RuntimeError: CUDA out of memory. Tried to allocate 768.00 MiB (GPU 0; 6.00 GiB total capacity; 3.63 GiB already allocated; 328.06 MiB free; 3.90 GiB reserved in total by PyTorch) If r eserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

[nanshanvv]

I set the batch_size to 1 and still get this error when I run the model.
Here is the gpu information of my computer.

Could you tell me how to fix this error?

[WuJunde]

1. if 3D, decrease -chunk, -num_sample, and some other parameters I mentioned in README
2. parameter pruning to fp32 or fp16. I have not implemented it, so you need to implement yourself. Luckily, there are many relevant sources on GitHub. I may update it in the future
3. change a bigger GPU, 6GB memory is not enough for most DL projects now

[nanshanvv]

Thank you for your reply

[AmrinKareem]

Hi @nanshanvv have you solved this issue? Could you please share your solution? Thanks!

[nanshanvv]

Hi @nanshanvv have you solved this issue? Could you please share your solution? Thanks!

Hi, I solved this problem by buying cloud servers online to get more computing power.

[Nimophilist]

Hi，Could you share what the configuration of GPU used is?