search

NVlabs / contact_graspnet

Efficient 6-DoF Grasp Generation in Cluttered Scenes
Other
324 stars 111 forks source link

Result different during inference #13

Closed y556zhao closed 2 years ago

y556zhao commented 2 years ago

Hi,

I was using default parameters to test the model on 7.npy, and the resulting grasps are sparse as shown in the image below. May I ask if I was doing anything wrong?

Here is the command I used: python contact_graspnet/inference.py --np_path=test_data/7.npy --local_regions --filter_grasps

Checkpoint: checkpoints/scene_test_2048_bs3_hor_sigma_001

Thank you! @MartinSmeyer grasp

MartinSmeyer commented 2 years ago

Hi,

This looks like issue #9

Apparently, users of newer GPUs require tensorflow-gpu=2.5 with CUDA >= 11 to run this code. Recompile the tf_ops if you use a newer version. Does this help?

I will update the requirements soon.

Best,

Martin

y556zhao commented 2 years ago

Yes, it works! Thank you so much for helping!