search

eriksandstroem / Point-SLAM

Point-SLAM: Dense Neural Point Cloud-based SLAM
Apache License 2.0
392 stars 32 forks source link

About run #16

Closed ywaison closed 10 months ago

ywaison commented 10 months ago

Hello, thank you very much for the code, when I configured the environment according to reademe, running the code “python run.py configs/TUM_RGBD/freiburg1_desk.yaml”, I got an error as shown below RuntimeError: Attempting to deserialize object on a CUDA device but torch.cuda.is_available() is False. If you are running on a CPU-only machine, please use torch.load with map_location=torch.device('cpu') to map your storages to the CPU. I can't wait for your advice and help, thank you again！

eriksandstroem commented 10 months ago

Hi, You are welcome! This error would suggest that you do not have a GPU available as the "torch.cuda.is_available()" evidently returns False. First confirm with some simple tests that you can load torch tensors to the GPU. What does the command "nvidia-smi" return in the terminal?

ywaison commented 10 months ago

Hi, You are welcome! This error would suggest that you do not have a GPU available as the "torch.cuda.is_available()" evidently returns False. First confirm with some simple tests that you can load torch tensors to the GPU. What does the command "nvidia-smi" return in the terminal?

Glad to get your prompt answer, my computer cuda version is as follows | NVIDIA-SMI 535.72 Driver Version: 536.45 CUDA Version: 12.2 And I have cuda version 11.3 installed and pytorch version 1.12.1

eriksandstroem commented 10 months ago

And what kind of GPU does it list? E.g. how much memory does it have?

ywaison commented 10 months ago

nvidia-smi

ZI{02QW_9A HX7$_013(P(O

eriksandstroem commented 10 months ago

Hi @ywaison, Did you figure out the issue?

eriksandstroem commented 10 months ago

Closing for now.