Creating video issue

[DavidRees87]

When using the default set up I am able to create a video without any problem. The issue I am having is when locating the .pkl file in my GDrive. The GDrive is connected, the training is being saved there and I am able to resume training with no problem. However when I use the path to the .pkl file neither images or video can be created. Do you have any ideas why this may be happening and any solutions?

Thank you for any help you can give, I greatly appreciate it.

[woctezuma]

Maybe try to copy the .pkl file from your Google Drive to your local session on Google Colab.

[xl-sr]

Hi David :) definitely try copying the pkl file to your local session. Also, could you post the error message you receive? I assume it simply does not find the pkl?

[DavidRees87]

Hi Axel, Thank you for your reply and help. I did try that but kept getting disconnected before the large file could upload. However I did find a solution. For some reason if I ran the example project first and generated a video I was then able to generate one from the .pkl located on my Gdrive. I’m not sure why as everything else was exactly the same but it seems to have worked. Is it possible to change how the seeds transform in the videos generated? So that they change at a constant rate rather than speeding up and down? Thank you so much for your help, I greatly appreciate it! David

…

Sent from my iPhone

On 7 Feb 2022, at 10:03, Axel Sauer ***@***.***> wrote:  Hi David :) definitely try copying the pkl file to your local session. Also, could you post the error message you receive? I assume it simply does not find the pkl? — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[DavidRees87]

When running the training with 1024x1024 images I get this error:

/usr/local/lib/python3.7/dist-packages/torch/nn/modules/module.py in convert(t)
895 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None,
896 non_blocking, memory_format=convert_to_format)
--> 897 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
898
899 return self._apply(convert)

RuntimeError: CUDA out of memory. Tried to allocate 64.00 MiB (GPU 0; 15.90 GiB total capacity; 14.87 GiB already allocated; 27.75 MiB free; 14.87 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

I have Colab Pro+ and it doesn't look like there is a lack of memory. Any suggestions on what I can do?

Thanks

[woctezuma]

RuntimeError: CUDA out of memory

Relevant:

• Config for 11GB GPU #5
• colab k80 is very slow #28
• Out of memory with batch at 1 #40

[DavidRees87]

Thank you, #5 worked. I also tried using 4096x4096 images but got an error message. Are those files too large to use? Thanks David

…

Sent from my iPhone

On 9 Feb 2022, at 16:59, Wok ***@***.***> wrote:  RuntimeError: CUDA out of memory Relevant: #5 #28 #40 — Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you authored the thread.

[xl-sr]

yes, those are very likely too large and need some special treatment.