Skip to content

Inference on ScanNet v2 RuntimeError: CUDA error: invalid argument #34

Open
@hazelAutumn

Description

Hi, thanks for your excellent work.
I tried to reproduce the result of the pretrained model on the validation set of ScanNet v2.0. I followed the instructions in your repo to process the data and installation, but I faced this issue:
image

Does anyone have any suggestions about the root of this issue? Thank you so much.

My environment:

  • CUDA 10.2.
  • spconv 1.0
  • pytorch 1.1.0
  • python 3.7

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions