import os from dotenv import load_dotenv load_dotenv() class Settings: APP_LEVEL_PATH: str = os.path.split(os.path.dirname(os.path.abspath(__file__)))[0] ROOT_LEVEL_PATH: str = os.path.split(APP_LEVEL_PATH)[0] APP_NAME: str = os.getenv('APP_NAME', "Classification-Backend") ENVIRONMENT: str = os.getenv('ENVIRONMENT', "development") PORT: int = os.getenv('PORT', 8000) HF_ACCESS_TOKEN = os.getenv("HF_TOKEN") BERT_TOKENIZER_PATH = os.getenv("BERT_MODEL_PATH") LSTM_TOKENIZER_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/lstm/tokenizer.json') ML_TFIDF_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/tfidf_vectorizer.pkl') MLP_TOKENIZER_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/mlp/tokenizer.json') BERT_MODEL_PATH = os.getenv("BERT_MODEL_PATH") LSTM_MODEL_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/lstm/LSTM_nepali_news_classifier_model.keras') ML_NAIVE_BAYES_MODEL_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/Naive_Bayes_Classifier.joblib') ML_GRADIENT_BOOSTING_MODEL_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/Gradient_Boosting_Classifier.joblib') ML_XG_BOOST_MODEL_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/XG_Boost_Classifier.joblib') MLP_MODEL_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/mlp/MLP_nepali_news_classifier_model.keras') BERT_LABEL_ENCODER = os.path.join(ROOT_LEVEL_PATH, 'outputs/bert/label_encoder.pkl') LSTM_LABEL_ENCODER = os.path.join(ROOT_LEVEL_PATH, 'outputs/lstm/label_encoder.pkl') ML_LABEL_ENCODER = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/label_encoder.pkl') MLP_LABEL_ENCODER = os.path.join(ROOT_LEVEL_PATH, 'outputs/mlp/label_encoder.pkl') LSTM_PARAMETERS_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/lstm/parameters.json') ML_PARAMETERS_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/ml/parameters.json') MLP_PARAMETERS_PATH = os.path.join(ROOT_LEVEL_PATH, 'outputs/mlp/parameters.json') settings = Settings()