修复atleast函数中，”输入为tensor的list，输出不是tensor的list“的bug #73102

Qin-sx · 2025-06-04T15:34:49Z

PR Category

User Experience

PR Types

Bug fixes

Description

当前问题
当输入为tensor的list时，输出为tensor

解决方案
当输入为tensor的list时，进行判断，现在输出也为tensor的list

modified: python/paddle/tensor/manipulation.py modified: test/legacy_test/test_atleast_xd.py

paddle-bot · 2025-06-04T15:34:56Z

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.

codecov-commenter · 2025-06-04T19:04:34Z

Codecov Report

Attention: Patch coverage is 0% with 9 lines in your changes missing coverage. Please review.

Please upload report for BASE (develop@917ecbd). Learn more about missing BASE report.

Files with missing lines	Patch %	Lines
python/paddle/tensor/manipulation.py	0.00%	9 Missing ⚠️

❌ Your patch status has failed because the patch coverage (0.00%) is below the target coverage (90.00%). You can increase the patch coverage or adjust the target coverage.

Additional details and impacted files

@@            Coverage Diff             @@
##             develop   #73102   +/-   ##
==========================================
  Coverage           ?    0.00%           
==========================================
  Files              ?        1           
  Lines              ?        9           
  Branches           ?        0           
==========================================
  Hits               ?        0           
  Misses             ?        9           
  Partials           ?        0

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

Qin-sx · 2025-06-05T01:58:58Z

CI: https://xly.bce.baidu.com/paddlepaddle/paddle/newipipe/detail/12905852/job/29455975
可否麻烦帮忙看一下，我在代码中加入print，测试中的代码应该是覆盖了新增代码的

    if len(inputs) == 1 and isinstance(inputs[0], list):
        if all(
            isinstance(
                item,
                (
                    paddle.Tensor,
                    paddle.base.framework.Variable,
                    paddle.base.libpaddle.pir.Value,
                ),
            )
            for item in inputs[0]
        ):
            inputs = inputs[0]
            print("in atleast_1d inputs:", inputs)

in atleast_1d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]
in atleast_2d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]
in atleast_3d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]

Qin-sx · 2025-06-05T14:26:00Z

PR-CI-Coverage
这个CI好像无法重新执行，“该流水线禁止重新执行！”

zhwesky2010 · 2025-06-06T06:59:08Z

python/paddle/tensor/manipulation.py

@@ -5277,6 +5277,20 @@ def atleast_1d(*inputs, name=None):
            [123]), Tensor(shape=[1, 1], dtype=float32, place=Place(cpu), stop_gradient=True,
            [[1.23000002]])]
    """
+    if len(inputs) == 1 and isinstance(inputs[0], list):


if len(inputs) == 1 and isinstance(inputs[0], (list, tuple)): inputs = inputs[0]

不用判断list里是什么

您好，请问是指不用判断list里面是不是paddle.tensor吗？因为好像原本的测试中有list包含list的情况，所以不判断的话测试无法通过。

具体是哪个case呢？

list包含list情况按这么写，应该也没问题吧？

if len(inputs) == 1 and isinstance(inputs[0], (list, tuple)): inputs = inputs[0]

import numpy as np import paddle import torch # inputs = ([[[137]]],) inputs = [[[[137]]]] print("Inputs:", inputs) print("type Inputs:", type(inputs)) print("type Inputs[0]:", type(inputs[0])) result = paddle.atleast_1d(*inputs) print("Result:", type(result)) print("Result:", result) # result_pt = torch.atleast_1d(*inputs) # print("Result_pt:", type(result_pt)) # print("Result_pt:", result_pt)

Inputs: [[[[137]]]] type Inputs: <class 'list'> type Inputs[0]: <class 'list'> len(out) == 1 Result: <class 'paddle.Tensor'> Result: Tensor(shape=[1, 1, 1], dtype=int64, place=Place(gpu:0), stop_gradient=True, [[[137]]])

对于上面这个例子，paddle会将[[[[137]]]]的list改为[[[137]]]的tensor。pytorch应该不支持这个输入。
如果按照下面的方式写的话，paddle会将[[[[137]]]]的list改为[[137]]的tensor。会比之前少一个维度。

if len(inputs) == 1 and isinstance(inputs[0], (list, tuple)): inputs = inputs[0]

Inputs: [[[[137]]]] type Inputs: <class 'list'> type Inputs[0]: <class 'list'> len(out) == 1 Result: <class 'paddle.Tensor'> Result: Tensor(shape=[1, 1], dtype=int64, place=Place(gpu:0), stop_gradient=True, [[137]])

zhwesky2010 · 2025-06-11T03:59:28Z

python/paddle/tensor/manipulation.py

@@ -5277,6 +5277,20 @@ def atleast_1d(*inputs, name=None):
            [123]), Tensor(shape=[1, 1], dtype=float32, place=Place(cpu), stop_gradient=True,
            [[1.23000002]])]
    """
+    if len(inputs) == 1 and isinstance(inputs[0], list):
+        if all(


