GitHub Repository: ManuSanchez02/7506R-2c2022-GRUPO09
Path: blob/main/tp2/7506R_TP2_GRUPO09_ENTREGA_N2_(redes_neuronales).ipynb
⁹¹ views

Kernel: Python 3

Configuración inicial

Importamos e instalamos las bibliotecas necesarias.

In [19]:

!pip install keras_tuner

Out[19]:

Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/
Requirement already satisfied: keras_tuner in /usr/local/lib/python3.8/dist-packages (1.1.3)
Requirement already satisfied: numpy in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (1.21.6)
Requirement already satisfied: packaging in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (21.3)
Requirement already satisfied: kt-legacy in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (1.0.4)
Requirement already satisfied: tensorboard in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (2.9.1)
Requirement already satisfied: requests in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (2.23.0)
Requirement already satisfied: ipython in /usr/local/lib/python3.8/dist-packages (from keras_tuner) (7.9.0)
Requirement already satisfied: pygments in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (2.6.1)
Requirement already satisfied: jedi>=0.10 in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (0.18.2)
Requirement already satisfied: backcall in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (0.2.0)
Requirement already satisfied: prompt-toolkit<2.1.0,>=2.0.0 in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (2.0.10)
Requirement already satisfied: pexpect in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (4.8.0)
Requirement already satisfied: decorator in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (4.4.2)
Requirement already satisfied: pickleshare in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (0.7.5)
Requirement already satisfied: traitlets>=4.2 in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (5.1.1)
Requirement already satisfied: setuptools>=18.5 in /usr/local/lib/python3.8/dist-packages (from ipython->keras_tuner) (57.4.0)
Requirement already satisfied: parso<0.9.0,>=0.8.0 in /usr/local/lib/python3.8/dist-packages (from jedi>=0.10->ipython->keras_tuner) (0.8.3)
Requirement already satisfied: wcwidth in /usr/local/lib/python3.8/dist-packages (from prompt-toolkit<2.1.0,>=2.0.0->ipython->keras_tuner) (0.2.5)
Requirement already satisfied: six>=1.9.0 in /usr/local/lib/python3.8/dist-packages (from prompt-toolkit<2.1.0,>=2.0.0->ipython->keras_tuner) (1.15.0)
Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.8/dist-packages (from packaging->keras_tuner) (3.0.9)
Requirement already satisfied: ptyprocess>=0.5 in /usr/local/lib/python3.8/dist-packages (from pexpect->ipython->keras_tuner) (0.7.0)
Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.8/dist-packages (from requests->keras_tuner) (3.0.4)
Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.8/dist-packages (from requests->keras_tuner) (1.24.3)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.8/dist-packages (from requests->keras_tuner) (2022.9.24)
Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.8/dist-packages (from requests->keras_tuner) (2.10)
Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (1.8.1)
Requirement already satisfied: werkzeug>=1.0.1 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (1.0.1)
Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (0.6.1)
Requirement already satisfied: grpcio>=1.24.3 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (1.50.0)
Requirement already satisfied: google-auth<3,>=1.6.3 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (2.14.1)
Requirement already satisfied: wheel>=0.26 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (0.38.4)
Requirement already satisfied: absl-py>=0.4 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (1.3.0)
Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (0.4.6)
Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (3.4.1)
Requirement already satisfied: protobuf<3.20,>=3.9.2 in /usr/local/lib/python3.8/dist-packages (from tensorboard->keras_tuner) (3.19.6)
Requirement already satisfied: cachetools<6.0,>=2.0.0 in /usr/local/lib/python3.8/dist-packages (from google-auth<3,>=1.6.3->tensorboard->keras_tuner) (5.2.0)
Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.8/dist-packages (from google-auth<3,>=1.6.3->tensorboard->keras_tuner) (0.2.8)
Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.8/dist-packages (from google-auth<3,>=1.6.3->tensorboard->keras_tuner) (4.9)
Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.8/dist-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard->keras_tuner) (1.3.1)
Requirement already satisfied: importlib-metadata>=4.4 in /usr/local/lib/python3.8/dist-packages (from markdown>=2.6.8->tensorboard->keras_tuner) (4.13.0)
Requirement already satisfied: zipp>=0.5 in /usr/local/lib/python3.8/dist-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard->keras_tuner) (3.10.0)
Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /usr/local/lib/python3.8/dist-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard->keras_tuner) (0.4.8)
Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.8/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard->keras_tuner) (3.2.2)

In [20]:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_absolute_error, mean_squared_error
from sklearn.preprocessing import StandardScaler
from sklearn import preprocessing

import tensorflow as tf
from tensorflow import keras
import keras_tuner as kt
from keras.models import Sequential
from keras.layers.core import Dense  
from sklearn.metrics import (
    confusion_matrix,
    recall_score, 
    accuracy_score, 
    f1_score,
    mean_squared_error,
    r2_score,
)
from sklearn.model_selection import KFold
from keras.losses import sparse_categorical_crossentropy
from keras.optimizers import Adam

np.random.seed(1)
tf.random.set_seed(1)

Funciones útiles

In [21]:

def graficar_matriz_de_confusion(y_test, y_pred, labels):
    cm = confusion_matrix(y_test, y_pred)
    fig = plt.figure(figsize = (12, 10))
    sns.heatmap(
        cm, annot=True, xticklabels=labels, yticklabels=labels, cmap="GnBu", fmt="g"
    )
    plt.title("Matriz de confusión")
    plt.xlabel("Predichos")
    plt.ylabel("Verdaderos")
    plt.show()

def imprimir_metricas_de_clasificacion(y_test, y_pred):
    accuracy = accuracy_score(y_test, y_pred)
    recall = recall_score(y_test, y_pred, average="weighted")
    f1 = f1_score(y_test, y_pred, average="weighted")

    print("\nMétricas de clasificación\n")
    print("Accuracy: " + str(accuracy))
    print("Recall: " + str(recall))
    print("F1 score: " + str(f1))

def imprimir_metricas_de_regresion(target_test, precios_predichos):
    # Error cuadrático medio
    mse = mean_squared_error(
        y_true=target_test, y_pred=precios_predichos, squared=True
    )

    # Raíz del error cuadrático medio
    rmse = mean_squared_error(
        y_true=target_test, y_pred=precios_predichos, squared=False
    )

    # Score R2
    r2 = r2_score(target_test, precios_predichos)

    print("\nMétricas de regresión\n")
    print(f"El error (mse) de test es: {mse}")
    print(f"El error (rmse) de test es: {rmse}")
    print(f"El score R2 es: {r2}")

Importamos los datasets que se utilizarán para trabajar.

In [22]:

dataset_train_clasificacion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/clasificacion_train_feature.csv?raw=True")
dataset_test_clasificacion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/clasificacion_test_feature.csv?raw=True")

dataset_train_regresion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/regresion_train_feature.csv?raw=True")
dataset_test_regresion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/regresion_test_feature.csv?raw=True")

In [23]:

y_train_clasificacion_categorico = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/clasificacion_train_target.csv?raw=True")
y_test_clasificacion_categorico = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/clasificacion_test_target.csv?raw=True")

y_train_regresion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/regresion_train_target.csv?raw=True")
y_test_regresion = pd.read_csv("https://github.com/ManuSanchez02/7506R-2c2022-GRUPO09/blob/main/tp2/datasets/regresion_test_target.csv?raw=True")

In [24]:

def preparar_dataset(dataset):
  dataset.drop("Unnamed: 0", axis = 1, inplace=True)
  dataset.drop("id", axis = 1, inplace=True)
  return dataset

def preparar_target(target):
  target.drop("Unnamed: 0", axis = 1, inplace=True)
  return target

dataset_train_clasificacion = preparar_dataset(dataset_train_clasificacion)
dataset_test_clasificacion = preparar_dataset(dataset_test_clasificacion)
dataset_train_regresion = preparar_dataset(dataset_train_regresion)
dataset_test_regresion = preparar_dataset(dataset_test_regresion)

y_train_clasificacion_categorico = preparar_target(y_train_clasificacion_categorico)
y_test_clasificacion_categorico = preparar_target(y_test_clasificacion_categorico)
y_train_regresion = preparar_target(y_train_regresion)
y_test_regresion = preparar_target(y_test_regresion)

Convertimos las categorías del target de clasificación en categorías numéricas para poder utilizarlas en la red neuronal.

In [25]:

le = preprocessing.LabelEncoder()
y_train_clasificacion = le.fit_transform(y_train_clasificacion_categorico)
y_test_clasificacion = le.transform(y_test_clasificacion_categorico)

Out[25]:

/usr/local/lib/python3.8/dist-packages/sklearn/preprocessing/_label.py:115: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
  y = column_or_1d(y, warn=True)
/usr/local/lib/python3.8/dist-packages/sklearn/preprocessing/_label.py:133: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
  y = column_or_1d(y, warn=True)

Redes neuronales

Escalamos los datos para que performe mejor.

In [26]:

scaler_clasificacion = StandardScaler()
scaler_clasificacion.fit(dataset_train_clasificacion)

scaler_regresion = StandardScaler()
scaler_regresion.fit(dataset_train_regresion)

Out[26]:

StandardScaler()

In [27]:

x_train_clasificacion = scaler_clasificacion.transform(dataset_train_clasificacion)
x_test_clasificacion = scaler_clasificacion.transform(dataset_test_clasificacion)

x_train_regresion = scaler_regresion.transform(dataset_train_regresion)
x_test_regresion = scaler_regresion.transform(dataset_test_regresion)

Clasificación

Definimos las salidas del modelo.

In [28]:

outputs = {
    "bajo": 0,
    "medio": 1,
    "alto": 2,
}

Creamos el modelo secuencial de Keras.

In [39]:

def build_model(hp):
  model = keras.Sequential()
  model.add(keras.layers.Dense(
      hp.Int('units', min_value=32, max_value=512, step=16),
      activation=hp.Choice("activation", ["relu", "tanh"])))
  model.add(keras.layers.Dense(10, activation="tanh"))
  model.add(keras.layers.Dense(3, activation="softmax"))
  learning_rate = hp.Float("lr", min_value=1e-4, max_value=1e-1, sampling="log")

  model.compile(loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False), 
                optimizer=keras.optimizers.Adam(learning_rate=learning_rate),
                metrics=["accuracy"])
  return model

Hacemos un Randomized Search para buscar los mejores hiperparámetros.

In [42]:

tuner = kt.RandomSearch(
    build_model,
    objective='loss',
    max_trials=10,
    project_name="Clasificacion",
    overwrite=True)

In [44]:

stop_early = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=1)
tuner.search(x_train_clasificacion, y_train_clasificacion, epochs=100, validation_split=0.2, callbacks=[stop_early])
best_model = tuner.get_best_models()[0]

Out[44]:

Trial 10 Complete [00h 00m 31s]
loss: 0.8263664841651917

Best loss So Far: 0.8020757436752319
Total elapsed time: 00h 06m 13s

Imprimimos los resultados del mejor modelo obtenido.

In [57]:

best_hps = tuner.get_best_hyperparameters()[0]
best_hps.values

Out[57]:

{'units': 496, 'activation': 'tanh', 'lr': 0.00046002804367732453}

In [59]:

best_model = build_model(best_hps)
best_model.fit(x=x_train_clasificacion, y=y_train_clasificacion, epochs=500)

Out[59]:

Epoch 1/500
2163/2163 [==============================] - 7s 2ms/step - loss: 0.8633 - accuracy: 0.5845
Epoch 2/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8333 - accuracy: 0.5998
Epoch 3/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8243 - accuracy: 0.6036
Epoch 4/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8171 - accuracy: 0.6073
Epoch 5/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8122 - accuracy: 0.6077
Epoch 6/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8082 - accuracy: 0.6105
Epoch 7/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8054 - accuracy: 0.6107
Epoch 8/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8036 - accuracy: 0.6121
Epoch 9/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.8013 - accuracy: 0.6113
Epoch 10/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7998 - accuracy: 0.6128
Epoch 11/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7984 - accuracy: 0.6146
Epoch 12/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7967 - accuracy: 0.6147
Epoch 13/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7955 - accuracy: 0.6150
Epoch 14/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7937 - accuracy: 0.6149
Epoch 15/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7923 - accuracy: 0.6155
Epoch 16/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7918 - accuracy: 0.6171
Epoch 17/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7896 - accuracy: 0.6173
Epoch 18/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7893 - accuracy: 0.6190
Epoch 19/500
2163/2163 [==============================] - 10s 5ms/step - loss: 0.7878 - accuracy: 0.6202
Epoch 20/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7865 - accuracy: 0.6195
Epoch 21/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7854 - accuracy: 0.6196
Epoch 22/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7849 - accuracy: 0.6200
Epoch 23/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7837 - accuracy: 0.6215
Epoch 24/500
2163/2163 [==============================] - 12s 5ms/step - loss: 0.7832 - accuracy: 0.6212
Epoch 25/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7820 - accuracy: 0.6230
Epoch 26/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7809 - accuracy: 0.6246
Epoch 27/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7803 - accuracy: 0.6231
Epoch 28/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7802 - accuracy: 0.6230
Epoch 29/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7782 - accuracy: 0.6246
Epoch 30/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7780 - accuracy: 0.6247
Epoch 31/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7775 - accuracy: 0.6253
Epoch 32/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7765 - accuracy: 0.6258
Epoch 33/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7756 - accuracy: 0.6259
Epoch 34/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7749 - accuracy: 0.6263
Epoch 35/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7744 - accuracy: 0.6275
Epoch 36/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7734 - accuracy: 0.6269
Epoch 37/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7729 - accuracy: 0.6282
Epoch 38/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7723 - accuracy: 0.6283
Epoch 39/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7716 - accuracy: 0.6277
Epoch 40/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7707 - accuracy: 0.6295
Epoch 41/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7704 - accuracy: 0.6280
Epoch 42/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7696 - accuracy: 0.6294
Epoch 43/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7682 - accuracy: 0.6301
Epoch 44/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7683 - accuracy: 0.6307
Epoch 45/500
2163/2163 [==============================] - 10s 4ms/step - loss: 0.7672 - accuracy: 0.6308
Epoch 46/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7669 - accuracy: 0.6308
Epoch 47/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7658 - accuracy: 0.6317
Epoch 48/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7655 - accuracy: 0.6322
Epoch 49/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7651 - accuracy: 0.6328
Epoch 50/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7639 - accuracy: 0.6325
Epoch 51/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7632 - accuracy: 0.6325
Epoch 52/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7621 - accuracy: 0.6344
Epoch 53/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7626 - accuracy: 0.6339
Epoch 54/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7617 - accuracy: 0.6338
Epoch 55/500
2163/2163 [==============================] - 12s 5ms/step - loss: 0.7611 - accuracy: 0.6343
Epoch 56/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7609 - accuracy: 0.6346
Epoch 57/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7600 - accuracy: 0.6344
Epoch 58/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7591 - accuracy: 0.6349
Epoch 59/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7588 - accuracy: 0.6350
Epoch 60/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7578 - accuracy: 0.6364
Epoch 61/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7578 - accuracy: 0.6362
Epoch 62/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7576 - accuracy: 0.6351
Epoch 63/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7560 - accuracy: 0.6366
Epoch 64/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7560 - accuracy: 0.6376
Epoch 65/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7556 - accuracy: 0.6375
Epoch 66/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7552 - accuracy: 0.6396
Epoch 67/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7544 - accuracy: 0.6379
Epoch 68/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7539 - accuracy: 0.6395
Epoch 69/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7534 - accuracy: 0.6382
Epoch 70/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7527 - accuracy: 0.6391
Epoch 71/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7525 - accuracy: 0.6394
Epoch 72/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7520 - accuracy: 0.6394
Epoch 73/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7517 - accuracy: 0.6392
Epoch 74/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7511 - accuracy: 0.6397
Epoch 75/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7512 - accuracy: 0.6401
Epoch 76/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7502 - accuracy: 0.6404
Epoch 77/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7499 - accuracy: 0.6404
Epoch 78/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7496 - accuracy: 0.6406
Epoch 79/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7487 - accuracy: 0.6418
Epoch 80/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7485 - accuracy: 0.6434
Epoch 81/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7480 - accuracy: 0.6423
Epoch 82/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7471 - accuracy: 0.6425
Epoch 83/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7468 - accuracy: 0.6428
Epoch 84/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7466 - accuracy: 0.6422
Epoch 85/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7457 - accuracy: 0.6433
Epoch 86/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7457 - accuracy: 0.6434
Epoch 87/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7451 - accuracy: 0.6429
Epoch 88/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7442 - accuracy: 0.6447
Epoch 89/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7441 - accuracy: 0.6437
Epoch 90/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7439 - accuracy: 0.6442
Epoch 91/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7438 - accuracy: 0.6440
Epoch 92/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7432 - accuracy: 0.6435
Epoch 93/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7426 - accuracy: 0.6452
Epoch 94/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7421 - accuracy: 0.6458
Epoch 95/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7419 - accuracy: 0.6445
Epoch 96/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7418 - accuracy: 0.6455
Epoch 97/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7412 - accuracy: 0.6460
Epoch 98/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7407 - accuracy: 0.6467
Epoch 99/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7404 - accuracy: 0.6457
Epoch 100/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7397 - accuracy: 0.6462
Epoch 101/500
2163/2163 [==============================] - 15s 7ms/step - loss: 0.7396 - accuracy: 0.6466
Epoch 102/500
2163/2163 [==============================] - 10s 5ms/step - loss: 0.7391 - accuracy: 0.6462
Epoch 103/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7388 - accuracy: 0.6473
Epoch 104/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7379 - accuracy: 0.6481
Epoch 105/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7380 - accuracy: 0.6471
Epoch 106/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7373 - accuracy: 0.6487
Epoch 107/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7369 - accuracy: 0.6481
Epoch 108/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7366 - accuracy: 0.6495
Epoch 109/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7366 - accuracy: 0.6490
Epoch 110/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7361 - accuracy: 0.6483
Epoch 111/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7352 - accuracy: 0.6498
Epoch 112/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7345 - accuracy: 0.6491
Epoch 113/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7345 - accuracy: 0.6499
Epoch 114/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7338 - accuracy: 0.6493
Epoch 115/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7339 - accuracy: 0.6500
Epoch 116/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7336 - accuracy: 0.6506
Epoch 117/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7331 - accuracy: 0.6508
Epoch 118/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7328 - accuracy: 0.6495
Epoch 119/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7330 - accuracy: 0.6502
Epoch 120/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7314 - accuracy: 0.6518
Epoch 121/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7321 - accuracy: 0.6507
Epoch 122/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7313 - accuracy: 0.6524
Epoch 123/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7313 - accuracy: 0.6528
Epoch 124/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7307 - accuracy: 0.6524
Epoch 125/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7298 - accuracy: 0.6523
Epoch 126/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7295 - accuracy: 0.6538
Epoch 127/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7295 - accuracy: 0.6521
Epoch 128/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7291 - accuracy: 0.6534
Epoch 129/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7284 - accuracy: 0.6527
Epoch 130/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7286 - accuracy: 0.6540
Epoch 131/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7284 - accuracy: 0.6531
Epoch 132/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7278 - accuracy: 0.6545
Epoch 133/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7268 - accuracy: 0.6557
Epoch 134/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7269 - accuracy: 0.6547
Epoch 135/500
2163/2163 [==============================] - 10s 5ms/step - loss: 0.7268 - accuracy: 0.6556
Epoch 136/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7264 - accuracy: 0.6539
Epoch 137/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7263 - accuracy: 0.6549
Epoch 138/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7256 - accuracy: 0.6562
Epoch 139/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7260 - accuracy: 0.6548
Epoch 140/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7255 - accuracy: 0.6550
Epoch 141/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7247 - accuracy: 0.6556
Epoch 142/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7247 - accuracy: 0.6552
Epoch 143/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7241 - accuracy: 0.6568
Epoch 144/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7239 - accuracy: 0.6553
Epoch 145/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7240 - accuracy: 0.6568
Epoch 146/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7232 - accuracy: 0.6572
Epoch 147/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7228 - accuracy: 0.6561
Epoch 148/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7227 - accuracy: 0.6563
Epoch 149/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7221 - accuracy: 0.6582
Epoch 150/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7221 - accuracy: 0.6578
Epoch 151/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7218 - accuracy: 0.6572
Epoch 152/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7212 - accuracy: 0.6587
Epoch 153/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7210 - accuracy: 0.6567
Epoch 154/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7207 - accuracy: 0.6581
Epoch 155/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7207 - accuracy: 0.6574
Epoch 156/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7198 - accuracy: 0.6590
Epoch 157/500
2163/2163 [==============================] - 10s 5ms/step - loss: 0.7200 - accuracy: 0.6583
Epoch 158/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7195 - accuracy: 0.6584
Epoch 159/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7194 - accuracy: 0.6585
Epoch 160/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7190 - accuracy: 0.6590
Epoch 161/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7190 - accuracy: 0.6588
Epoch 162/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7183 - accuracy: 0.6595
Epoch 163/500
2163/2163 [==============================] - 11s 5ms/step - loss: 0.7181 - accuracy: 0.6592
Epoch 164/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7179 - accuracy: 0.6590
Epoch 165/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7174 - accuracy: 0.6589
Epoch 166/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7172 - accuracy: 0.6618
Epoch 167/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7167 - accuracy: 0.6596
Epoch 168/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7162 - accuracy: 0.6614
Epoch 169/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7169 - accuracy: 0.6605
Epoch 170/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7162 - accuracy: 0.6604
Epoch 171/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7159 - accuracy: 0.6615
Epoch 172/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7157 - accuracy: 0.6609
Epoch 173/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7154 - accuracy: 0.6613
Epoch 174/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7146 - accuracy: 0.6615
Epoch 175/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7148 - accuracy: 0.6621
Epoch 176/500
2163/2163 [==============================] - 14s 6ms/step - loss: 0.7149 - accuracy: 0.6619
Epoch 177/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7145 - accuracy: 0.6617
Epoch 178/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7139 - accuracy: 0.6613
Epoch 179/500
2163/2163 [==============================] - 11s 5ms/step - loss: 0.7143 - accuracy: 0.6618
Epoch 180/500
2163/2163 [==============================] - 10s 5ms/step - loss: 0.7133 - accuracy: 0.6623
Epoch 181/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7130 - accuracy: 0.6626
Epoch 182/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7133 - accuracy: 0.6629
Epoch 183/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7132 - accuracy: 0.6618
Epoch 184/500
2163/2163 [==============================] - 8s 3ms/step - loss: 0.7125 - accuracy: 0.6623
Epoch 185/500
2163/2163 [==============================] - 8s 4ms/step - loss: 0.7117 - accuracy: 0.6624
Epoch 186/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7116 - accuracy: 0.6631
Epoch 187/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7116 - accuracy: 0.6632
Epoch 188/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7110 - accuracy: 0.6625
Epoch 189/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7111 - accuracy: 0.6633
Epoch 190/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7106 - accuracy: 0.6652
Epoch 191/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7106 - accuracy: 0.6637
Epoch 192/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7110 - accuracy: 0.6645
Epoch 193/500
2163/2163 [==============================] - 9s 4ms/step - loss: 0.7103 - accuracy: 0.6636
Epoch 194/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7104 - accuracy: 0.6638
Epoch 195/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7096 - accuracy: 0.6637
Epoch 196/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7093 - accuracy: 0.6648
Epoch 197/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7090 - accuracy: 0.6653
Epoch 198/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7088 - accuracy: 0.6651
Epoch 199/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7083 - accuracy: 0.6641
Epoch 200/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7092 - accuracy: 0.6642
Epoch 201/500
2163/2163 [==============================] - 10s 4ms/step - loss: 0.7083 - accuracy: 0.6647
Epoch 202/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7078 - accuracy: 0.6642
Epoch 203/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7080 - accuracy: 0.6660
Epoch 204/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7070 - accuracy: 0.6650
Epoch 205/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7079 - accuracy: 0.6656
Epoch 206/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7067 - accuracy: 0.6648
Epoch 207/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7068 - accuracy: 0.6662
Epoch 208/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7066 - accuracy: 0.6658
Epoch 209/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7064 - accuracy: 0.6659
Epoch 210/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7061 - accuracy: 0.6674
Epoch 211/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7058 - accuracy: 0.6659
Epoch 212/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7055 - accuracy: 0.6674
Epoch 213/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7059 - accuracy: 0.6661
Epoch 214/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7053 - accuracy: 0.6663
Epoch 215/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7051 - accuracy: 0.6656
Epoch 216/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7045 - accuracy: 0.6665
Epoch 217/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7045 - accuracy: 0.6673
Epoch 218/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7041 - accuracy: 0.6669
Epoch 219/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7042 - accuracy: 0.6678
Epoch 220/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7035 - accuracy: 0.6672
Epoch 221/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7037 - accuracy: 0.6690
Epoch 222/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7039 - accuracy: 0.6665
Epoch 223/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7034 - accuracy: 0.6678
Epoch 224/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7032 - accuracy: 0.6679
Epoch 225/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7026 - accuracy: 0.6680
Epoch 226/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7022 - accuracy: 0.6694
Epoch 227/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7022 - accuracy: 0.6689
Epoch 228/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7021 - accuracy: 0.6689
Epoch 229/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7018 - accuracy: 0.6690
Epoch 230/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7023 - accuracy: 0.6676
Epoch 231/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.7018 - accuracy: 0.6690
Epoch 232/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7017 - accuracy: 0.6685
Epoch 233/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7010 - accuracy: 0.6699
Epoch 234/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.7013 - accuracy: 0.6702
Epoch 235/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.7009 - accuracy: 0.6686
Epoch 236/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7009 - accuracy: 0.6703
Epoch 237/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7002 - accuracy: 0.6709
Epoch 238/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.7002 - accuracy: 0.6696
Epoch 239/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6995 - accuracy: 0.6699
Epoch 240/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6999 - accuracy: 0.6714
Epoch 241/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6996 - accuracy: 0.6702
Epoch 242/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6992 - accuracy: 0.6687
Epoch 243/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6987 - accuracy: 0.6704
Epoch 244/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6987 - accuracy: 0.6710
