import sys, os sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..'))) from utils import create_vector_store, save_embedded_data, load_documents_from_directory, process_directory from loggings.logging_config import logger import time # This module will load in the data, you only need to add the data path to it. data_path = './data' def load_data(data_path: str): logger.info(f"Loading data from {data_path}") start_time = time.time() # logging the start time logger.info(f"Start time: {start_time}") documents, docs_id, num_pages = load_documents_from_directory(data_path) logger.info(f"Data loaded") logger.info(f"Creating vector store") embed_db = create_vector_store(documents, docs_id, num_pages) logger.info(f"Vector store created") logger.info(f"Saving the vector store") # saving the embedded data save_embedded_data(embed_db) logger.info(f"Vector store saved") end_time = time.time() logger.info(f"End time: {end_time}") time_taken = end_time - start_time logger.info(f"Time taken: {time_taken}") print("Vector store created and saved") # creating the thumbnails logger.info(f"Creating thumbnails") status = process_directory(data_path) print(f"{status}: Thumbnails created.") logger.info(f"Thumbnails created") return embed_db if __name__ == "__main__": load_data(data_path)