运行Dynamic-DeepJSCC报错及解决

《Deep Joint Source-Channel Coding for Wireless Image Transmission with Adaptive Rate Control》论文代码复现

代码链接：GitHub - mingyuyng/Dynamic_JSCC: Codes for "Deep Joint Source-Channel Coding for Wireless Image Transmission with Adaptive Rate Control", ICASSP 2022

主要出现以下两个问题：

1.AttributeError: module 'torch._C' has no attribute '_cuda_setDevice'

报错原因：我的pytorch是CPU版本，命令行中的--gpu_ids 0是指使用GPU运行程序，所以出现以上报错

解决：①将命令行中的--gpu_ids 0改成--gpu_ids -1

           ②修改base_options.py中‘gpu_ids’的默认值为-1

2.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.

报错原因：多进程的原因

解决：①在运行epoch前，加上if __name__=='__main__'

           ②不使用多线程，num_workers参数设置为0