frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 12:54:11 +08:00
Code Issues Packages Projects Releases Wiki Activity

Render Example: and not Example:: in docs (#153978)

Browse Source 
Everything here is a grep except the changes in tools/autograd/load_derivatives.py which I manually corrected.

The correct notation is:
```
Example::

    >>> ...
```

It is common and wrong to have:
```
Example::
    >>> ...
```

In the wrong example, we get these pesky double colons:
![image](https://github.com/user-attachments/assets/20ffd349-68bb-4552-966c-e23923350476)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/153978
Approved by: https://github.com/soulitzer, https://github.com/malfet
This commit is contained in:
Jane Xu
2025-05-20 13:27:28 -07:00
committed by PyTorch MergeBot
parent 0959869683
commit 8817e5ac80
11 changed files with 51 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
tools/autograd/load_derivatives.py
View File

@ -986,7 +986,7 @@ def saved_variables(
def _create_op_prefix(name: str) -> str:
"""Takes a native function name converts to a op prefix name.
r"""Takes a native function name converts to an op prefix name.
Note that the "name" parameter must be the native function name
without the optional variant suffix, so "add" instead of
@ -995,8 +995,9 @@ def _create_op_prefix(name: str) -> str:
OP names correspond to classes, hence the change to title case.
Example::
>>> _create_op_prefix("add")
'AddBackward'
>>> _create_op_prefix("add")
'AddBackward'
"""
camel_case = "".join([p.title() for p in name.split("_")])
return (camel_case + "Backward").replace("ForwardBackward", "Backward")

2
torch/_custom_ops.py
View File

@ -52,6 +52,7 @@ def custom_op(qualname, func_or_schema=None):
schema string.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_CUDA)
>>> import torch
>>> import numpy as np
@ -134,6 +135,7 @@ def impl(qualname, *, device_types=("cpu", "cuda"), func=None):
device_types (str or Iterable[str]): the device type(s) to register the function for.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_CUDA)
>>> import torch
>>> import numpy as np

1
torch/_functorch/eager_transforms.py
View File

@ -1701,6 +1701,7 @@ def linearize(func: Callable, *primals) -> tuple[Any, Callable]:
with a single evaluation.
Example::
>>> import torch
>>> from torch.func import linearize
>>> def fn(x):

1
torch/_size_docs.py
View File

@ -19,6 +19,7 @@ More formally, for a tensor ``x = tensor.ones(10, 10)`` with size ``s = torch.Si
``x.numel() == x.size().numel() == s.numel() == 100`` holds true.
Example::
>>> x=torch.ones(10, 10)
>>> s=x.size()
>>> s

5
torch/_tensor_docs.py
View File

@ -2474,6 +2474,7 @@ Args:
value (float): the value to fill with
Example::
>>> x = torch.tensor([[1, 2, 3], [4, 5, 6], [7, 8, 9]], dtype=torch.float)
>>> index = torch.tensor([0, 2])
>>> x.index_fill_(1, index, -1)
@ -6871,6 +6872,7 @@ The returned tensor and :attr:`self` share the same underlying storage.
Returns :attr:`self` if :attr:`self` is a real-valued tensor tensor.
Example::
>>> x=torch.randn(4, dtype=torch.cfloat)
>>> x
tensor([(0.3100+0.3553j), (-0.5445-0.7896j), (-1.6492-0.0633j), (-0.0638-0.8119j)])
@ -6890,6 +6892,7 @@ The returned tensor and :attr:`self` share the same underlying storage.
:func:`imag` is only supported for tensors with complex dtypes.
Example::
>>> x=torch.randn(4, dtype=torch.cfloat)
>>> x
tensor([(0.3100+0.3553j), (-0.5445-0.7896j), (-1.6492-0.0633j), (-0.0638-0.8119j)])
@ -6923,6 +6926,7 @@ matrix multiplication, it is necessary to use ``int32`` indexing in order
to avoid downcasting and potentially losing information.
Example::
>>> csr = torch.eye(5,5).to_sparse_csr()
>>> csr.crow_indices()
tensor([0, 1, 2, 3, 4, 5], dtype=torch.int32)
@ -6943,6 +6947,7 @@ matrix multiplication, it is necessary to use ``int32`` indexing in order
to avoid downcasting and potentially losing information.
Example::
>>> csr = torch.eye(5,5).to_sparse_csr()
>>> csr.col_indices()
tensor([0, 1, 2, 3, 4], dtype=torch.int32)

