True)
+>>> linear . weight
+Parameter containing:
+tensor([[ 0.1913, -0.3420],
+ [-0.5113, -0.2325]], dtype=torch.float64)
+>>> gpu1 = torch . device ( "cuda:1" )
+>>> linear . to ( gpu1 , dtype = torch . half , non_blocking = True )
+Linear(in_features=2, out_features=2, bias=True)
+>>> linear . weight
+Parameter containing:
+tensor([[ 0.1914, -0.3420],
+ [-0.5112, -0.2324]], dtype=torch.float16, device='cuda:1')
+>>> cpu = torch . device ( "cpu" )
+>>> linear . to ( cpu )
+Linear(in_features=2, out_features=2, bias=True)
+>>> linear . weight
+Parameter containing:
+tensor([[ 0.1914, -0.3420],
+ [-0.5112, -0.2324]], dtype=torch.float16)
+
+
+
+
+
+train( mode: bool = True ) → T
+Sets the module in training mode.
+This has any effect only on certain modules. See documentations of
+particular modules for details of their behaviors in training/evaluation
+mode, if they are affected, e.g. Dropout BatchNorm
+
+Parameters
+mode (bool True False True
+Returns
+self
+Return type
+Module
+
+
+
+
+
+type( dst_type: Union[torch.dtype, str] ) → T
+Casts all parameters and buffers to dst_type
+
+Parameters
+dst_type (type or string ) – the desired type
+Returns
+self
+Return type
+Module
+
+
+
+
+
+zero_grad( ) → None
+Sets gradients of all model parameters to zero.
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Docs
+
Access comprehensive developer documentation for PyTorch
+
View Docs
+
+
+
+
Tutorials
+
Get in-depth tutorials for beginners and advanced developers
+
View Tutorials
+
+
+
+
Resources
+
Find development resources and get your questions answered
+
View Resources
+
+
+
+
+
+
To analyze traffic and optimize your experience, we serve cookies on this site. By clicking or navigating, you agree to allow our usage of cookies. As the current maintainers of this site, Facebook’s Cookies Policy applies. Learn more, including about available controls: Cookies Policy .
+
+
+
A
(torch.nn.Flatten method)
(torch.nn.Module method)
+
+
(torch.nn.Unflatten method)
add_observer_() (in module torch.quantization)
@@ -648,6 +650,8 @@ A
(torch.nn.Flatten method)
(torch.nn.Module method)
+
+
(torch.nn.Unflatten method)
(torch.nn.utils.prune.BasePruningMethod class method)
@@ -940,6 +944,8 @@
B
(torch.nn.Flatten method)
(torch.nn.Module method)
+
+
(torch.nn.Unflatten method)
(torch.Tensor method)
@@ -1036,6 +1042,8 @@
B
(torch.nn.Flatten method)
(torch.nn.Module method)
+
+
(torch.nn.Unflatten method)
BuildExtension() (in module torch.utils.cpp_extension)
@@ -1140,6 +1148,8 @@ C
(torch.nn.Flatten method)
(torch.nn.Module method)
+
+
(torch.nn.Unflatten method)
cholesky() (in module torch)
@@ -1260,10 +1270,10 @@ C
Constraint (class in torch.distributions.constraints)
-
-
+
Event (class in torch.cuda)
@@ -1998,8 +2016,6 @@ E
extra_repr() (torch.jit.ScriptModule method)
F
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.Tensor method)
F
(torch.Tensor method)
floor_() (torch.Tensor method)
-
named_children() (torch.jit.ScriptModule method)
@@ -3645,6 +3671,8 @@ N
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
named_modules() (torch.jit.ScriptModule method)
@@ -3653,6 +3681,8 @@ N
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
named_parameters() (torch.jit.ScriptModule method)
@@ -3661,6 +3691,8 @@ N
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
names (torch.Tensor attribute)
@@ -3699,12 +3731,12 @@ N
NegativeBinomial (class in torch.distributions.negative_binomial)
+
-
parameters_to_vector() (in module torch.nn.utils)
@@ -3920,11 +3954,11 @@ P
Poisson (class in torch.distributions.poisson)
poisson() (in module torch)
-poisson_nll_loss() (in module torch.nn.functional)
register_buffer() (torch.jit.ScriptModule method)
@@ -4225,6 +4261,8 @@ R
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
register_custom_op_symbolic() (in module torch.onnx)
@@ -4235,6 +4273,8 @@ R
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
register_forward_pre_hook() (torch.jit.ScriptModule method)
@@ -4243,6 +4283,8 @@ R
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
register_hook() (torch.Tensor method)
@@ -4255,6 +4297,8 @@ R
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