Epoch 245/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6987 - accuracy: 0.6727
Epoch 246/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6984 - accuracy: 0.6704
Epoch 247/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6975 - accuracy: 0.6696
Epoch 248/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6983 - accuracy: 0.6708
Epoch 249/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6981 - accuracy: 0.6694
Epoch 250/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6975 - accuracy: 0.6701
Epoch 251/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6972 - accuracy: 0.6716
Epoch 252/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6968 - accuracy: 0.6701
Epoch 253/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6966 - accuracy: 0.6720
Epoch 254/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6966 - accuracy: 0.6713
Epoch 255/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6963 - accuracy: 0.6717
Epoch 256/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6965 - accuracy: 0.6720
Epoch 257/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6965 - accuracy: 0.6723
Epoch 258/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6958 - accuracy: 0.6715
Epoch 259/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6957 - accuracy: 0.6721
Epoch 260/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6953 - accuracy: 0.6721
Epoch 261/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6948 - accuracy: 0.6717
Epoch 262/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6955 - accuracy: 0.6709
Epoch 263/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6949 - accuracy: 0.6721
Epoch 264/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6946 - accuracy: 0.6722
Epoch 265/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6941 - accuracy: 0.6726
Epoch 266/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6952 - accuracy: 0.6718
Epoch 267/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6942 - accuracy: 0.6714
Epoch 268/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6939 - accuracy: 0.6738
Epoch 269/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6941 - accuracy: 0.6729
Epoch 270/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6929 - accuracy: 0.6732
Epoch 271/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6935 - accuracy: 0.6725
Epoch 272/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6935 - accuracy: 0.6726
Epoch 273/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6934 - accuracy: 0.6737
Epoch 274/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6929 - accuracy: 0.6731
Epoch 275/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6935 - accuracy: 0.6720
Epoch 276/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6933 - accuracy: 0.6746
Epoch 277/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6925 - accuracy: 0.6734
Epoch 278/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6925 - accuracy: 0.6727
Epoch 279/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6920 - accuracy: 0.6735
Epoch 280/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6919 - accuracy: 0.6739
Epoch 281/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6921 - accuracy: 0.6742
Epoch 282/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6912 - accuracy: 0.6749
Epoch 283/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6913 - accuracy: 0.6745
Epoch 284/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6912 - accuracy: 0.6746
Epoch 285/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6909 - accuracy: 0.6741
Epoch 286/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.6912 - accuracy: 0.6745
Epoch 287/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6906 - accuracy: 0.6759
Epoch 288/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6912 - accuracy: 0.6751
Epoch 289/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6904 - accuracy: 0.6753
Epoch 290/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6908 - accuracy: 0.6752
Epoch 291/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6897 - accuracy: 0.6773
Epoch 292/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6901 - accuracy: 0.6756
Epoch 293/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6900 - accuracy: 0.6750
Epoch 294/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6901 - accuracy: 0.6745
Epoch 295/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6896 - accuracy: 0.6754
Epoch 296/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6892 - accuracy: 0.6749
Epoch 297/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6896 - accuracy: 0.6752
Epoch 298/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6890 - accuracy: 0.6763
Epoch 299/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6894 - accuracy: 0.6761
Epoch 300/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6883 - accuracy: 0.6756
Epoch 301/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6888 - accuracy: 0.6750
Epoch 302/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6885 - accuracy: 0.6765
Epoch 303/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6884 - accuracy: 0.6765
Epoch 304/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6884 - accuracy: 0.6774
Epoch 305/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6885 - accuracy: 0.6748
Epoch 306/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6878 - accuracy: 0.6760
Epoch 307/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6877 - accuracy: 0.6764
Epoch 308/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6879 - accuracy: 0.6751
Epoch 309/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6878 - accuracy: 0.6758
Epoch 310/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6877 - accuracy: 0.6756
Epoch 311/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6872 - accuracy: 0.6763
Epoch 312/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6867 - accuracy: 0.6769
Epoch 313/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6867 - accuracy: 0.6774
Epoch 314/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6868 - accuracy: 0.6758
Epoch 315/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6865 - accuracy: 0.6760
Epoch 316/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6865 - accuracy: 0.6773
Epoch 317/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6857 - accuracy: 0.6791
Epoch 318/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6858 - accuracy: 0.6767
Epoch 319/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6859 - accuracy: 0.6764
Epoch 320/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6862 - accuracy: 0.6767
Epoch 321/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6849 - accuracy: 0.6777
Epoch 322/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6857 - accuracy: 0.6772
Epoch 323/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6849 - accuracy: 0.6791
Epoch 324/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6854 - accuracy: 0.6776
Epoch 325/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6846 - accuracy: 0.6770
Epoch 326/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6847 - accuracy: 0.6779
Epoch 327/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6850 - accuracy: 0.6784
Epoch 328/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6838 - accuracy: 0.6775
Epoch 329/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6842 - accuracy: 0.6776
Epoch 330/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6847 - accuracy: 0.6769
Epoch 331/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6843 - accuracy: 0.6783
Epoch 332/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6838 - accuracy: 0.6777
Epoch 333/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6841 - accuracy: 0.6768
Epoch 334/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6833 - accuracy: 0.6781
Epoch 335/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6832 - accuracy: 0.6793
Epoch 336/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6831 - accuracy: 0.6790
Epoch 337/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6834 - accuracy: 0.6770
Epoch 338/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6832 - accuracy: 0.6793
Epoch 339/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6829 - accuracy: 0.6781
Epoch 340/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6828 - accuracy: 0.6788
Epoch 341/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6824 - accuracy: 0.6796
Epoch 342/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6824 - accuracy: 0.6787
Epoch 343/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6823 - accuracy: 0.6780
Epoch 344/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6823 - accuracy: 0.6793
Epoch 345/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6822 - accuracy: 0.6781
Epoch 346/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6818 - accuracy: 0.6797
Epoch 347/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6819 - accuracy: 0.6804
Epoch 348/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6818 - accuracy: 0.6790
Epoch 349/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6815 - accuracy: 0.6794
Epoch 350/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6815 - accuracy: 0.6791
Epoch 351/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6816 - accuracy: 0.6791
Epoch 352/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6812 - accuracy: 0.6791
Epoch 353/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6811 - accuracy: 0.6801
Epoch 354/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6804 - accuracy: 0.6813
Epoch 355/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6810 - accuracy: 0.6796
Epoch 356/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6805 - accuracy: 0.6801
Epoch 357/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6806 - accuracy: 0.6789
Epoch 358/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6800 - accuracy: 0.6807
Epoch 359/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6795 - accuracy: 0.6822
Epoch 360/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6798 - accuracy: 0.6804
Epoch 361/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6794 - accuracy: 0.6795
Epoch 362/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6799 - accuracy: 0.6819
Epoch 363/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6791 - accuracy: 0.6802
Epoch 364/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6794 - accuracy: 0.6809
Epoch 365/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6790 - accuracy: 0.6801
Epoch 366/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6791 - accuracy: 0.6808
Epoch 367/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6788 - accuracy: 0.6809
Epoch 368/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6790 - accuracy: 0.6809
Epoch 369/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6787 - accuracy: 0.6813
Epoch 370/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6785 - accuracy: 0.6804
Epoch 371/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6785 - accuracy: 0.6828
Epoch 372/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6784 - accuracy: 0.6819
Epoch 373/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6779 - accuracy: 0.6818
Epoch 374/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6779 - accuracy: 0.6792
Epoch 375/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6786 - accuracy: 0.6813
Epoch 376/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6773 - accuracy: 0.6827
Epoch 377/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6781 - accuracy: 0.6803
Epoch 378/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6772 - accuracy: 0.6824
Epoch 379/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6775 - accuracy: 0.6816
Epoch 380/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6776 - accuracy: 0.6816
Epoch 381/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6767 - accuracy: 0.6816
Epoch 382/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6774 - accuracy: 0.6821
Epoch 383/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6767 - accuracy: 0.6818
Epoch 384/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6765 - accuracy: 0.6833
Epoch 385/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6764 - accuracy: 0.6825
Epoch 386/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6768 - accuracy: 0.6822
Epoch 387/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6758 - accuracy: 0.6829
Epoch 388/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6767 - accuracy: 0.6834
Epoch 389/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6770 - accuracy: 0.6816
Epoch 390/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6758 - accuracy: 0.6819
Epoch 391/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6763 - accuracy: 0.6830
Epoch 392/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6759 - accuracy: 0.6827
Epoch 393/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6756 - accuracy: 0.6832
Epoch 394/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6759 - accuracy: 0.6827
Epoch 395/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6749 - accuracy: 0.6827
Epoch 396/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6758 - accuracy: 0.6826
Epoch 397/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6755 - accuracy: 0.6821
Epoch 398/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6750 - accuracy: 0.6832
Epoch 399/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6750 - accuracy: 0.6839
Epoch 400/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6745 - accuracy: 0.6839
Epoch 401/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6752 - accuracy: 0.6830
Epoch 402/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6743 - accuracy: 0.6841
Epoch 403/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6743 - accuracy: 0.6830
Epoch 404/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6745 - accuracy: 0.6836
Epoch 405/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6741 - accuracy: 0.6839
Epoch 406/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6740 - accuracy: 0.6834
Epoch 407/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6742 - accuracy: 0.6832
Epoch 408/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6740 - accuracy: 0.6833
Epoch 409/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6736 - accuracy: 0.6857
Epoch 410/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6737 - accuracy: 0.6838
Epoch 411/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6734 - accuracy: 0.6841
Epoch 412/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6733 - accuracy: 0.6833
Epoch 413/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6735 - accuracy: 0.6830
Epoch 414/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6729 - accuracy: 0.6847
Epoch 415/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6730 - accuracy: 0.6838
Epoch 416/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6726 - accuracy: 0.6849
Epoch 417/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6727 - accuracy: 0.6844
Epoch 418/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6720 - accuracy: 0.6857
Epoch 419/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6726 - accuracy: 0.6845
Epoch 420/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6723 - accuracy: 0.6857
Epoch 421/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6724 - accuracy: 0.6845
Epoch 422/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6726 - accuracy: 0.6837
Epoch 423/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6719 - accuracy: 0.6846
Epoch 424/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6719 - accuracy: 0.6838
Epoch 425/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6711 - accuracy: 0.6848
Epoch 426/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6723 - accuracy: 0.6850
Epoch 427/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6713 - accuracy: 0.6837
Epoch 428/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6720 - accuracy: 0.6838
Epoch 429/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6717 - accuracy: 0.6851
Epoch 430/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6710 - accuracy: 0.6857
Epoch 431/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6714 - accuracy: 0.6851
Epoch 432/500
2163/2163 [==============================] - 5s 2ms/step - loss: 0.6716 - accuracy: 0.6856
Epoch 433/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6707 - accuracy: 0.6850
Epoch 434/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6709 - accuracy: 0.6852
Epoch 435/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6711 - accuracy: 0.6860
Epoch 436/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6704 - accuracy: 0.6852
Epoch 437/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6706 - accuracy: 0.6852
Epoch 438/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6697 - accuracy: 0.6857
Epoch 439/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6705 - accuracy: 0.6858
Epoch 440/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6702 - accuracy: 0.6856
Epoch 441/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6696 - accuracy: 0.6872
Epoch 442/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6702 - accuracy: 0.6858
Epoch 443/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6691 - accuracy: 0.6857
Epoch 444/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6705 - accuracy: 0.6860
Epoch 445/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6701 - accuracy: 0.6863
Epoch 446/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6692 - accuracy: 0.6867
Epoch 447/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6692 - accuracy: 0.6864
Epoch 448/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6695 - accuracy: 0.6855
Epoch 449/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6689 - accuracy: 0.6873
Epoch 450/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6688 - accuracy: 0.6865
Epoch 451/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6693 - accuracy: 0.6848
Epoch 452/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6685 - accuracy: 0.6872
Epoch 453/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6685 - accuracy: 0.6881
Epoch 454/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6686 - accuracy: 0.6864
Epoch 455/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6680 - accuracy: 0.6879
Epoch 456/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6679 - accuracy: 0.6886
Epoch 457/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6688 - accuracy: 0.6873
Epoch 458/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6676 - accuracy: 0.6874
Epoch 459/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6678 - accuracy: 0.6862
Epoch 460/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6683 - accuracy: 0.6874
Epoch 461/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6669 - accuracy: 0.6890
Epoch 462/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6675 - accuracy: 0.6885
Epoch 463/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6682 - accuracy: 0.6871
Epoch 464/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6680 - accuracy: 0.6874
Epoch 465/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6667 - accuracy: 0.6877
Epoch 466/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6671 - accuracy: 0.6885
Epoch 467/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6672 - accuracy: 0.6870
Epoch 468/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6665 - accuracy: 0.6880
Epoch 469/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6674 - accuracy: 0.6875
Epoch 470/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6668 - accuracy: 0.6869
Epoch 471/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6663 - accuracy: 0.6894
Epoch 472/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6664 - accuracy: 0.6892
Epoch 473/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6664 - accuracy: 0.6890
Epoch 474/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6668 - accuracy: 0.6880
Epoch 475/500
2163/2163 [==============================] - 5s 3ms/step - loss: 0.6668 - accuracy: 0.6878
Epoch 476/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6663 - accuracy: 0.6897
Epoch 477/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6659 - accuracy: 0.6891
Epoch 478/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6664 - accuracy: 0.6873
Epoch 479/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6656 - accuracy: 0.6889
Epoch 480/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6657 - accuracy: 0.6884
Epoch 481/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6655 - accuracy: 0.6890
Epoch 482/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6659 - accuracy: 0.6888
Epoch 483/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6652 - accuracy: 0.6892
Epoch 484/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6656 - accuracy: 0.6889
Epoch 485/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6662 - accuracy: 0.6884
Epoch 486/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6655 - accuracy: 0.6876
Epoch 487/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6647 - accuracy: 0.6893
Epoch 488/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6652 - accuracy: 0.6893
Epoch 489/500
2163/2163 [==============================] - 7s 3ms/step - loss: 0.6651 - accuracy: 0.6890
Epoch 490/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6645 - accuracy: 0.6903
Epoch 491/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6644 - accuracy: 0.6901
Epoch 492/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6646 - accuracy: 0.6901
Epoch 493/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6644 - accuracy: 0.6898
Epoch 494/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6641 - accuracy: 0.6904
Epoch 495/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6643 - accuracy: 0.6903
Epoch 496/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6641 - accuracy: 0.6884
Epoch 497/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6646 - accuracy: 0.6899
Epoch 498/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6640 - accuracy: 0.6892
Epoch 499/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6641 - accuracy: 0.6902
Epoch 500/500
2163/2163 [==============================] - 6s 3ms/step - loss: 0.6639 - accuracy: 0.6909

