app.py

import streamlit as st
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

# Load model and tokenizer
@st.cache_resource()
def load_model():
    model_name = "deepseek-ai/deepseek-coder-1.3b-instruct"  # Using smaller model

    tokenizer = AutoTokenizer.from_pretrained(model_name)

    model = AutoModelForCausalLM.from_pretrained(
        model_name, 
        torch_dtype=torch.float32,  # Use float32 for CPU (float16 needs GPU)
        device_map={"": "cpu"}  # Force model to use CPU
    )

    return model, tokenizer

model, tokenizer = load_model()