search

DLHacks / mocogan

A pytorch implemention of MoCoGAN
102 stars 21 forks source link

ValueError: Range cannot be empty (low >= high) unless no samples are taken #1

Open mengyaaa opened 5 years ago

mengyaaa commented 5 years ago

/home/lab226/Desktop/mocogan-hong/models.py:147: UserWarning: nn.init.xavier_uniform is now deprecated in favor of nn.init.xavieruniform. init.xavieruniform(params) /home/lab226/Desktop/mocogan-hong/models.py:157: UserWarning: nn.init.constant is now deprecated in favor of nn.init.constant. init.constant(params, 0) !!!!!!!!!!!!!!!!!! -12 Traceback (most recent call last): File "train.py", line 211, in real_videos = random_choice() File "train.py", line 79, in random_choice video = torch.Tensor(trim(video)) File "train.py", line 63, in trim start = np.random.randint(0, video.shape[1] - (T+1)) File "mtrand.pyx", line 992, in mtrand.RandomState.randint ValueError: Range cannot be empty (low >= high) unless no samples are taken can any one help me?thx~

nidhal1231 commented 5 years ago

@mengyaaa Have you found a solution. I have the same issue. Thankyou for your help

cjiang2000 commented 5 years ago

Bump to this thread. I also have the same issue.

artikeshari commented 4 years ago

Length of the video might be zero. Check the size of video.