Error code 12 (Cannot allocate memory)

[DaWe35]

python3 visualize.py --song ooo002.mp3 --resolution 512

VPS:

• 6 core
• 16 GB RAM
• 50 GB+ free SSD sapce
• Ubuntu 18.04

Generating frames

  0%|                                                                                                                                                                 | 0/14 [00:00<?, ?it/s]Traceback (most recent call last):
  File "visualize.py", line 382, in <module>
    output = model(noise_vector, class_vector, truncation)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/user/.local/lib/python3.6/site-packages/pytorch_pretrained_biggan/model.py", line 295, in forward
    z = self.generator(cond_vector, truncation)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/user/.local/lib/python3.6/site-packages/pytorch_pretrained_biggan/model.py", line 239, in forward
    z = layer(z, cond_vector, truncation)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/user/.local/lib/python3.6/site-packages/pytorch_pretrained_biggan/model.py", line 190, in forward
    x = self.conv_3(x)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 345, in forward
    return self.conv2d_forward(input, self.weight)
  File "/home/user/.local/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 342, in conv2d_forward
    self.padding, self.dilation, self.groups)
RuntimeError: [enforce fail at CPUAllocator.cpp:64] . DefaultCPUAllocator: can't allocate memory: you tried to allocate 4026531840 bytes. Error code 12 (Cannot allocate memory)

  0%|                                                                                                                                                                 | 0/14 [01:36<?, ?it/s]

[mhsharifi96]

same problem ! :/
when try run occur error can't allocate memory

what can I solve it?

[didinium]

have the same problem over here

[illtellyoulater]

same thing here, anyone manged to fix this?

EDIT: Ah ok, just saw there's a PR that could fix it.

[didinium]

I didn't have enough ram. You could either get more ram or modify the code so the arrays aren't stored in ram

…

On Sun, Mar 5, 2023 at 9:54 AM illtellyoulater ***@***.***> wrote: same thing here, anyone manged to fix this? — Reply to this email directly, view it on GitHub <#3 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AB6UVWBHLMTBVZOLTWIMLNLW2SSKJANCNFSM4J3CHSGQ> . You are receiving this because you commented.Message ID: ***@***.***>

[illtellyoulater]

I could make it work using this updated fork which includes this and other fixes, in addition to a couple more features: https://github.com/zlattala/deep-music-visualizer

[DaWe35]

Thanks @illtellyoulater, it works fine for me, except is uses 0% of my GPU. How can I make it run on GPU instead of CPU?

[illtellyoulater]

@DaWe35

If you are using the code from the forked repository then you should open an issue about that on the forked repository.

That said, I am not observing the issue you are reporting, but usually when you see the CPU being used in place of the GPU it means you installed a wrong version of torch. To make sure everything works as expected please make sure to follow the forked repository package requirements and installing instructions.

[DaWe35]

You're right, but I don't think this issue is related to the new fork. With conda, I was able to use my GPU, so yes, it was something with the version of python/torch/env.
My next question is if there is a 1080p/4k pretrained model, and if not, how do I train it? 😄