from transformers import AutoModelForSequenceClassification, AutoTokenizer
from huggingface_hub import hf_hub_download

model = AutoModelForSequenceClassification.from_pretrained("shyeon5898/paxmeter")
tokenizer = AutoTokenizer.from_pretrained("shyeon5898/paxmeter")

rep=hf_hub_download(repo_id="shyeon5898/paxmeter", filename="id2label.json")
with open(rep, "r") as json_file:
    i2l = json.load(json_file)
f_pipe = pipeline("text-classification", model=f_model, tokenizer=tokenizer)
input_text = "The South's unification minister, Jeong Se-hyun, told Kim Yong-nam in Pyongyang that the revelation had darkened the mood, and urged the North to end its programme"
outputs = f_pipe(a)
outputs[0]['label']
print(i2l[outputs[0]['label'].split('_')[1]])

Result

-0.108

Positive results indicate that the world has become more peaceful, while negative results signify that the world has become more chaotic.