[xdoctest] reformat example code with google style in No.324, 333, 336, 353 #57565

longranger2 · 2023-09-20T15:34:59Z

PR types

Others

PR changes

Others

Description

修改如下文件的示例代码为新的格式，并通过 xdoctest 检查：

paddle/fluid/pybind/parallel_executor.cc
python/paddle/base/dygraph/base.py
python/paddle/base/executor.py
paddle/fluid/pybind/eager_method.cc

@sunzhongkai588 @SigureMo @megemini

[xdoctest] 分批次修改已有代码的示例 #55629

paddle-bot · 2023-09-20T15:35:06Z

你的PR提交成功，感谢你对开源项目的贡献!
请关注后续CI自动化测试结果，详情请参考Paddle-CI手册。
Your PR has been submitted. Thanks for your contribution!
Please wait for the result of CI firstly. See Paddle CI Manual for details.

SigureMo

CI 里的问题也需要看一下

SigureMo · 2023-09-21T01:38:24Z

paddle/fluid/pybind/eager_method.cc

-        data = paddle.to_tensor(data)
-        x = linear(data)
-        print(x.numpy())
+        >>> paddle.disable_static()


默认不是动态图么，为什么需要这个？是使用的 convert_doctest 太旧了还是？

SigureMo · 2023-09-21T01:40:29Z

paddle/fluid/pybind/eager_method.cc

-        y = x[1]
-        print(y.is_contiguous())
+        >>> x = paddle.to_tensor([1, 2, 3])
+        >>> y = x[1]


怎么下面那个 is_contiguous 直接删了？这不就是 is_contiguous 的 docstring 么？

SigureMo · 2023-09-21T01:41:13Z

paddle/fluid/pybind/parallel_executor.cc

-                        build_strategy = static.BuildStrategy()
-                        build_strategy.fuse_adamw = True
-                     )DOC")
+                        >>> import paddle


上面需要空一行

SigureMo · 2023-09-21T01:42:03Z

python/paddle/base/dygraph/base.py

-            import paddle
-            print(paddle.in_dynamic_mode())  # True, dynamic mode is turn ON by default since paddle 2.0.0
+            >>> import paddle
+            >>> print(paddle.in_dynamic_mode())  # True, dynamic mode is turn ON by default since paddle 2.0.0


后面注释里的 True 可以删掉了

SigureMo · 2023-09-21T01:42:12Z

python/paddle/base/dygraph/base.py

-            paddle.enable_static()
-            print(paddle.in_dynamic_mode())  # False, Now we are in static graph mode
+            >>> paddle.enable_static()
+            >>> print(paddle.in_dynamic_mode())  # False, Now we are in static graph mode


SigureMo · 2023-09-21T01:44:07Z

python/paddle/base/dygraph/base.py

+            >>> print(tmp.gradient() is None)
+            >>> print(l0.weight.gradient() is None)
+            True
+            False


输出位置不对吧，489 不是 487 的输出么

SigureMo · 2023-09-21T01:44:40Z

python/paddle/base/dygraph/base.py

+            >>> print(test_dygraph_grad(create_graph=False))
+            >>> print(test_dygraph_grad(create_graph=True))
+            [2.]
+            [4.]


同上，输出要跟随相应的语句

SigureMo · 2023-09-21T01:44:58Z

python/paddle/base/dygraph/base.py

+            ...     c = np.array([2+1j, 2])
+            ...     z = base.dygraph.to_variable(c)
+            ...     z.numpy()
+            ...     z.dtype


这里用 print 吧，不然会调用 __repr__ 而得到 VarType 而不是 paddle.float32 这样的字符串

SigureMo · 2023-09-21T01:45:10Z

python/paddle/base/dygraph/base.py

+            ...     y.shape
+            ...
+            ...     y = base.dygraph.to_variable(((0.1, 1.2), (2.2, 3.1), (4.9, 5.2)), dtype='int32')
+            ...     y.shape


其他几个也都加下 print 吧

>>> x 这种比较合适不写 print

... x 就很奇怪了

SigureMo · 2023-09-21T01:45:50Z

python/paddle/base/executor.py

-                #  [-0.24635398 -0.13003758]
-                #  [-0.49232286 -0.25939852]
-                #  [-0.44514108 -0.2345845 ]]
+                >>> # required: gpu


没改成 doctest directive

@megemini 顺师傅有时间在 bad statement 里加一个这个？好像现在只需要加一个正则规则就行了？

PR： #57578 🤗

longranger2 · 2023-09-21T09:25:55Z

done

megemini · 2023-09-21T09:35:10Z

paddle/fluid/pybind/eager_method.cc

+        >>> import paddle
+
+        >>> x = paddle.to_tensor(1.0, stop_gradient=False)
+        >>> clone_x = x.clone()


要加一下 clone_x.retain_grads()，不然 clone_x.stop_gradient 是 False，而 clone_x.grad 是 None，作为示例没有意义～

megemini · 2023-09-21T09:38:03Z

paddle/fluid/pybind/eager_method.cc


-      print(y.grad) # [1., 1., 1.]
+        >>> print(y.grad) # [1., 1., 1.]