这里改成any，规则是：如果有任意一个Tensor，就可以按list来分解，如果全部为子list，就不用分解。

嗯，收到，已修改

zhwesky2010 · 2025-06-11T04:00:08Z

python/paddle/tensor/manipulation.py

@@ -5277,6 +5277,20 @@ def atleast_1d(*inputs, name=None):
            [123]), Tensor(shape=[1, 1], dtype=float32, place=Place(cpu), stop_gradient=True,
            [[1.23000002]])]
    """
+    if len(inputs) == 1 and isinstance(inputs[0], list):


这里改成 isinstance(inputs[0], (list, tuple)):

覆盖输入一个tuple(Tensor)的情况

嗯，收到，已修改

modified: python/paddle/tensor/manipulation.py

Qin-sx · 2025-06-11T15:36:33Z

CI: https://xly.bce.baidu.com/paddlepaddle/paddle/newipipe/detail/12905852/job/29455975 可否麻烦帮忙看一下，我在代码中加入print，测试中的代码应该是覆盖了新增代码的

    if len(inputs) == 1 and isinstance(inputs[0], list):
        if all(
            isinstance(
                item,
                (
                    paddle.Tensor,
                    paddle.base.framework.Variable,
                    paddle.base.libpaddle.pir.Value,
                ),
            )
            for item in inputs[0]
        ):
            inputs = inputs[0]
            print("in atleast_1d inputs:", inputs)

in atleast_1d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]
in atleast_2d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]
in atleast_3d inputs: [Tensor(shape=[], dtype=int32, place=Place(cpu), stop_gradient=True,
       123), Tensor(shape=[3], dtype=float32, place=Place(cpu), stop_gradient=True,
       [1., 2., 3.]), Tensor(shape=[2, 2], dtype=int64, place=Place(cpu), stop_gradient=True,
       [[1, 2],
        [3, 4]])]

CI还是有同样的问题，测试的代码应该可以覆盖

zhwesky2010 · 2025-06-12T03:52:50Z

test/legacy_test/test_atleast_xd.py

+class TestAtleastWithTensorList(unittest.TestCase):
+    """Test when input is a list of paddle tensors"""
+
+    def test_tensor_list_input(self):


再加几个case，分别是 list包含list、Tensor list、tuple包含tuple、Tensor tuple，目前只有第二种Tensor list

嗯，收到，已添加

modified: test/legacy_test/test_atleast_xd.py

zhwesky2010

LGTM

zhwesky2010 · 2025-06-16T03:18:56Z

@Qin-sx 看下这个CE-Framework为何失败，是否有不兼容问题

Qin-sx · 2025-06-17T00:37:35Z

@Qin-sx 看下这个CE-Framework为何失败，是否有不兼容问题

嗯，我重新执行通过了，应该是偶发的。我记得之前提交的版本应该都是通过的。

modified for input of list of tensor in atleast func

4506195

modified: python/paddle/tensor/manipulation.py modified: test/legacy_test/test_atleast_xd.py

paddle-bot bot added the contributor External developers label Jun 4, 2025

zhwesky2010 reviewed Jun 6, 2025

View reviewed changes

Merge branch 'develop' into atleast_xd

30fe19d

zhwesky2010 reviewed Jun 11, 2025

View reviewed changes

modified tuple and any in condition

f8191b6

modified: python/paddle/tensor/manipulation.py

zhwesky2010 reviewed Jun 12, 2025

View reviewed changes

Qin-sx added 2 commits June 12, 2025 14:33

added a test for tuple of tensor

dc0e405

modified: test/legacy_test/test_atleast_xd.py

added tests of nested tuple and list

14eb29f

modified: test/legacy_test/test_atleast_xd.py

zhwesky2010 approved these changes Jun 13, 2025

View reviewed changes

zhwesky2010 merged commit b6b5d60 into PaddlePaddle:develop Jun 23, 2025
50 of 51 checks passed

修复atleast函数中，”输入为tensor的list，输出不是tensor的list“的bug #73102

修复atleast函数中，”输入为tensor的list，输出不是tensor的list“的bug #73102

Uh oh!

Conversation

Qin-sx commented Jun 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

PR Category

PR Types

Description

Uh oh!

paddle-bot bot commented Jun 4, 2025

Uh oh!

codecov-commenter commented Jun 4, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Qin-sx commented Jun 5, 2025

Uh oh!

Qin-sx commented Jun 5, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zhwesky2010 Jun 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Qin-sx commented Jun 11, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

zhwesky2010 left a comment

Choose a reason for hiding this comment

Uh oh!

zhwesky2010 commented Jun 16, 2025

Uh oh!

Qin-sx commented Jun 17, 2025

Uh oh!

Uh oh!

Uh oh!

Qin-sx commented Jun 4, 2025 •

edited

Loading

codecov-commenter commented Jun 4, 2025 •

edited

Loading

zhwesky2010 Jun 9, 2025 •

edited

Loading