How to run ToothGroupNetwork

[hx171220]

Hello:

I'm using the main branch an follow the procedure:

1. use challenger dataset to process,
2. use processing data to train
3. use the training results to verify However, the following error occurred My running environment：
   • The graphics card is RTX 3090
   • Python 3.7.0
   • torch: 1.13.1+cu117, cuda 11.7,cudnn 8500
   • Ubuntu 20.4 Now I don't know where the problem is, Can you help me?

[limhoyeon]

please tell me what options you used for train and inference!

[hx171220]

This is my options:

• train: python start_train.py --model_name "tgnet_fps" --config_path "train_configs/tgnet_fps.py" --experiment_name "tgnet" --input_data_dir_path "../processDir" --train_data_split_txt_path "../split/base_name_train_fold.txt" --val_data_split_txt_path "../split/base_name_val_fold.txt"
• verify: python start_inference.py --input_dir_path /home/ecs-user/ToothGroupNetwork-main/modelDir --split_txt_path /home/ecs-user/ToothGroupNetwork-main/split/base_name_test_fold.txt --save_path /home/ecs-user/ToothGroupNetwork-main/result --model_name tgnet --checkpoint_path /home/ecs-user/ToothGroupNetwork-main/ckpts/tgnet_75 --checkpoint_path_bdl /home/ecs-user/ToothGroupNetwork-main/ckpts/tgnet_val

tgnet_75.h5 and tgnet_val.h5 are generated training sets

I modified the following configuration：

Is there anything else that needs to be changed?

[JeffreyWardman]

@limhoyeon has not provided up to date pretrained model checkpoints and needs to reupload them

[limhoyeon]

Make sure to train both the tgnet_fps model and the tgnet_bdl model! Our tgnet needs two checkpoints(tgnet_bdl, tgnet_fps) to make inferences. For more details, see the train section of the readme. Let me know if you have any further questions.

[hx171220]

@limhoyeon Thank you. I'll check it first and get back to you if I have any questions