hzxie/GRNet

How to evaluate GRNet on Completion 3d Benchmark?

Closed this issue · 2 comments

Hi,
Thanks for your great work,and I saw the great result of the GRNet on the Leaderboard of Completion 3d Benchmark.

When I tried to evaluate GRNet on the completion 3d benchmark, I found some difficulties, such as the version of cuda and pytorch. Completion 3d benchmark uses cuda9.0 and pytorch0.4.0.

Could you please give me some advice ?

Thanks a lot!

hzxie commented

We don't rely on any code from the Completion3D benchmark.
If you want to inference on Completion3D, you can simply run:

python3 runner.py --inference --weights=/path/to/pretrained/model.pth

Meanwhile, please update the __C.DATASET.TEST_DATASET to Completion3D in config.py.

Thanks a lot for your reply!