要有输出～

megemini · 2023-09-21T09:40:55Z

paddle/fluid/pybind/eager_method.cc

+        >>> print(x.grad)
+        Tensor(shape=[1], dtype=float32, place=Place(cpu), stop_gradient=False, [20.])
+
+        >>> print(detach_x.grad)


原来的注释 'stop_gradient=True' by default 是说明性文字，可以保留～

megemini · 2023-09-21T09:51:36Z

paddle/fluid/pybind/eager_method.cc

+        >>> print(underline_x)
+         - place: Place(cpu)
+         - shape: [1]
+         - layout: NCHW
+         - dtype: 5
+         - data: [1]


# a Dense Tensor info 注释保留

这里的结果跟 aistudio 的不一样，麻烦确认一下：

- place: Place(cpu) - shape: [1] - layout: NCHW - dtype: float32 - data: [1]

megemini · 2023-09-21T09:52:17Z

paddle/fluid/pybind/eager_method.cc


-        x = paddle.to_tensor([1, 2, 3])
-        print(x.data_ptr())
+        >>> x = paddle.to_tensor([1, 2, 3])


这里的 print(x.data_ptr()) 不要省，可以把输出 skip 掉～

megemini · 2023-09-21T09:57:46Z

paddle/fluid/pybind/parallel_executor.cc

+                                ...     import paddle
+                                ...     import paddle.static as static
+                                ...
+                                ...     paddle.enable_static()
+                                ...
+                                ...     build_strategy = static.BuildStrategy()
+                                ...     build_strategy.build_cinn_pass = True


这里要用 >>> ，另外，空格留的多了～

是不是用的 convert-doctest 转的？工具在一些错误缩进的时候无法处理 ... ...

是的，使用convert-doctest 转的

megemini · 2023-09-21T09:59:35Z

paddle/fluid/pybind/parallel_executor.cc

+                                ...     import paddle
+                                ...     import paddle.static as static
+                                ...
+                                ...     paddle.enable_static()
+                                ...
+                                ...     build_strategy = static.BuildStrategy()
+                                ...     build_strategy.fuse_broadcast_ops = True


megemini · 2023-09-21T10:08:25Z

python/paddle/base/dygraph/base.py

+            >>> print(test_dygraph_grad(None))
+            7.
+
+            >>> # dy1 = [1], dy2 = [4]
+            >>> print(test_dygraph_grad([None, grad_value]))
+            [16.]
+
+            >>> # dy1 = [4], dy2 = [1]
+            >>> print(test_dygraph_grad([grad_value, None]))
+            [19.]
+
+            >>> # dy1 = [3], dy2 = [4]
+            >>> grad_y1 = paddle.to_tensor(3.0)
+            >>> print(test_dygraph_grad([grad_y1, grad_value]))
+            [24.]


aistudio 上面测试，这里的输出只有数字，不是 list，请确认一下～

megemini · 2023-09-21T10:10:28Z

python/paddle/base/dygraph/base.py

+            array(1, dtype=float32)
+            array(-1, dtype=float32)
+            array([2.+1.j, 2.+0.j])
+            complex128
+            [3, 2]
+            [3, 2]


aistudio 上用 develop 的 paddle，输出是：

1.0 -1.0 [2.+1.j 2.+0.j] paddle.complex128 [3, 2] [3, 2]

请确认一下～

megemini · 2023-09-21T10:27:53Z

python/paddle/base/executor.py

+            >>> import paddle.base as base
+            >>> place = base.CPUPlace()
+            >>> exe = base.executor(place)
+            >>> data = np.array(size=(100, 200, 300))
+            >>> np_outs = map(lambda x: base.executor._as_lodtensor(x, place), data)


需要 import numpy

base.Executor

np.array 不对～

longranger2 · 2023-09-26T14:14:44Z

@megemini done

megemini · 2023-09-27T05:28:37Z

paddle/fluid/pybind/eager_method.cc

+        >>> y = x[1]
+        >>> y = y.contiguous()
+        >>> print(y)
+        ensor(shape=[], dtype=int64, place=Place(cpu), stop_gradient=True, 2)


漏了一个 T

megemini · 2023-09-27T06:45:55Z

python/paddle/base/dygraph/base.py

+            >>> print(test_dygraph_grad(None))
+            7.
+
+            >>> # dy1 = [1], dy2 = [4]
+            >>> print(test_dygraph_grad([None, grad_value]))
+            16.
+
+            >>> # dy1 = [4], dy2 = [1]
+            >>> print(test_dygraph_grad([grad_value, None]))
+            19.
+
+            >>> # dy1 = [3], dy2 = [4]
+            >>> grad_y1 = paddle.to_tensor(3.0)
+            >>> print(test_dygraph_grad([grad_y1, grad_value]))
+            24.


