qq_40146935的博客

元素去重 import itertools label = [k for k,g in itertools.groupby(list('AAAABBBCCDAABBB'))] print(label)

import numpy as np indices = "hongnngn" target = np.zeros(shape=(5,),dtype=np.long) target[:6]= indices print(target) ValueError: invalid literal for int() with base 10: 'hongnngn' indices = [1,2,4,5,7] target = np.zeros(shape=(10,),dtype=np.long) tar.

问题 可视化网络时，出现下面错误 RuntimeError: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same 解决方法 原因：可视化网络使用的时cpu 将网络转成cuda net = CRNN(num_classes=1000) 修改后 net = CRNN(num_classes=1000) net.cuda() ...