<keras.callbacks.History at 0x7f5bf53a63a0>

Generamos las predicciones para train y vemos cómo performa.

In [60]:

def clasificar_propiedad(model, entrada):
  prediccion = model.predict(entrada)
  arg_max = np.argmax(prediccion, axis=1)
  return arg_max

In [61]:

y_pred = clasificar_propiedad(best_model, x_train_clasificacion)
graficar_matriz_de_confusion(y_train_clasificacion, y_pred, outputs.keys())
imprimir_metricas_de_clasificacion(y_train_clasificacion, y_pred)

Out[61]:

2163/2163 [==============================] - 3s 1ms/step

Métricas de clasificación

Accuracy: 0.6928990491604289
Recall: 0.6928990491604289
F1 score: 0.6931316050718525

Vemos cómo performa con los datos de test.

In [62]:

y_pred = clasificar_propiedad(best_model, x_test_clasificacion)
graficar_matriz_de_confusion(y_test_clasificacion, y_pred, outputs.keys())
imprimir_metricas_de_clasificacion(y_test_clasificacion, y_pred)

Out[62]:

540/540 [==============================] - 1s 1ms/step

Métricas de clasificación

Accuracy: 0.6342565766600997
Recall: 0.6342565766600997
F1 score: 0.6340227995538984

Regresión

Construimos el modelo.

In [63]:

d_in = x_train_regresion.shape[1]
d_out = 1

In [66]:

def build_model(hp):
  model = keras.Sequential()
  model.add(keras.layers.Dense(
      hp.Int('units', min_value=32, max_value=512, step=16),
      activation=hp.Choice("activation", ["relu", "tanh"]), input_shape=(d_in,)))
  model.add(Dense(128, activation='relu'))
  model.add(Dense(32, activation='relu'))
  model.add(Dense(d_out))
  learning_rate = hp.Float("lr", min_value=1e-4, max_value=1e-1, sampling="log")

  model.compile(optimizer=keras.optimizers.Adam(learning_rate=learning_rate), 
  loss = 'mse', 
  metrics = ['mse'],)
  return model

Buscamos los mejores hiperparámetros.

In [67]:

tuner = kt.RandomSearch(
    build_model,
    objective='loss',
    max_trials=10,
    project_name="Regresion",
    overwrite=True)

In [68]:

stop_early = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=1)
tuner.search(x_train_clasificacion, y_train_clasificacion, epochs=100, validation_split=0.2, callbacks=[stop_early])
best_model = tuner.get_best_models()[0]

Out[68]:

Trial 10 Complete [00h 00m 29s]
loss: 0.5874912738800049

Best loss So Far: 0.587153971195221
Total elapsed time: 00h 03m 50s

Imprimimos los mejores hiperparámetros encontrados y entrenamos nuevamente al modelo.

In [69]:

best_hps = tuner.get_best_hyperparameters()[0]
best_hps.values

Out[69]:

{'units': 224, 'activation': 'relu', 'lr': 0.00010700549249923322}

In [70]:

best_model = build_model(best_hps)
best_model.fit(x=x_train_regresion, y=y_train_regresion, epochs=500)

Out[70]:

WARNING:tensorflow:Detecting that an object or model or tf.train.Checkpoint is being deleted with unrestored values. See the following logs for the specific values in question. To silence these warnings, use `status.expect_partial()`. See https://www.tensorflow.org/api_docs/python/tf/train/Checkpoint#restorefor details about the status object returned by the restore function.
WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer.iter
WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer.beta_1
WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer.beta_2
WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer.decay
WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer.learning_rate

