bug描述 Describe the Bug

我使用的环境是aistudio的V100 16GB环境，在创建环境时配置如下：

• paddlepaddle 2.3.2
• python 3.7

当我尝试使用自定义Layer推理时，出现报错如下：

aistudio@jupyter-368487-4527015:~$ python test.py 
W0912 14:11:56.456024  1107 gpu_resources.cc:61] Please NOTE: device: 0, GPU Compute Capability: 7.0, Driver API Version: 11.2, Runtime API Version: 11.2
W0912 14:11:56.460989  1107 gpu_resources.cc:91] device: 0, cuDNN Version: 8.2.

--------------------------------------
C++ Traceback (most recent call last):
--------------------------------------
No stack trace in paddle, may be caused by external reasons.

----------------------
Error Message Summary:
----------------------
FatalError: `Segmentation fault` is detected by the operating system.
  [TimeInfo: *** Aborted at 1662963117 (unix time) try "date -d @1662963117" if you are using GNU date ***]
  [SignalInfo: *** SIGSEGV (@0x0) received by PID 1107 (TID 0x7f4579ad2700) from PID 0 ***]

段错误 (核心已转储)
aistudio@jupyter-368487-4527015:~$

我的完成代码在test.py中，test.py完整内容如下：

import paddle
from paddle import nn
import paddle.nn.functional as F

class Net(nn.Layer):
    def __init__(self):
        super(Net, self).__init__()
        self.head = nn.Sequential(
            nn.Conv2D(in_channels=1, out_channels=64, kernel_size=3, stride=1, padding=1, bias_attr=False),
            nn.BatchNorm2D(64)
        )

    def forward(self, x):  # output shape:
        x = F.relu(self.head(x))  # 1,64,40,40
        return x

input=paddle.ones([1,1,112,112])
net=Net()
out=net(input)
print(out.shape)

我将仅包含test.py文件的项目公开，直接使用命令： python test.py 即可执行复现出上述bug

我发现了一篇高质量的实训项目，使用免费算力即可一键运行，还能额外获取8小时免费GPU运行时长，快来Fork一下体验吧。
FatalError： https://aistudio.baidu.com/aistudio/projectdetail/4529590?sUid=368487&shared=1&ts=1662963462896

其他补充信息 Additional Supplementary Information

No response