PyTorch：The “freeze_support()” line can be omitted if the program is not going to be frozen

小鱼儿 2022-09-05 12:48 127阅读 0赞

在windows上运行pytorch时，稍不注意就会遇到freeze\_support()的错误。解决这种错误只要把代码放到if name == “main”: 中运行就可以了。就忍不住来深究下这个问题。

这是一个关于windows上多进程实现的恩特。在windows上，子进程会自动import启动它的这个文件，而在import的时候是会自动执行这些语句的。如果不加\_\_main\_\_限制的化，就会无限递归创建子进程，进而报错。于是import的时候使用 name == “main” 保护起来就可以了。

出现错误：

Traceback (most recent call last):  
File "<string>", line 1, in <module>  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 105, in spawn\_main  
exitcode = \_main(fd)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 114, in \_main  
prepare(preparation\_data)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 225, in prepare  
\_fixup\_main\_from\_path(data\['init\_main\_from\_path'\])  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 277, in \_fixup\_main\_from\_path  
run\_name="\_\_mp\_main\_\_")  
File "D:\\opt\\anaconda3\\lib\\runpy.py", line 263, in run\_path  
pkg\_name=pkg\_name, script\_name=fname)  
File "D:\\opt\\anaconda3\\lib\\runpy.py", line 96, in \_run\_module\_code  
mod\_name, mod\_spec, pkg\_name, script\_name)  
File "D:\\opt\\anaconda3\\lib\\runpy.py", line 85, in \_run\_code  
exec(code, run\_globals)  
File "D:\\ML\_Study\\PyTorch-garbage-classify\\garbage-classification-using-pytorch.py", line 59, in <module>  
image,label = next(iter(train\_loader))  
File "D:\\opt\\anaconda3\\lib\\site-packages\\torch\\utils\\data\\dataloader.py", line 819, in \_\_iter\_\_  
return \_DataLoaderIter(self)  
File "D:\\opt\\anaconda3\\lib\\site-packages\\torch\\utils\\data\\dataloader.py", line 560, in \_\_init\_\_  
w.start()  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\process.py", line 105, in start  
self.\_popen = self.\_Popen(self)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\context.py", line 223, in \_Popen  
return \_default\_context.get\_context().Process.\_Popen(process\_obj)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\context.py", line 322, in \_Popen  
return Popen(process\_obj)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\popen\_spawn\_win32.py", line 33, in \_\_init\_\_  
prep\_data = spawn.get\_preparation\_data(process\_obj.\_name)  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 143, in get\_preparation\_data  
\_check\_not\_importing\_main()  
File "D:\\opt\\anaconda3\\lib\\multiprocessing\\spawn.py", line 136, in \_check\_not\_importing\_main  
is not going to be frozen to produce an executable.''')  
RuntimeError:  
An attempt has been made to start a new process before the  
current process has finished its bootstrapping phase.

This probably means that you are not using fork to start your  
child processes and you have forgotten to use the proper idiom  
in the main module:

if \_\_name\_\_ == '\_\_main\_\_':  
freeze\_support()  
...

The "freeze\_support()" line can be omitted if the program  
is not going to be frozen to produce an executable.  
  
解决方案：

if \_\_name\_\_ == '\_\_main\_\_':  
\# 代码  
1  
2  
通过上述方式实验完美解决，这里说下我的实验版本

In \[1\]: import torch

In \[2\]: import torchvision

In \[3\]: torch.version  
Out\[3\]: ‘1.0.1’

In \[4\]: torchvision.version  
Out\[4\]: ‘0.2.2’  
————————————————  
版权声明：本文为CSDN博主「aiwen2100」的原创文章，遵循CC 4.0 BY-SA版权协议，转载请附上原文出处链接及本声明。  
原文链接：https://blog.csdn.net/shenfuli/article/details/103969964