TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

[28YOGESH]

Traceback (most recent call last):
File "api_for_multistep.py", line 110, in
model_dump="model_saved/model_for_multi_step.ckpt"
File "api_for_multistep.py", line 41, in init
one_step = prepare_single_step_model(model_dump)
File "/home/csn/Desktop/RetroExplainer/model/util.py", line 119, in prepare_single_step_model
single_step_model = RetroAGT.load_from_checkpoint(ckpt, strict=False)
File "/home/csn/anaconda3/envs/RetroExplainer/lib/python3.7/site-packages/pytorch_lightning/core/saving.py", line 169, in load_from_checkpoint
model = cls._load_model_state(checkpoint, *args, **kwargs)
File "/home/csn/anaconda3/envs/RetroExplainer/lib/python3.7/site-packages/pytorch_lightning/core/saving.py", line 205, in _load_model_state
model = cls(*cls_args, **cls_kwargs)
File "/home/csn/Desktop/RetroExplainer/model/RetroAGT.py", line 90, in init
assert os.path.isfile(leaving_group_path)
File "/home/csn/anaconda3/envs/RetroExplainer/lib/python3.7/genericpath.py", line 30, in isfile
st = os.stat(path)
TypeError: stat: path should be string, bytes, os.PathLike or integer, not NoneType

Please sir give a solution for this error. How i can solve this error.