model.py

import torch
from typing import Tuple
from transformers import AutoTokenizer, AutoModelForCausalLM

# Initialize the model and tokenizer variables as None
tokenizer = None
model = None


def get_model_and_tokenizer() -> Tuple[AutoModelForCausalLM, AutoTokenizer]:
    """
    Returns the preloaded model and tokenizer. If they haven't been loaded before, loads them.

    Returns:
        tuple: A tuple containing the preloaded model and tokenizer.
    """
    global model, tokenizer
    if model is None or tokenizer is None:
        # Set device
        device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

        # Load the tokenizer and the model
        tokenizer = AutoTokenizer.from_pretrained("MikeMpapa/lmd_mmm_tokenizer_tutorial_artist", use_auth_token=True)
        model = AutoModelForCausalLM.from_pretrained(
            "MikeMpapa/lmd-4bars-2048-epochs7",
            use_auth_token=True
        )

        # Move model to device
        model = model.to(device)

    return model, tokenizer