| """Carga model.safetensors + config.json y predice el precio de casas nuevas.
|
|
|
| Uso:
|
| python predict.py
|
| """
|
|
|
| import json
|
| from pathlib import Path
|
|
|
| import numpy as np
|
| import torch
|
| from safetensors.torch import load_file
|
|
|
| from modelo import RegresionLineal
|
|
|
| CARPETA = Path(__file__).parent
|
|
|
|
|
| config = json.loads((CARPETA / "config.json").read_text(encoding="utf-8"))
|
| media = np.array(config["normalizacion"]["media"], dtype="float32")
|
| desvio = np.array(config["normalizacion"]["desvio"], dtype="float32")
|
|
|
|
|
| modelo = RegresionLineal(n_entradas=len(config["entradas"]))
|
| modelo.load_state_dict(load_file(CARPETA / "model.safetensors"))
|
| modelo.eval()
|
|
|
|
|
| casas_nuevas = np.array([[50, 0, 0], [80, 1, 0], [110, 1, 1]], dtype="float32")
|
| with torch.no_grad():
|
| precios = modelo(torch.tensor((casas_nuevas - media) / desvio)).numpy().ravel()
|
|
|
| for casa, precio in zip(casas_nuevas, precios):
|
| print(f"{dict(zip(config['entradas'], casa.astype(int).tolist()))} -> {precio:.1f} miles de US$")
|
|
|