Epoch 1/500
2163/2163 [==============================] - 6s 3ms/step - loss: 32732413952.0000 - mse: 32732413952.0000
Epoch 2/500
2163/2163 [==============================] - 5s 2ms/step - loss: 6488169984.0000 - mse: 6488169984.0000
Epoch 3/500
2163/2163 [==============================] - 6s 3ms/step - loss: 3065523968.0000 - mse: 3065523968.0000
Epoch 4/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2878243584.0000 - mse: 2878243584.0000
Epoch 5/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2804150016.0000 - mse: 2804150016.0000
Epoch 6/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2758620160.0000 - mse: 2758620160.0000
Epoch 7/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2727923712.0000 - mse: 2727923712.0000
Epoch 8/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2705176320.0000 - mse: 2705176320.0000
Epoch 9/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2687800832.0000 - mse: 2687800832.0000
Epoch 10/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2673324288.0000 - mse: 2673324288.0000
Epoch 11/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2661199616.0000 - mse: 2661199616.0000
Epoch 12/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2651013888.0000 - mse: 2651013888.0000
Epoch 13/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2642069248.0000 - mse: 2642069248.0000
Epoch 14/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2634811392.0000 - mse: 2634811392.0000
Epoch 15/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2628414976.0000 - mse: 2628414976.0000
Epoch 16/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2622315264.0000 - mse: 2622315264.0000
Epoch 17/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2616353280.0000 - mse: 2616353280.0000
Epoch 18/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2611685120.0000 - mse: 2611685120.0000
Epoch 19/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2606664448.0000 - mse: 2606664448.0000
Epoch 20/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2603040000.0000 - mse: 2603040000.0000
Epoch 21/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2598931968.0000 - mse: 2598931968.0000
Epoch 22/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2595731712.0000 - mse: 2595731712.0000
Epoch 23/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2592449280.0000 - mse: 2592449280.0000
Epoch 24/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2589044224.0000 - mse: 2589044224.0000
Epoch 25/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2586116608.0000 - mse: 2586116608.0000
Epoch 26/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2583699712.0000 - mse: 2583699712.0000
Epoch 27/500
2163/2163 [==============================] - 5s 2ms/step - loss: 2580941568.0000 - mse: 2580941568.0000
Epoch 28/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2579114752.0000 - mse: 2579114752.0000
Epoch 29/500
2163/2163 [==============================] - 5s 3ms/step - loss: 2576287488.0000 - mse: 2576287488.0000
Epoch 30/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2574503680.0000 - mse: 2574503680.0000
Epoch 31/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2571964416.0000 - mse: 2571964416.0000
Epoch 32/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2569635072.0000 - mse: 2569635072.0000
Epoch 33/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2568297984.0000 - mse: 2568297984.0000
Epoch 34/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2566500864.0000 - mse: 2566500864.0000
Epoch 35/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2564893952.0000 - mse: 2564893952.0000
Epoch 36/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2563148032.0000 - mse: 2563148032.0000
Epoch 37/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2561884416.0000 - mse: 2561884416.0000
Epoch 38/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2559502848.0000 - mse: 2559502848.0000
Epoch 39/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2558888448.0000 - mse: 2558888448.0000
Epoch 40/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2557254400.0000 - mse: 2557254400.0000
Epoch 41/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2555282944.0000 - mse: 2555282944.0000
Epoch 42/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2554207232.0000 - mse: 2554207232.0000
Epoch 43/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2552147968.0000 - mse: 2552147968.0000
Epoch 44/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2551307264.0000 - mse: 2551307264.0000
Epoch 45/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2549370112.0000 - mse: 2549370112.0000
Epoch 46/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2548667648.0000 - mse: 2548667648.0000
Epoch 47/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2547274496.0000 - mse: 2547274496.0000
Epoch 48/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2545530624.0000 - mse: 2545530624.0000
Epoch 49/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2544046080.0000 - mse: 2544046080.0000
Epoch 50/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2543001088.0000 - mse: 2543001088.0000
Epoch 51/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2541800704.0000 - mse: 2541800704.0000
Epoch 52/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2540523264.0000 - mse: 2540523264.0000
Epoch 53/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2539549184.0000 - mse: 2539549184.0000
Epoch 54/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2537867008.0000 - mse: 2537867008.0000
Epoch 55/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2536996864.0000 - mse: 2536996864.0000
Epoch 56/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2535478784.0000 - mse: 2535478784.0000
Epoch 57/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2533908736.0000 - mse: 2533908736.0000
Epoch 58/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2533269248.0000 - mse: 2533269248.0000
Epoch 59/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2531567872.0000 - mse: 2531567872.0000
Epoch 60/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2530421504.0000 - mse: 2530421504.0000
Epoch 61/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2528589312.0000 - mse: 2528589312.0000
Epoch 62/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2527364608.0000 - mse: 2527364608.0000
Epoch 63/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2526354688.0000 - mse: 2526354688.0000
Epoch 64/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2525059840.0000 - mse: 2525059840.0000
Epoch 65/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2523678208.0000 - mse: 2523678208.0000
Epoch 66/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2522267392.0000 - mse: 2522267392.0000
Epoch 67/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2521103872.0000 - mse: 2521103872.0000
Epoch 68/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2519354368.0000 - mse: 2519354368.0000
Epoch 69/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2517817088.0000 - mse: 2517817088.0000
Epoch 70/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2516071424.0000 - mse: 2516071424.0000
Epoch 71/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2514662656.0000 - mse: 2514662656.0000
Epoch 72/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2513143552.0000 - mse: 2513143552.0000
Epoch 73/500
2163/2163 [==============================] - 8s 4ms/step - loss: 2511861760.0000 - mse: 2511861760.0000
Epoch 74/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2510546688.0000 - mse: 2510546688.0000
Epoch 75/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2508431360.0000 - mse: 2508431360.0000
Epoch 76/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2506638080.0000 - mse: 2506638080.0000
Epoch 77/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2504494336.0000 - mse: 2504494336.0000
Epoch 78/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2502560256.0000 - mse: 2502560256.0000
Epoch 79/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2501280000.0000 - mse: 2501280000.0000
Epoch 80/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2499499264.0000 - mse: 2499499264.0000
Epoch 81/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2497174272.0000 - mse: 2497174272.0000
Epoch 82/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2495312128.0000 - mse: 2495312128.0000
Epoch 83/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2492897024.0000 - mse: 2492897024.0000
Epoch 84/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2490497792.0000 - mse: 2490497792.0000
Epoch 85/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2488095744.0000 - mse: 2488095744.0000
Epoch 86/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2485402112.0000 - mse: 2485402112.0000
Epoch 87/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2483392512.0000 - mse: 2483392512.0000
Epoch 88/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2480520448.0000 - mse: 2480520448.0000
Epoch 89/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2477408768.0000 - mse: 2477408768.0000
Epoch 90/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2475335680.0000 - mse: 2475335680.0000
Epoch 91/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2472004864.0000 - mse: 2472004864.0000
Epoch 92/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2470129408.0000 - mse: 2470129408.0000
Epoch 93/500
2163/2163 [==============================] - 8s 4ms/step - loss: 2466523648.0000 - mse: 2466523648.0000
Epoch 94/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2463677696.0000 - mse: 2463677696.0000
Epoch 95/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2459663360.0000 - mse: 2459663360.0000
Epoch 96/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2457854976.0000 - mse: 2457854976.0000
Epoch 97/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2454249472.0000 - mse: 2454249472.0000
Epoch 98/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2451349248.0000 - mse: 2451349248.0000
Epoch 99/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2447956480.0000 - mse: 2447956480.0000
Epoch 100/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2444744192.0000 - mse: 2444744192.0000
Epoch 101/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2441849344.0000 - mse: 2441849344.0000
Epoch 102/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2438049536.0000 - mse: 2438049536.0000
Epoch 103/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2435214848.0000 - mse: 2435214848.0000
Epoch 104/500
2163/2163 [==============================] - 8s 4ms/step - loss: 2431680512.0000 - mse: 2431680512.0000
Epoch 105/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2428575488.0000 - mse: 2428575488.0000
Epoch 106/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2424966144.0000 - mse: 2424966144.0000
Epoch 107/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2421763328.0000 - mse: 2421763328.0000
Epoch 108/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2418650368.0000 - mse: 2418650368.0000
Epoch 109/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2415320576.0000 - mse: 2415320576.0000
Epoch 110/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2411934720.0000 - mse: 2411934720.0000
Epoch 111/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2408559616.0000 - mse: 2408559616.0000
Epoch 112/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2405138176.0000 - mse: 2405138176.0000
Epoch 113/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2401988352.0000 - mse: 2401988352.0000
Epoch 114/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2398920192.0000 - mse: 2398920192.0000
Epoch 115/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2396034816.0000 - mse: 2396034816.0000
Epoch 116/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2392830976.0000 - mse: 2392830976.0000
Epoch 117/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2390137856.0000 - mse: 2390137856.0000
Epoch 118/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2387352064.0000 - mse: 2387352064.0000
Epoch 119/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2384671744.0000 - mse: 2384671744.0000
Epoch 120/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2381797120.0000 - mse: 2381797120.0000
Epoch 121/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2378999808.0000 - mse: 2378999808.0000
Epoch 122/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2376595968.0000 - mse: 2376595968.0000
Epoch 123/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2373659648.0000 - mse: 2373659648.0000
Epoch 124/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2371530752.0000 - mse: 2371530752.0000
Epoch 125/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2368811008.0000 - mse: 2368811008.0000
Epoch 126/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2366212352.0000 - mse: 2366212352.0000
Epoch 127/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2362543360.0000 - mse: 2362543360.0000
Epoch 128/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2361534976.0000 - mse: 2361534976.0000
Epoch 129/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2358573056.0000 - mse: 2358573056.0000
Epoch 130/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2356213504.0000 - mse: 2356213504.0000
Epoch 131/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2353868800.0000 - mse: 2353868800.0000
Epoch 132/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2351515904.0000 - mse: 2351515904.0000
Epoch 133/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2349476352.0000 - mse: 2349476352.0000
Epoch 134/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2346836480.0000 - mse: 2346836480.0000
Epoch 135/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2345253888.0000 - mse: 2345253888.0000
Epoch 136/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2343099648.0000 - mse: 2343099648.0000
Epoch 137/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2341102592.0000 - mse: 2341102592.0000
Epoch 138/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2338604544.0000 - mse: 2338604544.0000
Epoch 139/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2336492800.0000 - mse: 2336492800.0000
Epoch 140/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2334276608.0000 - mse: 2334276608.0000
Epoch 141/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2332720640.0000 - mse: 2332720640.0000
Epoch 142/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2330574336.0000 - mse: 2330574336.0000
Epoch 143/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2328431104.0000 - mse: 2328431104.0000
Epoch 144/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2325844992.0000 - mse: 2325844992.0000
Epoch 145/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2323933696.0000 - mse: 2323933696.0000
Epoch 146/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2323211776.0000 - mse: 2323211776.0000
Epoch 147/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2321231616.0000 - mse: 2321231616.0000
Epoch 148/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2319376640.0000 - mse: 2319376640.0000
Epoch 149/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2317438208.0000 - mse: 2317438208.0000
Epoch 150/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2316134144.0000 - mse: 2316134144.0000
Epoch 151/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2314102272.0000 - mse: 2314102272.0000
Epoch 152/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2311626240.0000 - mse: 2311626240.0000
Epoch 153/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2310930688.0000 - mse: 2310930688.0000
Epoch 154/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2308738048.0000 - mse: 2308738048.0000
Epoch 155/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2307095040.0000 - mse: 2307095040.0000
Epoch 156/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2305234688.0000 - mse: 2305234688.0000
Epoch 157/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2303770112.0000 - mse: 2303770112.0000
Epoch 158/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2301957632.0000 - mse: 2301957632.0000
Epoch 159/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2300369664.0000 - mse: 2300369664.0000
Epoch 160/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2298743040.0000 - mse: 2298743040.0000
Epoch 161/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2297137664.0000 - mse: 2297137664.0000
Epoch 162/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2295391488.0000 - mse: 2295391488.0000
Epoch 163/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2293946624.0000 - mse: 2293946624.0000
Epoch 164/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2291612928.0000 - mse: 2291612928.0000
Epoch 165/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2289646592.0000 - mse: 2289646592.0000
Epoch 166/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2288915200.0000 - mse: 2288915200.0000
Epoch 167/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2287650816.0000 - mse: 2287650816.0000
