Generate a new ConflictQA-popQA dataset with other language model

[acDante]

Hi, thanks for the great work! I would like to generate a new version of ConflictQA-popQA dataset using Llama3-8b model. Could you explain how to use your code base to generate the knowledge conflict data? What steps should I follow? I find the build_prompt functions for popQA in prompt_preparation.py do not seem to work for the original data/popQA.tsv dataset. Would you mind sharing the popQA data in json format?

I got the following error when running run.py:

Traceback (most recent call last):
  File "/mnt/ceph_rbd/LLM-Knowledge-Conflict/code/run.py", line 19, in <module>
    test_data = build_zeroshot_prompt_popQA(args.input_file, model_name=args.model_name)
  File "/mnt/ceph_rbd/LLM-Knowledge-Conflict/code/prompt_preparation.py", line 64, in build_zeroshot_prompt_popQA
    unit = json.loads(line)
  File "/mnt/ceph_rbd/miniconda3/envs/kc/lib/python3.10/json/__init__.py", line 346, in loads
    return _default_decoder.decode(s)
  File "/mnt/ceph_rbd/miniconda3/envs/kc/lib/python3.10/json/decoder.py", line 337, in decode
    obj, end = self.raw_decode(s, idx=_w(s, 0).end())
  File "/mnt/ceph_rbd/miniconda3/envs/kc/lib/python3.10/json/decoder.py", line 355, in raw_decode
    raise JSONDecodeError("Expecting value", s, err.value) from None
json.decoder.JSONDecodeError: Expecting value: line 1 column 1 (char 0)