Using pretrained JointBERT model

Models
1

Hi there,
I am working on JointBERT from huggingface. I trained the model on my custom dataset and saved locally as:

main.py:
...
parser.add_argument("--model_dir", default=custom_model, required=True, type=str, help="Path to save, load model")

Now, I am using this pretrained custom_model in predict.py:

model = MODEL_CLASSES[args.model_type][1].from_pretrained(args.model_dir,
                                                                  args=args,
                                                                  intent_label_lst=get_intent_labels(args),
                                                                  slot_label_lst=get_slot_labels(args))
...
parser.add_argument("--model_dir", default="./saved-models/custom_model", type=str, help="Path to save, load model")

I can assure that /saved-models/custom_model folder exists in the directory, and the custom_model folder contains:config.json, intent_label.txt, pytorch_model.bin, slot_label.txt, training_args.bin.

The error message is:

OSError: ticketmachine_model is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models'
If this is a private repository, make sure to pass a token having permission to this repo with `use_auth_token` or log in with `huggingface-cli login` and pass `use_auth_token=True`.

Can anyone help? Thank you.