（转）RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.DoubleTensor) should be

最新推荐文章于 2024-09-19 18:05:46 发布

JY丫丫

最新推荐文章于 2024-09-19 18:05:46 发布

阅读量1.4k

点赞数

分类专栏： pytorch

原文链接：https://blog.csdn.net/jizhidexiaoming/article/details/82502280

版权

pytorch 专栏收录该内容

16 篇文章 2 订阅

订阅专栏

出错位置：

train_label_batch = torch.from_numpy(train_label_batch)

解决办法：将数据类型转换成FloatTensor即可，如下，加一行代码Tensor.type(torch.FloatTensor)


 
 
   
   
    
    
   
   
   
   
    
    
     
     train_label_batch = torch.from_numpy(train_label_batch)
    
    
   
   

   
   
    
    
   
   
   
   
    
    
     
     train_label_batch = train_label_batch.type(torch.FloatTensor)  
     
     # 转Float
    
    
   
   

   
   
    
    
   
   
   
   
    
    
     
     train_label_batch = train_label_batch.cuda()  
     
     # 转cuda

转自
RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.DoubleTensor) should be the - 一千零一夜的博客 - CSDN博客
https://blog.csdn.net/jizhidexiaoming/article/details/82502280