12
torch/_torch_docs.py
View File

@ -611,6 +611,7 @@ Args:
{input}
Example::
>>> x = torch.arange(4, dtype=torch.float)
>>> A = torch.complex(x, x).reshape(2, 2)
>>> A
@ -2050,6 +2051,7 @@ Args:
indices_or_sections (int or list or tuple of ints): See argument in :func:`torch.tensor_split`.
Example::
>>> t = torch.arange(16.0).reshape(4,4)
>>> t
tensor([[ 0., 1., 2., 3.],
@ -2099,6 +2101,7 @@ Args:
indices_or_sections (int or list or tuple of ints): See argument in :func:`torch.tensor_split`.
Example::
>>> t = torch.arange(16.0).reshape(4,4)
>>> t
tensor([[ 0., 1., 2., 3.],
@ -2140,6 +2143,7 @@ Args:
indices_or_sections (int or list or tuple of ints): See argument in :func:`torch.tensor_split`.
Example::
>>> t = torch.arange(16.0).reshape(2, 2, 4)
>>> t
tensor([[[ 0., 1., 2., 3.],
@ -2295,6 +2299,7 @@ Returns:
:func:`torch.corrcoef` normalized covariance matrix.
Example::
>>> x = torch.tensor([[0, 2], [1, 1], [2, 0]]).T
>>> x
tensor([[0, 1, 2],
@ -4477,6 +4482,7 @@ Keyword args:
{pin_memory}
Example::
>>> t = torch.randn(2, 5, dtype=torch.float64)
>>> t.numpy().tofile('storage.pt')
>>> t_mapped = torch.from_file('storage.pt', shared=False, size=10, dtype=torch.float64)
@ -5148,6 +5154,7 @@ Returns:
bin_edges(Tensor[]): sequence of N 1D Tensors containing the bin edges.
Example::
>>> torch.histogramdd(torch.tensor([[0., 1.], [1., 0.], [2., 0.], [2., 2.]]), bins=[3, 3],
... weight=torch.tensor([1., 2., 4., 8.]))
torch.return_types.histogramdd(
@ -10042,6 +10049,7 @@ Keyword args:
{check_invariants}
Example::
>>> compressed_indices = [0, 2, 4]
>>> plain_indices = [0, 1, 0, 1]
>>> values = [1, 2, 3, 4]
@ -10102,6 +10110,7 @@ Keyword args:
{check_invariants}
Example::
>>> crow_indices = [0, 2, 4]
>>> col_indices = [0, 1, 0, 1]
>>> values = [1, 2, 3, 4]
@ -10164,6 +10173,7 @@ Keyword args:
{check_invariants}
Example::
>>> ccol_indices = [0, 2, 4]
>>> row_indices = [0, 1, 0, 1]
>>> values = [1, 2, 3, 4]
@ -10228,6 +10238,7 @@ Keyword args:
{check_invariants}
Example::
>>> crow_indices = [0, 1, 2]
>>> col_indices = [0, 1]
>>> values = [[[1, 2], [3, 4]], [[5, 6], [7, 8]]]
@ -10294,6 +10305,7 @@ Keyword args:
{check_invariants}
Example::
>>> ccol_indices = [0, 1, 2]
>>> row_indices = [0, 1]
>>> values = [[[1, 2], [3, 4]], [[5, 6], [7, 8]]]

3
torch/autograd/function.py
View File

@ -64,6 +64,7 @@ class FunctionCtx:
See :ref:`extending-autograd` for more details on how to use this method.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_AUTOGRAD)
>>> class Func(Function):
>>> @staticmethod
@ -107,6 +108,7 @@ class FunctionCtx:
See :ref:`extending-autograd` for more details on how to use this method.
Example::
>>> # xdoctest: +SKIP
>>> class Func(torch.autograd.Function):
>>> @staticmethod
@ -234,6 +236,7 @@ class FunctionCtx:
prior to calling the :func:`backward` and :func:`jvp` methods.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_AUTOGRAD)
>>> class SimpleFunc(Function):
>>> @staticmethod

4
torch/distributed/device_mesh.py
View File

@ -417,6 +417,7 @@ else:
of mesh reduces across rows (0, 1, 2, 3) and (4, 5, 6, 7).
Example::
>>> # xdoctest: +SKIP("no rank")
>>> from torch.distributed.device_mesh import DeviceMesh
>>>
@ -700,6 +701,7 @@ else:
Calling mesh_3d["cp", "dp"] on rank 2, 3, 6, 7 returns a 2D submesh of DeviceMesh:([[2, 6], [3, 7]]).
Example::
>>> # xdoctest: +SKIP("no rank")
>>> from torch.distributed.device_mesh import DeviceMesh
>>>
@ -926,6 +928,7 @@ else:
Calling mesh_2d.get_local_rank(mesh_dim=1) on rank 3, 7 would return 3.
Example::
>>> # xdoctest: +SKIP("no rank")
>>> from torch.distributed.device_mesh import DeviceMesh
>>>
@ -1008,6 +1011,7 @@ else:
DeviceMesh: A :class:`DeviceMesh` object representing the device layout.
Example::
>>> # xdoctest: +SKIP("no rank")
>>> from torch.distributed.device_mesh import init_device_mesh
>>>

