Error in training the model

[anas-zafar]

Thanks for the amazing work @Sun-XH. I am trying to reproduce the results mentioned in the paper. I am having this issue while training on OPD Multi dataset, using pretrained weights of OPD Real to train OPDFormer-O. Could you please guide me if I am missing out something

Config '/home/research/OPDMulti/data/OPDMulti/obj_info.json' has no VERSION. Assuming it to be compatible with latest v2.
Traceback (most recent call last):
  File "train.py", line 179, in <module>
    args=(args,),
  File "/home/ubuntu/anaconda3/envs/opdmulti/lib/python3.7/site-packages/detectron2/engine/launch.py", line 82, in launch
    main_func(*args)
  File "train.py", line 160, in main
    cfg = setup_cfg(args)
  File "train.py", line 51, in setup_cfg
    cfg.merge_from_file(args.config_file)
  File "/home/ubuntu/anaconda3/envs/opdmulti/lib/python3.7/site-packages/detectron2/config/config.py", line 54, in merge_from_file
    self.merge_from_other_cfg(loaded_cfg)
  File "/home/ubuntu/anaconda3/envs/opdmulti/lib/python3.7/site-packages/fvcore/common/config.py", line 132, in merge_from_other_cfg
    return super().merge_from_other_cfg(cfg_other)
  File "/home/ubuntu/anaconda3/envs/opdmulti/lib/python3.7/site-packages/yacs/config.py", line 217, in merge_from_other_cfg
    _merge_a_into_b(cfg_other, self, self, [])
  File "/home/ubuntu/anaconda3/envs/opdmulti/lib/python3.7/site-packages/yacs/config.py", line 491, in _merge_a_into_b
    raise KeyError("Non-existent config key: {}".format(full_key))
KeyError: 'Non-existent config key: scene_00000_00_0_1'

[Sun-XH]

Hi there, Thanks for your interest. From the error, I believe that you're using the wrong config file. The obj_info.json is the model_attr_file, which should be filled to the model_attr_path. The README file has a detailed instruction on what config file you should use. For example, if you want to train the -O version, the config file should be OPDMulit/opdformer/configs/opd_o_real.yaml Hope this can solve your issue.