Train full batch gradient descent model from scratch

Implement full training loop with vectorized batch gradient descent using NumPy only.
Plot loss across iterations and report final MAE, RMSE, and R2 on test data.

Start by inspecting the repository's Jupyter notebook and locate the model-training and evaluation sections. Implement the full vectorized batch gradient descent loop with NumPy, plot loss across iterations, and verify final MAE, RMSE, and R2 on the test data.

Written by the indexing model from the issue text.