Epoch 168/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2285879040.0000 - mse: 2285879040.0000
Epoch 169/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2283743744.0000 - mse: 2283743744.0000
Epoch 170/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2282806016.0000 - mse: 2282806016.0000
Epoch 171/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2281210368.0000 - mse: 2281210368.0000
Epoch 172/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2279241984.0000 - mse: 2279241984.0000
Epoch 173/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2277723136.0000 - mse: 2277723136.0000
Epoch 174/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2275785984.0000 - mse: 2275785984.0000
Epoch 175/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2274328832.0000 - mse: 2274328832.0000
Epoch 176/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2273187072.0000 - mse: 2273187072.0000
Epoch 177/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2270886656.0000 - mse: 2270886656.0000
Epoch 178/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2269403648.0000 - mse: 2269403648.0000
Epoch 179/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2267573248.0000 - mse: 2267573248.0000
Epoch 180/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2266040320.0000 - mse: 2266040320.0000
Epoch 181/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2264138240.0000 - mse: 2264138240.0000
Epoch 182/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2262481920.0000 - mse: 2262481920.0000
Epoch 183/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2260829952.0000 - mse: 2260829952.0000
Epoch 184/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2258974464.0000 - mse: 2258974464.0000
Epoch 185/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2256883200.0000 - mse: 2256883200.0000
Epoch 186/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2255362304.0000 - mse: 2255362304.0000
Epoch 187/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2252989440.0000 - mse: 2252989440.0000
Epoch 188/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2251091968.0000 - mse: 2251091968.0000
Epoch 189/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2249382656.0000 - mse: 2249382656.0000
Epoch 190/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2247369728.0000 - mse: 2247369728.0000
Epoch 191/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2244731392.0000 - mse: 2244731392.0000
Epoch 192/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2243359232.0000 - mse: 2243359232.0000
Epoch 193/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2240570624.0000 - mse: 2240570624.0000
Epoch 194/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2239549952.0000 - mse: 2239549952.0000
Epoch 195/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2237370112.0000 - mse: 2237370112.0000
Epoch 196/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2235218944.0000 - mse: 2235218944.0000
Epoch 197/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2232405760.0000 - mse: 2232405760.0000
Epoch 198/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2230899968.0000 - mse: 2230899968.0000
Epoch 199/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2229107968.0000 - mse: 2229107968.0000
Epoch 200/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2227221504.0000 - mse: 2227221504.0000
Epoch 201/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2224955904.0000 - mse: 2224955904.0000
Epoch 202/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2222689024.0000 - mse: 2222689024.0000
Epoch 203/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2220928512.0000 - mse: 2220928512.0000
Epoch 204/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2218331392.0000 - mse: 2218331392.0000
Epoch 205/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2216781568.0000 - mse: 2216781568.0000
Epoch 206/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2214934528.0000 - mse: 2214934528.0000
Epoch 207/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2212673792.0000 - mse: 2212673792.0000
Epoch 208/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2210515712.0000 - mse: 2210515712.0000
Epoch 209/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2208527360.0000 - mse: 2208527360.0000
Epoch 210/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2206739456.0000 - mse: 2206739456.0000
Epoch 211/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2204417536.0000 - mse: 2204417536.0000
Epoch 212/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2202453504.0000 - mse: 2202453504.0000
Epoch 213/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2199995648.0000 - mse: 2199995648.0000
Epoch 214/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2197742336.0000 - mse: 2197742336.0000
Epoch 215/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2196084224.0000 - mse: 2196084224.0000
Epoch 216/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2193434880.0000 - mse: 2193434880.0000
Epoch 217/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2190644992.0000 - mse: 2190644992.0000
Epoch 218/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2188349696.0000 - mse: 2188349696.0000
Epoch 219/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2185961728.0000 - mse: 2185961728.0000
Epoch 220/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2183688960.0000 - mse: 2183688960.0000
Epoch 221/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2181942016.0000 - mse: 2181942016.0000
Epoch 222/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2178946560.0000 - mse: 2178946560.0000
Epoch 223/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2176656384.0000 - mse: 2176656384.0000
Epoch 224/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2174544384.0000 - mse: 2174544384.0000
Epoch 225/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2172067840.0000 - mse: 2172067840.0000
Epoch 226/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2169718272.0000 - mse: 2169718272.0000
Epoch 227/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2167760896.0000 - mse: 2167760896.0000
Epoch 228/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2165219584.0000 - mse: 2165219584.0000
Epoch 229/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2162918656.0000 - mse: 2162918656.0000
Epoch 230/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2160761088.0000 - mse: 2160761088.0000
Epoch 231/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2157750784.0000 - mse: 2157750784.0000
Epoch 232/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2155457024.0000 - mse: 2155457024.0000
Epoch 233/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2153140480.0000 - mse: 2153140480.0000
Epoch 234/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2151234304.0000 - mse: 2151234304.0000
Epoch 235/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2149476608.0000 - mse: 2149476608.0000
Epoch 236/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2147162880.0000 - mse: 2147162880.0000
Epoch 237/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2145323520.0000 - mse: 2145323520.0000
Epoch 238/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2143689216.0000 - mse: 2143689216.0000
Epoch 239/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2141670912.0000 - mse: 2141670912.0000
Epoch 240/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2140091264.0000 - mse: 2140091264.0000
Epoch 241/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2138448256.0000 - mse: 2138448256.0000
Epoch 242/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2136956672.0000 - mse: 2136956672.0000
Epoch 243/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2135331968.0000 - mse: 2135331968.0000
Epoch 244/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2133418368.0000 - mse: 2133418368.0000
Epoch 245/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2132027008.0000 - mse: 2132027008.0000
Epoch 246/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2130913664.0000 - mse: 2130913664.0000
Epoch 247/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2128594560.0000 - mse: 2128594560.0000
Epoch 248/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2127545472.0000 - mse: 2127545472.0000
Epoch 249/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2125839232.0000 - mse: 2125839232.0000
Epoch 250/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2124123520.0000 - mse: 2124123520.0000
Epoch 251/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2122974080.0000 - mse: 2122974080.0000
Epoch 252/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2121911296.0000 - mse: 2121911296.0000
Epoch 253/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2120468480.0000 - mse: 2120468480.0000
Epoch 254/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2118862336.0000 - mse: 2118862336.0000
Epoch 255/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2117618176.0000 - mse: 2117618176.0000
Epoch 256/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2117063936.0000 - mse: 2117063936.0000
Epoch 257/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2115476480.0000 - mse: 2115476480.0000
Epoch 258/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2114498688.0000 - mse: 2114498688.0000
Epoch 259/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2113769984.0000 - mse: 2113769984.0000
Epoch 260/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2112615936.0000 - mse: 2112615936.0000
Epoch 261/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2111396224.0000 - mse: 2111396224.0000
Epoch 262/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2110364160.0000 - mse: 2110364160.0000
Epoch 263/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2109166080.0000 - mse: 2109166080.0000
Epoch 264/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2108489600.0000 - mse: 2108489600.0000
Epoch 265/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2107309696.0000 - mse: 2107309696.0000
Epoch 266/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2106131200.0000 - mse: 2106131200.0000
Epoch 267/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2104596608.0000 - mse: 2104596608.0000
Epoch 268/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2104349696.0000 - mse: 2104349696.0000
Epoch 269/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2103064320.0000 - mse: 2103064320.0000
Epoch 270/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2102227456.0000 - mse: 2102227456.0000
Epoch 271/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2101116160.0000 - mse: 2101116160.0000
Epoch 272/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2100165248.0000 - mse: 2100165248.0000
Epoch 273/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2099200640.0000 - mse: 2099200640.0000
Epoch 274/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2097852928.0000 - mse: 2097852928.0000
Epoch 275/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2097716864.0000 - mse: 2097716864.0000
Epoch 276/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2095800832.0000 - mse: 2095800832.0000
Epoch 277/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2095388288.0000 - mse: 2095388288.0000
Epoch 278/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2094837888.0000 - mse: 2094837888.0000
Epoch 279/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2094057216.0000 - mse: 2094057216.0000
Epoch 280/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2092701312.0000 - mse: 2092701312.0000
Epoch 281/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2091678208.0000 - mse: 2091678208.0000
Epoch 282/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2091098752.0000 - mse: 2091098752.0000
Epoch 283/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2089991808.0000 - mse: 2089991808.0000
Epoch 284/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2088961920.0000 - mse: 2088961920.0000
Epoch 285/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2088597120.0000 - mse: 2088597120.0000
Epoch 286/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2087990784.0000 - mse: 2087990784.0000
Epoch 287/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2086846976.0000 - mse: 2086846976.0000
Epoch 288/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2085781888.0000 - mse: 2085781888.0000
Epoch 289/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2085000320.0000 - mse: 2085000320.0000
Epoch 290/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2084291072.0000 - mse: 2084291072.0000
Epoch 291/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2083326208.0000 - mse: 2083326208.0000
Epoch 292/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2082906368.0000 - mse: 2082906368.0000
Epoch 293/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2082359168.0000 - mse: 2082359168.0000
Epoch 294/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2081659904.0000 - mse: 2081659904.0000
Epoch 295/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2080017152.0000 - mse: 2080017152.0000
Epoch 296/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2079872896.0000 - mse: 2079872896.0000
Epoch 297/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2078719616.0000 - mse: 2078719616.0000
Epoch 298/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2078270208.0000 - mse: 2078270208.0000
Epoch 299/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2077491200.0000 - mse: 2077491200.0000
Epoch 300/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2076549888.0000 - mse: 2076549888.0000
Epoch 301/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2076395904.0000 - mse: 2076395904.0000
Epoch 302/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2075286912.0000 - mse: 2075286912.0000
Epoch 303/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2074562048.0000 - mse: 2074562048.0000
Epoch 304/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2074218752.0000 - mse: 2074218752.0000
Epoch 305/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2073260416.0000 - mse: 2073260416.0000
Epoch 306/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2072327552.0000 - mse: 2072327552.0000
Epoch 307/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2071856768.0000 - mse: 2071856768.0000
Epoch 308/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2070459008.0000 - mse: 2070459008.0000
Epoch 309/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2070356992.0000 - mse: 2070356992.0000
Epoch 310/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2069599360.0000 - mse: 2069599360.0000
Epoch 311/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2069073024.0000 - mse: 2069073024.0000
Epoch 312/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2067941632.0000 - mse: 2067941632.0000
Epoch 313/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2067634176.0000 - mse: 2067634176.0000
Epoch 314/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2066787072.0000 - mse: 2066787072.0000
Epoch 315/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2066710912.0000 - mse: 2066710912.0000
Epoch 316/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2065188992.0000 - mse: 2065188992.0000
Epoch 317/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2064588672.0000 - mse: 2064588672.0000
Epoch 318/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2064127744.0000 - mse: 2064127744.0000
Epoch 319/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2063104256.0000 - mse: 2063104256.0000
Epoch 320/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2062897408.0000 - mse: 2062897408.0000
Epoch 321/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2061904640.0000 - mse: 2061904640.0000
Epoch 322/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2061803264.0000 - mse: 2061803264.0000
Epoch 323/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2060518144.0000 - mse: 2060518144.0000
Epoch 324/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2060257792.0000 - mse: 2060257792.0000
Epoch 325/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2059518336.0000 - mse: 2059518336.0000
Epoch 326/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2058759424.0000 - mse: 2058759424.0000
Epoch 327/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2057922816.0000 - mse: 2057922816.0000
Epoch 328/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2057270912.0000 - mse: 2057270912.0000
Epoch 329/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2057008384.0000 - mse: 2057008384.0000
Epoch 330/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2055609216.0000 - mse: 2055609216.0000
Epoch 331/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2055638016.0000 - mse: 2055638016.0000
Epoch 332/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2054640256.0000 - mse: 2054640256.0000
Epoch 333/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2054021248.0000 - mse: 2054021248.0000
Epoch 334/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2053634816.0000 - mse: 2053634816.0000