@SigureMo 这里是我这边 patch float 精度的时候有问题，对于字符串开头就是数字的情况，之前没有考虑到，已经提交 PR：#57806 🤕

#57806 已合入，这个 PR 可以 merge 一下 develop 了

paddle-ci-bot · 2023-10-05T03:20:09Z

Sorry to inform you that 592d353's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually.

SigureMo

LGTMeow 🐾

@megemini 顺师傅可以再 review 一下～

megemini

LGTM ~ 辛苦 🫡

…6, 353 (PaddlePaddle#57565) * fix sample codes * fix bug * fix bug * fix bug * Update eager_method.cc

fix sample codes

bfd2071

paddle-bot bot added the contributor External developers label Sep 20, 2023

SigureMo reviewed Sep 21, 2023

View reviewed changes

luotao1 added the HappyOpenSource Pro 进阶版快乐开源活动，更具挑战性的任务 label Sep 21, 2023

luotao1 assigned luotao1, SigureMo and sunzhongkai588 Sep 21, 2023

fix bug

c58ac97

megemini reviewed Sep 21, 2023

View reviewed changes

luotao1 mentioned this pull request Sep 26, 2023

[xdoctest] 分批次修改已有代码的示例 #55629

Closed

fix bug

5d020f8

longranger2 requested review from megemini and SigureMo September 26, 2023 14:15

fix bug

a3ff95a

megemini reviewed Sep 27, 2023

View reviewed changes

Update eager_method.cc

592d353

Merge branch 'PaddlePaddle:develop' into xdoctest8

322af91

longranger2 requested a review from megemini October 10, 2023 10:52

SigureMo approved these changes Oct 10, 2023

View reviewed changes

megemini approved these changes Oct 11, 2023

View reviewed changes

luotao1 merged commit a392e4d into PaddlePaddle:develop Oct 11, 2023

Frida-a pushed a commit to Frida-a/Paddle that referenced this pull request Oct 14, 2023

[xdoctest] reformat example code with google style in No.324, 333, 33…

fd846d2

…6, 353 (PaddlePaddle#57565) * fix sample codes * fix bug * fix bug * fix bug * Update eager_method.cc

jiahy0825 pushed a commit to jiahy0825/Paddle that referenced this pull request Oct 16, 2023

[xdoctest] reformat example code with google style in No.324, 333, 33…

d0a763b

…6, 353 (PaddlePaddle#57565) * fix sample codes * fix bug * fix bug * fix bug * Update eager_method.cc

longranger2 deleted the xdoctest8 branch October 29, 2023 03:00

danleifeng pushed a commit to danleifeng/Paddle that referenced this pull request Nov 14, 2023

[xdoctest] reformat example code with google style in No.324, 333, 33…

6c43c1c

…6, 353 (PaddlePaddle#57565) * fix sample codes * fix bug * fix bug * fix bug * Update eager_method.cc

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[xdoctest] reformat example code with google style in No.324, 333, 336, 353 #57565

[xdoctest] reformat example code with google style in No.324, 333, 336, 353 #57565

longranger2 commented Sep 20, 2023

paddle-bot bot commented Sep 20, 2023

SigureMo left a comment

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

SigureMo Sep 21, 2023

megemini Sep 21, 2023

longranger2 commented Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

longranger2 Sep 26, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

megemini Sep 21, 2023

longranger2 commented Sep 26, 2023

megemini Sep 27, 2023

megemini Sep 27, 2023

SigureMo Oct 10, 2023

longranger2 Oct 10, 2023

paddle-ci-bot bot commented Oct 5, 2023

SigureMo left a comment

megemini left a comment


		print(y.grad) # [1., 1., 1.]
		>>> print(y.grad) # [1., 1., 1.]

[xdoctest] reformat example code with google style in No.324, 333, 336, 353 #57565

[xdoctest] reformat example code with google style in No.324, 333, 336, 353 #57565

Conversation

longranger2 commented Sep 20, 2023

PR types

PR changes

Description

paddle-bot bot commented Sep 20, 2023

SigureMo left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

longranger2 commented Sep 21, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

longranger2 commented Sep 26, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

paddle-ci-bot bot commented Oct 5, 2023

SigureMo left a comment

Choose a reason for hiding this comment

megemini left a comment

Choose a reason for hiding this comment