CommissarMa/MCNN-pytorch

Issues

• 最好的结果是多少

  #3 opened 5 years ago by ZhongyuanW
  8

• How can we evaluate the results with the official MSE loss?

  #13 opened 2 years ago by freshn
  2

• 请问如何得到具体的预测的人数呢？

  #9 opened 4 years ago by bigship
  3

• 数据集

  #15 opened a year ago by ihyyy
  0

• FileNotFoundError about IMG_XXX.npy file

  #6 opened a year ago by ft-z
  2

• 您好，可以提供下权重文件吗，我发现我在我的显卡上batchsize调为1以上的数字会报错，只能batchsize为1训练，但是很慢。

  #11 opened 3 years ago by wong1998
  1

• FileNotFoundError about IMG_XXX.npy file

  #14 opened a year ago by BlouseDong
  0

• 您好，您的model大概跑了多少轮，有最好的结果。

  #4 opened 5 years ago by wbchief
  2

• Dataset link failed

  #12 opened 3 years ago by handsn
  0

• Label for custom dataset

  #10 opened 4 years ago by enesdoruk
  0

• 您好，想问您如何解决visdom连接超时的问题？

  #5 opened 4 years ago by Cyprus-hy
  0

• Hi! I have a question, these are so many epochs to run so that i can't work it out in one time. If i need to save the model, should i save the min_mae epoch' model or the newest(break point) epoch.And,in the next time,which one(best one?or break one?) should i load to continue train.Hoping to get your help,Thx

  #8 opened 5 years ago by simonchance
  0

• Tried to allocate 18.07 GiB

  #7 opened 5 years ago by anhminh3105
  8

• 关于训练得问题

  #2 opened 6 years ago by EillotY
  2

• 在这个train.py下，我没有找到需要链接url，但是源码运行的时候，会报connection error的错误，并且不能够顺利的运行train.py;还存在的一个问题是您的my_dataloader.py里面的np.load其实是failed to interpret file as pikle.希望您能解决我的问题

  #1 opened 6 years ago by EillotY
  5