Epoch 335/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2052874496.0000 - mse: 2052874496.0000
Epoch 336/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2052269696.0000 - mse: 2052269696.0000
Epoch 337/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2051739136.0000 - mse: 2051739136.0000
Epoch 338/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2051123840.0000 - mse: 2051123840.0000
Epoch 339/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2050267392.0000 - mse: 2050267392.0000
Epoch 340/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2049518464.0000 - mse: 2049518464.0000
Epoch 341/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2049112320.0000 - mse: 2049112320.0000
Epoch 342/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2047985024.0000 - mse: 2047985024.0000
Epoch 343/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2047634176.0000 - mse: 2047634176.0000
Epoch 344/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2047021568.0000 - mse: 2047021568.0000
Epoch 345/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2046436608.0000 - mse: 2046436608.0000
Epoch 346/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2045575936.0000 - mse: 2045575936.0000
Epoch 347/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2045228288.0000 - mse: 2045228288.0000
Epoch 348/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2044561536.0000 - mse: 2044561536.0000
Epoch 349/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2044165120.0000 - mse: 2044165120.0000
Epoch 350/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2043304320.0000 - mse: 2043304320.0000
Epoch 351/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2043119232.0000 - mse: 2043119232.0000
Epoch 352/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2042302720.0000 - mse: 2042302720.0000
Epoch 353/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2041371520.0000 - mse: 2041371520.0000
Epoch 354/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2041420544.0000 - mse: 2041420544.0000
Epoch 355/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2040527104.0000 - mse: 2040527104.0000
Epoch 356/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2039593472.0000 - mse: 2039593472.0000
Epoch 357/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2039214336.0000 - mse: 2039214336.0000
Epoch 358/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2039300352.0000 - mse: 2039300352.0000
Epoch 359/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2038657664.0000 - mse: 2038657664.0000
Epoch 360/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2037669760.0000 - mse: 2037669760.0000
Epoch 361/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2037539584.0000 - mse: 2037539584.0000
Epoch 362/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2036517248.0000 - mse: 2036517248.0000
Epoch 363/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2035881984.0000 - mse: 2035881984.0000
Epoch 364/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2035748480.0000 - mse: 2035748480.0000
Epoch 365/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2034717312.0000 - mse: 2034717312.0000
Epoch 366/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2034252800.0000 - mse: 2034252800.0000
Epoch 367/500
2163/2163 [==============================] - 7s 3ms/step - loss: 2033885568.0000 - mse: 2033885568.0000
Epoch 368/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2033089152.0000 - mse: 2033089152.0000
Epoch 369/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2033026432.0000 - mse: 2033026432.0000
Epoch 370/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2032561280.0000 - mse: 2032561280.0000
Epoch 371/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2031345792.0000 - mse: 2031345792.0000
Epoch 372/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2031750528.0000 - mse: 2031750528.0000
Epoch 373/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2030470656.0000 - mse: 2030470656.0000
Epoch 374/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2030168192.0000 - mse: 2030168192.0000
Epoch 375/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2029815936.0000 - mse: 2029815936.0000
Epoch 376/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2029331712.0000 - mse: 2029331712.0000
Epoch 377/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2028473472.0000 - mse: 2028473472.0000
Epoch 378/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2028176256.0000 - mse: 2028176256.0000
Epoch 379/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2027708160.0000 - mse: 2027708160.0000
Epoch 380/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2026744960.0000 - mse: 2026744960.0000
Epoch 381/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2026845568.0000 - mse: 2026845568.0000
Epoch 382/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2025995520.0000 - mse: 2025995520.0000
Epoch 383/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2024545408.0000 - mse: 2024545408.0000
Epoch 384/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2024927104.0000 - mse: 2024927104.0000
Epoch 385/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2024144640.0000 - mse: 2024144640.0000
Epoch 386/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2023972480.0000 - mse: 2023972480.0000
Epoch 387/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2022915840.0000 - mse: 2022915840.0000
Epoch 388/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2022712064.0000 - mse: 2022712064.0000
Epoch 389/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2022108928.0000 - mse: 2022108928.0000
Epoch 390/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2021739776.0000 - mse: 2021739776.0000
Epoch 391/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2021405184.0000 - mse: 2021405184.0000
Epoch 392/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2021406208.0000 - mse: 2021406208.0000
Epoch 393/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2020467456.0000 - mse: 2020467456.0000
Epoch 394/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2019744256.0000 - mse: 2019744256.0000
Epoch 395/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2019272064.0000 - mse: 2019272064.0000
Epoch 396/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2018926592.0000 - mse: 2018926592.0000
Epoch 397/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2018640512.0000 - mse: 2018640512.0000
Epoch 398/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2017386240.0000 - mse: 2017386240.0000
Epoch 399/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2017445632.0000 - mse: 2017445632.0000
Epoch 400/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2016327552.0000 - mse: 2016327552.0000
Epoch 401/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2016413696.0000 - mse: 2016413696.0000
Epoch 402/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2016065664.0000 - mse: 2016065664.0000
Epoch 403/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2015186432.0000 - mse: 2015186432.0000
Epoch 404/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2014348544.0000 - mse: 2014348544.0000
Epoch 405/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2013836032.0000 - mse: 2013836032.0000
Epoch 406/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2013958656.0000 - mse: 2013958656.0000
Epoch 407/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2012827904.0000 - mse: 2012827904.0000
Epoch 408/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2011954176.0000 - mse: 2011954176.0000
Epoch 409/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2012296704.0000 - mse: 2012296704.0000
Epoch 410/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2011863552.0000 - mse: 2011863552.0000
Epoch 411/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2011214720.0000 - mse: 2011214720.0000
Epoch 412/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2010453632.0000 - mse: 2010453632.0000
Epoch 413/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2010195072.0000 - mse: 2010195072.0000
Epoch 414/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2008566272.0000 - mse: 2008566272.0000
Epoch 415/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2008842880.0000 - mse: 2008842880.0000
Epoch 416/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2008287872.0000 - mse: 2008287872.0000
Epoch 417/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2007309312.0000 - mse: 2007309312.0000
Epoch 418/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2007162752.0000 - mse: 2007162752.0000
Epoch 419/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2006625280.0000 - mse: 2006625280.0000
Epoch 420/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2005723520.0000 - mse: 2005723520.0000
Epoch 421/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2005281792.0000 - mse: 2005281792.0000
Epoch 422/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2006061952.0000 - mse: 2006061952.0000
Epoch 423/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2004874752.0000 - mse: 2004874752.0000
Epoch 424/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2004610304.0000 - mse: 2004610304.0000
Epoch 425/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2004032512.0000 - mse: 2004032512.0000
Epoch 426/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2002935040.0000 - mse: 2002935040.0000
Epoch 427/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2002384128.0000 - mse: 2002384128.0000
Epoch 428/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2002178560.0000 - mse: 2002178560.0000
Epoch 429/500
2163/2163 [==============================] - 8s 4ms/step - loss: 2001542016.0000 - mse: 2001542016.0000
Epoch 430/500
2163/2163 [==============================] - 8s 4ms/step - loss: 2000992896.0000 - mse: 2000992896.0000
Epoch 431/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2000921472.0000 - mse: 2000921472.0000
Epoch 432/500
2163/2163 [==============================] - 6s 3ms/step - loss: 2000393600.0000 - mse: 2000393600.0000
Epoch 433/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1999758592.0000 - mse: 1999758592.0000
Epoch 434/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1999191424.0000 - mse: 1999191424.0000
Epoch 435/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1998681984.0000 - mse: 1998681984.0000
Epoch 436/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1998474240.0000 - mse: 1998474240.0000
Epoch 437/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1998102144.0000 - mse: 1998102144.0000
Epoch 438/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1997153792.0000 - mse: 1997153792.0000
Epoch 439/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1996696704.0000 - mse: 1996696704.0000
Epoch 440/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1996844800.0000 - mse: 1996844800.0000
Epoch 441/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1996585472.0000 - mse: 1996585472.0000
Epoch 442/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1995675904.0000 - mse: 1995675904.0000
Epoch 443/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1994835712.0000 - mse: 1994835712.0000
Epoch 444/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1994020736.0000 - mse: 1994020736.0000
Epoch 445/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1993544960.0000 - mse: 1993544960.0000
Epoch 446/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1993480320.0000 - mse: 1993480320.0000
Epoch 447/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1993120512.0000 - mse: 1993120512.0000
Epoch 448/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1992767232.0000 - mse: 1992767232.0000
Epoch 449/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1992621440.0000 - mse: 1992621440.0000
Epoch 450/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1991967104.0000 - mse: 1991967104.0000
Epoch 451/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1991883264.0000 - mse: 1991883264.0000
Epoch 452/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1991063040.0000 - mse: 1991063040.0000
Epoch 453/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1990682368.0000 - mse: 1990682368.0000
Epoch 454/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1989977216.0000 - mse: 1989977216.0000
Epoch 455/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1989668352.0000 - mse: 1989668352.0000
Epoch 456/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1989548928.0000 - mse: 1989548928.0000
Epoch 457/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1988149376.0000 - mse: 1988149376.0000
Epoch 458/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1988414848.0000 - mse: 1988414848.0000
Epoch 459/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1987443968.0000 - mse: 1987443968.0000
Epoch 460/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1986937600.0000 - mse: 1986937600.0000
Epoch 461/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1986810368.0000 - mse: 1986810368.0000
Epoch 462/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1985968512.0000 - mse: 1985968512.0000
Epoch 463/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1985556736.0000 - mse: 1985556736.0000
Epoch 464/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1985214464.0000 - mse: 1985214464.0000
Epoch 465/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1985339136.0000 - mse: 1985339136.0000
Epoch 466/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1984574976.0000 - mse: 1984574976.0000
Epoch 467/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1983832448.0000 - mse: 1983832448.0000
Epoch 468/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1983630976.0000 - mse: 1983630976.0000
Epoch 469/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1983239168.0000 - mse: 1983239168.0000
Epoch 470/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1982715904.0000 - mse: 1982715904.0000
Epoch 471/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1982576896.0000 - mse: 1982576896.0000
Epoch 472/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1982456064.0000 - mse: 1982456064.0000
Epoch 473/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1981919232.0000 - mse: 1981919232.0000
Epoch 474/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1981184128.0000 - mse: 1981184128.0000
Epoch 475/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1981062144.0000 - mse: 1981062144.0000
Epoch 476/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1979913216.0000 - mse: 1979913216.0000
Epoch 477/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1979885696.0000 - mse: 1979885696.0000
Epoch 478/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1979623296.0000 - mse: 1979623296.0000
Epoch 479/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1978377728.0000 - mse: 1978377728.0000
Epoch 480/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1978752000.0000 - mse: 1978752000.0000
Epoch 481/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1977744128.0000 - mse: 1977744128.0000
Epoch 482/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1977124992.0000 - mse: 1977124992.0000
Epoch 483/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1977038464.0000 - mse: 1977038464.0000
Epoch 484/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1976687872.0000 - mse: 1976687872.0000
Epoch 485/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1975982080.0000 - mse: 1975982080.0000
Epoch 486/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1975365376.0000 - mse: 1975365376.0000
Epoch 487/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1975362048.0000 - mse: 1975362048.0000
Epoch 488/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1974885376.0000 - mse: 1974885376.0000
Epoch 489/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1974827392.0000 - mse: 1974827392.0000
Epoch 490/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1974362752.0000 - mse: 1974362752.0000
Epoch 491/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1972717568.0000 - mse: 1972717568.0000
Epoch 492/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1973244160.0000 - mse: 1973244160.0000
Epoch 493/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1972879616.0000 - mse: 1972879616.0000
Epoch 494/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1972418816.0000 - mse: 1972418816.0000
Epoch 495/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1972549120.0000 - mse: 1972549120.0000
Epoch 496/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1971509888.0000 - mse: 1971509888.0000
Epoch 497/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1971847936.0000 - mse: 1971847936.0000
Epoch 498/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1971114112.0000 - mse: 1971114112.0000
Epoch 499/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1970574592.0000 - mse: 1970574592.0000
Epoch 500/500
2163/2163 [==============================] - 6s 3ms/step - loss: 1970048384.0000 - mse: 1970048384.0000

