|
|
from llm.llm import LLM |
|
|
from utils.utils import read_json_file, write_json_file, write_text_file, json_to_markdown, read_text_file, parse_llm_output_to_json |
|
|
from agent.create_charts import Chart |
|
|
|
|
|
import os |
|
|
from prompt.template import PROBLEM_EXTRACT_PROMPT |
|
|
|
|
|
|
|
|
|
|
|
if __name__ == "__main__": |
|
|
import glob |
|
|
files = glob.glob('/Users/ann/Downloads/methmatical_paper_extraction/parse/2025_*/content/*.md') |
|
|
llm = LLM('chatgpt-4o-latest') |
|
|
data = {} |
|
|
for file in files: |
|
|
year, _, _, problem_type = file.split('/')[-1].split('.')[0].split('_') |
|
|
problem_name = f'{year}_{problem_type}' |
|
|
problem_str = read_text_file(file) |
|
|
problem = llm.generate(PROBLEM_EXTRACT_PROMPT.format(question=problem_str), problem_str) |
|
|
problem = parse_llm_output_to_json(problem) |
|
|
|
|
|
print(problem) |
|
|
write_json_file(f'/Users/ann/Documents/projects/math_modeling/data/actor_data/input/problem/{problem_name}.json', problem) |
|
|
|
