laughtervv / 3dn Goto Github PK

Can you please add an appropriate license file.

The pretrained model provided in the google drive only has xxx.ckpt.meta and xxx.ckpt.index, and there is no xxx.ckpt.data-0000-of-00001 in it. Without such files, the pretrained model can not be restored.

I am here to request the releasing of these xxx.ckpt.data-00000-of-00001

dear author, thank for you provide code, i have some problem when i run your code in my own dataset.

"tensorflow.python.framework.errors_impl.InvalidArgumentError: No OpKernel was registered to support Op 'MeshSampling' with these attrs. Registered devices: [CPU], Registered kernels:
device='GPU‘"

When I run the code, I did not find the data files. Can you provide the download link?

I compiled the tf_ops/mesh_sampling .
2019-09-07 16:13:41.098034: F tensorflow/core/lib/monitoring/collection_registry.cc:77] Cannot register 2 metrics with the same name: /tensorflow/cc/saved_model/load_attempt_count

I used the official ShapeNetCore.v2 dataset. But I do not know how to organize the pair of source and target mesh. How to generate the filelist the training code used:

parser.add_argument('--train_lst_mesh', default='/media/hdd2/data/ShapeNet/filelists/ShapeNetCore.v2.h5/all_template_train_50000_200000.lst', help='train mesh data list')
parser.add_argument('--valid_lst_mesh', default='/media/hdd2/data/ShapeNet/filelists/ShapeNetCore.v2.h5/all_template_valid_50000_200000.lst', help='test mesh data list')

this operation is not found in the repo.