search

IDEA-Research / DINO

[ICLR 2023] Official implementation of the paper "DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection"
Apache License 2.0
2.19k stars 243 forks source link

No module named 'MultiScaleDeformableAttention' #194

Closed IAmNewToCode closed 1 year ago

IAmNewToCode commented 1 year ago

Hi, I've set up everything as listed in the documentation, and I want to do inference with my custom model.

I have already done python setup.py build install

but it is still throwing me an error when I try to run inference.

args = SLConfig.fromfile(model_config_path) args.device = 'cuda' model, criterion, postprocessors = build_model_main(args) checkpoint = torch.load(model_chk_path, map_location='cpu') model.load_statedict(checkpoint['model']) = model.eval()

it throws an error at model, criterion, postprocessors = build_model_main(args)

ModuleNotFoundError: No module named 'MultiScaleDeformableAttention'

I am using google colab, but I'm not exactly sure how the fix here works: https://github.com/IDEA-Research/DINO/issues/62 Any help is appreciated!