Open delonixsen opened 2 years ago
I got the same error:
RuntimeError: CUDA out of memory. Tried to allocate 1.51 GiB (GPU 0; 3.82 GiB total capacity; 1.04 GiB already allocated; 99.75 MiB free; 1.04 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 error happened when I ran:
./experiments/scripts/demo.sh
GPU : NVIDIA Quadro T2000 with Max-Q
I'm using:
Full Traceback:
finished loading sdf !
Traceback (most recent call last):
File "./tools/test_images.py", line 193, in <module>
cfg.sdf_optimizer = sdf_optimizer(cfg.TEST.CLASSES[1:], sdf_files)
File "/home/.../PoseCNN-PyTorch/tools/../lib/sdf/sdf_optimizer.py", line 73, in __init__
self.sdf_torch = self.sdf_torch.cuda()
RuntimeError: CUDA out of memory. Tried to allocate 1.51 GiB (GPU 0; 3.82 GiB total capacity; 1.04 GiB already allocated; 99.75 MiB free; 1.04 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
run ./experiments/scripts/demo.sh 0
RuntimeError: CUDA out of memory. Tried to allocate 1.51 GiB (GPU 0; 3.82 GiB total capacity; 1.04 GiB already allocated; 74.38 MiB free; 1.04 GiB reserved in total by PyTorch)
Can someone help me with this problem? since I tried to reduce batch size, and nothing works
GPU : NVIDIA GeForce GTX 1650 4GB
Should I upgrade the Graphics Card?