import streamlit as st
import os

def safe_import(module_name):
    try:
        return __import__(module_name)
    except ImportError:
        return None

torch = safe_import('torch')
if torch is None:
    st.error("Torch is not installed yet. Please wait a moment for the dependencies to install.")
    st.stop()

import torch.nn as nn

# architecture
class AddModel(nn.Module):
    def __init__(self):
        super(AddModel, self).__init__()
        self.fc1 = nn.Linear(2, 32)
        self.relu1 = nn.ReLU()
        self.fc2 = nn.Linear(32, 64)
        self.relu2 = nn.ReLU()
        self.fc3 = nn.Linear(64, 1)

    def forward(self, x):
        x = self.relu1(self.fc1(x))
        x = self.relu2(self.fc2(x))
        x = self.fc3(x)
        return x

def load_model(model_path):
    model = AddModel()
    model.load_state_dict(torch.load(model_path, map_location=torch.device('cpu')))
    model.eval()  # evaluation mode
    return model

def predict_sum(model, x1, x2):
    with torch.no_grad():
        input_tensor = torch.tensor([[x1, x2]], dtype=torch.float32)
        prediction = model(input_tensor)
        return prediction.item()

# Streamlit app
def main():
    st.title("Sum Predictor using Neural Network (Using version M3)")

    model_path = "MA3T.pth"  # Update with your model path if necessary
    if os.path.exists(model_path):
        model = load_model(model_path)
        st.success("Model loaded successfully.")
        
        x1 = st.number_input("Enter the first number:", value=0.0)
        x2 = st.number_input("Enter the second number:", value=0.0)

        if st.button("Predict"):
            predicted_sum = predict_sum(model, x1, x2)
            correct_sum = x1 + x2  # Calculate the correct answer
            st.write(f"The predicted sum of {x1} and {x2} is: {predicted_sum:.2f}")
            st.write(f"The correct sum of {x1} and {x2} is: {correct_sum:.2f}")
    else:
        st.error("Model file not found. Please upload the model.")

if __name__ == "__main__":
    main()