<keras.callbacks.History at 0x7f5bf384eb20>

Generamos las predicciones para train y vemos cómo performa.

In [73]:

y_pred_regresion = best_model.predict(x_train_regresion)
imprimir_metricas_de_regresion(y_train_regresion, y_pred_regresion)

Out[73]:

2163/2163 [==============================] - 3s 1ms/step

Métricas de regresión

El error (mse) de test es: 1965073952.8753333
El error (rmse) de test es: 44329.1546600579
El score R2 es: 0.8167946762967582

Ahora lo probamos con los datos de test.

In [74]:

y_pred_regresion = best_model.predict(x_test_regresion)
imprimir_metricas_de_regresion(y_test_regresion, y_pred_regresion)

Out[74]:

540/540 [==============================] - 1s 1ms/step

Métricas de regresión

El error (mse) de test es: 2076981265.8100636
El error (rmse) de test es: 45573.90992453976
El score R2 es: 0.8044855283509453

Conclusiones

Las redes neuronales son muy poderosas pero el hecho de que su entrenamiento sea lento y la búsqueda de hiperparámetros tenga que ser manual hace que no podamos sacarle el provecho completo.

En comparación, XGBoost tiene mejores métricas en menos tiempo, y nos permite hacer una mejor búsqueda de hiperparámetros, por lo que para nuestros objetivos nos sirve más.

Configuración inicial

Funciones útiles

Redes neuronales

Clasificación

Regresión

Conclusiones

Product

Resources

Company