4
torch/futures/__init__.py
View File

@ -149,6 +149,7 @@ class Future(torch._C.Future, Generic[T], metaclass=_PyFutureMeta):
on those futures independently.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_FUTURES)
>>> def callback(fut):
... print(f"RPC return value is {fut.wait()}.")
@ -197,6 +198,7 @@ class Future(torch._C.Future, Generic[T], metaclass=_PyFutureMeta):
for handling completion/waiting on those futures independently.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_FUTURES)
>>> def callback(fut):
... print("This will run after the future has finished.")
@ -230,6 +232,7 @@ class Future(torch._C.Future, Generic[T], metaclass=_PyFutureMeta):
result (object): the result object of this ``Future``.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_FUTURES)
>>> import threading
>>> import time
@ -259,6 +262,7 @@ class Future(torch._C.Future, Generic[T], metaclass=_PyFutureMeta):
result (BaseException): the exception for this ``Future``.
Example::
>>> # xdoctest: +REQUIRES(env:TORCH_DOCTEST_FUTURES)
>>> fut = torch.futures.Future()
>>> fut.set_exception(ValueError("foo"))

4
torch/library.py
View File

@ -152,6 +152,7 @@ class Library:
name of the operator as inferred from the schema.
Example::
>>> my_lib = Library("mylib", "DEF")
>>> my_lib.define("sum(Tensor self) -> Tensor")
"""
@ -254,6 +255,7 @@ class Library:
the dispatch key that the library was created with.
Example::
>>> my_lib = Library("aten", "IMPL")
>>> my_lib._impl_with_aoti_compile("div.Tensor", "CPU")
"""
@ -316,6 +318,7 @@ class Library:
registered.
Example::
>>> my_lib = Library("aten", "IMPL")
>>> def div_cpu(self, other):
>>> return self * (1 / other)
@ -399,6 +402,7 @@ class Library:
to :attr:`fn` when calling. This should be used to create the appropriate keyset for redispatch calls.
Example::
>>> my_lib = Library("_", "IMPL")
>>> def fallback_kernel(op, *args, **kwargs):
>>> # Handle all autocast ops generically

11
torch/special/__init__.py
View File

@ -88,6 +88,7 @@ Keyword args:
out (Tensor, optional): the output tensor.
Example::
>>> a = torch.arange(-0.5, 1, 0.5)
>>> a
tensor([-0.5000, 0.0000, 0.5000])
@ -189,6 +190,7 @@ Keyword args:
{out}
Example::
>>> a = torch.tensor([1, 0.5])
>>> torch.special.polygamma(1, a)
tensor([1.64493, 4.9348])
@ -592,6 +594,7 @@ Keyword args:
{out}
Example::
>>> torch.special.i0e(torch.arange(5, dtype=torch.float32))
tensor([1.0000, 0.4658, 0.3085, 0.2430, 0.2070])
""".format(
@ -618,6 +621,7 @@ Keyword args:
{out}
Example::
>>> torch.special.i1(torch.arange(5, dtype=torch.float32))
tensor([0.0000, 0.5652, 1.5906, 3.9534, 9.7595])
""".format(
@ -645,6 +649,7 @@ Keyword args:
{out}
Example::
>>> torch.special.i1e(torch.arange(5, dtype=torch.float32))
tensor([0.0000, 0.2079, 0.2153, 0.1968, 0.1788])
""".format(
@ -671,6 +676,7 @@ Keyword args:
{out}
Example::
>>> torch.special.ndtr(torch.tensor([-3., -2, -1, 0, 1, 2, 3]))
tensor([0.0013, 0.0228, 0.1587, 0.5000, 0.8413, 0.9772, 0.9987])
""".format(
@ -700,6 +706,7 @@ Keyword args:
{out}
Example::
>>> torch.special.ndtri(torch.tensor([0, 0.25, 0.5, 0.75, 1]))
tensor([ -inf, -0.6745, 0.0000, 0.6745, inf])
""".format(
@ -726,6 +733,7 @@ Keyword args:
{out}
Example::
>>> torch.special.log_ndtr(torch.tensor([-3., -2, -1, 0, 1, 2, 3]))
tensor([-6.6077 -3.7832 -1.841 -0.6931 -0.1728 -0.023 -0.0014])
""".format(
@ -765,6 +773,7 @@ Keyword args:
{out}
Example::
>>> t = torch.randn(4)
>>> t
tensor([ 0.2252, -0.2948, 1.0267, -1.1566])
@ -838,6 +847,7 @@ Args:
is performed. This is useful for preventing data type overflows. Default: None.
Example::
>>> t = torch.ones(2, 2)
>>> torch.special.log_softmax(t, 0)
tensor([[-0.6931, -0.6931],
@ -868,6 +878,7 @@ Keyword args:
{out}
Example::
>>> x = torch.tensor([2., 4.])
>>> torch.special.zeta(x, 1)
tensor([1.6449, 1.0823])