RelaxedBernoulli (class in torch.distributions.relaxed_bernoulli)
@@ -4357,6 +4401,8 @@ R
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.Tensor method)
S
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.optim.lr_scheduler.LambdaLR method)
T
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.nn.utils.rnn.PackedSequence method)
T
torch.finfo (class in torch)
-
+
Transform (class in torch.distributions.transforms)
@@ -5461,6 +5513,8 @@ T
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.Tensor method)
U
(torch.Tensor method)
Unflatten (class in torch.nn)
+unflatten() (torch.Tensor method)
+Unflatten.to() (in module torch.nn) , [1] , [2] , [3]
Unfold (class in torch.nn)
U
unscale_() (torch.cuda.amp.GradScaler method)
-unsorted_indices() (torch.nn.utils.rnn.PackedSequence property)
+ unsorted_indices() (torch.nn.utils.rnn.PackedSequence property)
+unsqueeze() (in module torch)
@@ -5767,6 +5825,8 @@ Z
(torch.nn.Flatten method)
(torch.nn.Module method)
+(torch.nn.Unflatten method)
(torch.optim.Optimizer method)
return_counts : bool = False ,
dim : Optional [ int ]) -> Tuple [ Tensor , Tensor , Tensor ]
+Tensor . unsafe_chunk ( chunks : int ,
+ dim : int = 0 ) -> List [ Tensor ]
+
+Tensor . unsafe_split ( split_size : int ,
+ dim : int = 0 ) -> List [ Tensor ]
+
+Tensor . unsafe_split_with_sizes ( split_sizes : List [ int ],
+ dim : int = 0 ) -> List [ Tensor ]
+
Tensor . unsqueeze ( dim : int ) -> Tensor
Tensor . unsqueeze_ ( dim : int ) -> Tensor
@@ -6919,6 +6928,18 @@
torch . unbind ( self : Tensor ,
dim : int = 0 ) -> List [ Tensor ]
+torch . unsafe_chunk ( self : Tensor ,
+ chunks : int ,
+ dim : int = 0 ) -> List [ Tensor ]
+
+torch . unsafe_split ( self : Tensor ,
+ split_size : int ,
+ dim : int = 0 ) -> List [ Tensor ]
+
+torch . unsafe_split_with_sizes ( self : Tensor ,
+ split_sizes : List [ int ],
+ dim : int = 0 ) -> List [ Tensor ]
+
torch . unsqueeze ( self : Tensor ,
dim : int ) -> Tensor
diff --git a/docs/master/jit_language_reference.html b/docs/master/jit_language_reference.html
index e1718e7c090c..4ed427a78543 100644
--- a/docs/master/jit_language_reference.html
+++ b/docs/master/jit_language_reference.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/jit_python_reference.html b/docs/master/jit_python_reference.html
index 965820d4f715..9d9f9131b281 100644
--- a/docs/master/jit_python_reference.html
+++ b/docs/master/jit_python_reference.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/jit_unsupported.html b/docs/master/jit_unsupported.html
index ab5ea400b0c1..b96159ee2acf 100644
--- a/docs/master/jit_unsupported.html
+++ b/docs/master/jit_unsupported.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/mobile_optimizer.html b/docs/master/mobile_optimizer.html
index 9f9d43978e74..bf39824ab2b4 100644
--- a/docs/master/mobile_optimizer.html
+++ b/docs/master/mobile_optimizer.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/model_zoo.html b/docs/master/model_zoo.html
index d1fa25085ba5..2579bd4e8258 100644
--- a/docs/master/model_zoo.html
+++ b/docs/master/model_zoo.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/multiprocessing.html b/docs/master/multiprocessing.html
index fa342ccd4e56..9402f7ae16db 100644
--- a/docs/master/multiprocessing.html
+++ b/docs/master/multiprocessing.html
@@ -157,7 +157,7 @@
diff --git a/docs/master/name_inference.html b/docs/master/name_inference.html
index d3279f2c7c85..8adac2c552cd 100644
--- a/docs/master/name_inference.html
+++ b/docs/master/name_inference.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/named_tensor.html b/docs/master/named_tensor.html
index 40b0d436d5a2..3e19bf11ea5c 100644
--- a/docs/master/named_tensor.html
+++ b/docs/master/named_tensor.html
@@ -159,7 +159,7 @@
diff --git a/docs/master/nn.functional.html b/docs/master/nn.functional.html
index 3f60ce96f4e5..ea3d1cfa6b09 100644
--- a/docs/master/nn.functional.html
+++ b/docs/master/nn.functional.html
@@ -36,7 +36,7 @@
data_parallel
- Utilities nn.Flatten
Flattens a contiguous range of dims into a tensor.
nn.Unflatten
Unflattens a tensor into another tensor of a desired shape.
PyTorch 
+
+ 
+ 
