You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, mafia inference crashes due to out of memory error on the test dataset. I'm running it on RTX3080 Ti with 12 GB vRAM.
Please specify GPU memory requirements and which hardware is supported.
File "/home/lpryszcz/src/mAFiA/RODAN/models.py", line 41, in forward
return x * (torch.tanh(torch.nn.functional.softplus(x)))
torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 1.91 GiB (GPU 0; 11.76 GiB total capacity; 6.44 GiB already allocated; 296.19 MiB free; 11.09 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
The text was updated successfully, but these errors were encountered:
Hi @lpryszcz , the settings giving in the readme were tested on a Quadro RTX 6000 with 24GB memory. If you have less than that, please set --max_num_reads to a smaller number, say, 500.
Hi, mafia inference crashes due to out of memory error on the test dataset. I'm running it on RTX3080 Ti with 12 GB vRAM.
Please specify GPU memory requirements and which hardware is supported.
The text was updated successfully, but these errors were encountered: