In [1]:
import numpy as np
import matplotlib.pyplot as plt
import tensorflow as tf
keras = tf.keras
In [2]:
def plot_series(time, series, format="-", start=0, end=None, label=None):
plt.plot(time[start:end], series[start:end], format, label=label)
plt.xlabel("Time")
plt.ylabel("Value")
if label:
plt.legend(fontsize=14)
plt.grid(True)
def trend(time, slope=0):
return slope * time
def seasonal_pattern(season_time):
"""Just an arbitrary pattern, you can change it if you wish"""
return np.where(season_time < 0.4,
np.cos(season_time * 2 * np.pi),
1 / np.exp(3 * season_time))
def seasonality(time, period, amplitude=1, phase=0):
"""Repeats the same pattern at each period"""
season_time = ((time + phase) % period) / period
return amplitude * seasonal_pattern(season_time)
def white_noise(time, noise_level=1, seed=None):
rnd = np.random.RandomState(seed)
return rnd.randn(len(time)) * noise_level
def sequential_window_dataset(series, window_size):
series = tf.expand_dims(series, axis=-1)
ds = tf.data.Dataset.from_tensor_slices(series)
ds = ds.window(window_size + 1, shift=window_size, drop_remainder=True)
ds = ds.flat_map(lambda window: window.batch(window_size + 1))
ds = ds.map(lambda window: (window[:-1], window[1:]))
return ds.batch(1).prefetch(1)
In [3]:
time = np.arange(4 * 365 + 1)
slope = 0.05
baseline = 10
amplitude = 40
series = baseline + trend(time, slope) + seasonality(time, period=365, amplitude=amplitude)
noise_level = 5
noise = white_noise(time, noise_level, seed=42)
series += noise
plt.figure(figsize=(10, 6))
plot_series(time, series)
plt.show()
In [4]:
split_time = 1000
time_train = time[:split_time]
x_train = series[:split_time]
time_valid = time[split_time:]
x_valid = series[split_time:]
In [5]:
class ResetStatesCallback(keras.callbacks.Callback):
def on_epoch_begin(self, epoch, logs):
self.model.reset_states()
LSTM RNN Forecasting¶
In [6]:
keras.backend.clear_session()
tf.random.set_seed(42)
np.random.seed(42)
window_size = 30
train_set = sequential_window_dataset(x_train, window_size)
model = keras.models.Sequential([
keras.layers.LSTM(100, return_sequences=True, stateful=True,
batch_input_shape=[1, None, 1]),
keras.layers.LSTM(100, return_sequences=True, stateful=True),
keras.layers.Dense(1),
keras.layers.Lambda(lambda x: x * 200.0)
])
lr_schedule = keras.callbacks.LearningRateScheduler(
lambda epoch: 1e-8 * 10**(epoch / 20))
reset_states = ResetStatesCallback()
optimizer = keras.optimizers.SGD(lr=1e-8, momentum=0.9)
model.compile(loss=keras.losses.Huber(),
optimizer=optimizer,
metrics=["mae"])
history = model.fit(train_set, epochs=100,
callbacks=[lr_schedule, reset_states])
Epoch 1/100 33/33 [==============================] - 0s 7ms/step - loss: 90.9191 - mae: 91.4191 Epoch 2/100 33/33 [==============================] - 0s 6ms/step - loss: 83.8978 - mae: 84.3978 Epoch 3/100 33/33 [==============================] - 0s 6ms/step - loss: 75.4974 - mae: 75.9974 Epoch 4/100 33/33 [==============================] - 0s 6ms/step - loss: 66.0676 - mae: 66.5676 Epoch 5/100 33/33 [==============================] - 0s 6ms/step - loss: 55.5131 - mae: 56.0131 Epoch 6/100 33/33 [==============================] - 0s 6ms/step - loss: 43.7599 - mae: 44.2597 Epoch 7/100 33/33 [==============================] - 0s 6ms/step - loss: 30.7909 - mae: 31.2900 Epoch 8/100 33/33 [==============================] - 0s 6ms/step - loss: 17.1478 - mae: 17.6436 Epoch 9/100 33/33 [==============================] - 0s 6ms/step - loss: 10.0311 - mae: 10.5178 Epoch 10/100 33/33 [==============================] - 0s 7ms/step - loss: 10.0607 - mae: 10.5524 Epoch 11/100 33/33 [==============================] - 0s 6ms/step - loss: 9.8063 - mae: 10.2981 Epoch 12/100 33/33 [==============================] - 0s 6ms/step - loss: 9.3912 - mae: 9.8815 Epoch 13/100 33/33 [==============================] - 0s 6ms/step - loss: 9.0001 - mae: 9.4885 Epoch 14/100 33/33 [==============================] - 0s 6ms/step - loss: 8.6473 - mae: 9.1320 Epoch 15/100 33/33 [==============================] - 0s 6ms/step - loss: 8.3308 - mae: 8.8157 Epoch 16/100 33/33 [==============================] - 0s 6ms/step - loss: 8.0550 - mae: 8.5387 Epoch 17/100 33/33 [==============================] - 0s 6ms/step - loss: 7.8428 - mae: 8.3241 Epoch 18/100 33/33 [==============================] - 0s 6ms/step - loss: 7.6701 - mae: 8.1520 Epoch 19/100 33/33 [==============================] - 0s 6ms/step - loss: 7.5326 - mae: 8.0152 Epoch 20/100 33/33 [==============================] - 0s 6ms/step - loss: 7.4123 - mae: 7.8949 Epoch 21/100 33/33 [==============================] - 0s 6ms/step - loss: 7.3345 - mae: 7.8177 Epoch 22/100 33/33 [==============================] - 0s 7ms/step - loss: 7.2263 - mae: 7.7074 Epoch 23/100 33/33 [==============================] - 0s 7ms/step - loss: 7.0942 - mae: 7.5725 Epoch 24/100 33/33 [==============================] - 0s 6ms/step - loss: 6.9832 - mae: 7.4620 Epoch 25/100 33/33 [==============================] - 0s 6ms/step - loss: 6.8444 - mae: 7.3253 Epoch 26/100 33/33 [==============================] - 0s 7ms/step - loss: 6.7290 - mae: 7.2108 Epoch 27/100 33/33 [==============================] - 0s 6ms/step - loss: 6.5833 - mae: 7.0666 Epoch 28/100 33/33 [==============================] - 0s 6ms/step - loss: 6.4450 - mae: 6.9279 Epoch 29/100 33/33 [==============================] - 0s 6ms/step - loss: 6.2907 - mae: 6.7729 Epoch 30/100 33/33 [==============================] - 0s 6ms/step - loss: 6.1173 - mae: 6.5962 Epoch 31/100 33/33 [==============================] - 0s 6ms/step - loss: 5.9434 - mae: 6.4210 Epoch 32/100 33/33 [==============================] - 0s 6ms/step - loss: 5.7794 - mae: 6.2585 Epoch 33/100 33/33 [==============================] - 0s 6ms/step - loss: 5.6534 - mae: 6.1340 Epoch 34/100 33/33 [==============================] - 0s 6ms/step - loss: 5.5627 - mae: 6.0405 Epoch 35/100 33/33 [==============================] - 0s 6ms/step - loss: 5.4656 - mae: 5.9413 Epoch 36/100 33/33 [==============================] - 0s 7ms/step - loss: 5.3491 - mae: 5.8258 Epoch 37/100 33/33 [==============================] - 0s 6ms/step - loss: 5.2397 - mae: 5.7178 Epoch 38/100 33/33 [==============================] - 0s 6ms/step - loss: 5.1747 - mae: 5.6543 Epoch 39/100 33/33 [==============================] - 0s 6ms/step - loss: 5.1366 - mae: 5.6164 Epoch 40/100 33/33 [==============================] - 0s 6ms/step - loss: 5.0991 - mae: 5.5778 Epoch 41/100 33/33 [==============================] - 0s 6ms/step - loss: 5.1034 - mae: 5.5833 Epoch 42/100 33/33 [==============================] - 0s 6ms/step - loss: 5.1371 - mae: 5.6149 Epoch 43/100 33/33 [==============================] - 0s 6ms/step - loss: 5.1539 - mae: 5.6325 Epoch 44/100 33/33 [==============================] - 0s 7ms/step - loss: 5.1025 - mae: 5.5813 Epoch 45/100 33/33 [==============================] - 0s 6ms/step - loss: 5.5236 - mae: 6.0050 Epoch 46/100 33/33 [==============================] - 0s 6ms/step - loss: 6.0236 - mae: 6.5054 Epoch 47/100 33/33 [==============================] - 0s 6ms/step - loss: 6.3322 - mae: 6.8183 Epoch 48/100 33/33 [==============================] - 0s 7ms/step - loss: 7.4924 - mae: 7.9806 Epoch 49/100 33/33 [==============================] - 0s 6ms/step - loss: 8.1303 - mae: 8.6211 Epoch 50/100 33/33 [==============================] - 0s 6ms/step - loss: 11.0320 - mae: 11.5218 Epoch 51/100 33/33 [==============================] - 0s 6ms/step - loss: 5.9608 - mae: 6.4464 Epoch 52/100 33/33 [==============================] - 0s 6ms/step - loss: 8.0853 - mae: 8.5729 Epoch 53/100 33/33 [==============================] - 0s 7ms/step - loss: 8.6677 - mae: 9.1565 Epoch 54/100 33/33 [==============================] - 0s 7ms/step - loss: 11.0260 - mae: 11.5162 Epoch 55/100 33/33 [==============================] - 0s 6ms/step - loss: 9.2377 - mae: 9.7287 Epoch 56/100 33/33 [==============================] - 0s 6ms/step - loss: 14.0616 - mae: 14.5566 Epoch 57/100 33/33 [==============================] - 0s 6ms/step - loss: 11.2927 - mae: 11.7865 Epoch 58/100 33/33 [==============================] - 0s 7ms/step - loss: 9.7924 - mae: 10.2825 Epoch 59/100 33/33 [==============================] - 0s 6ms/step - loss: 20.5439 - mae: 21.0352 Epoch 60/100 33/33 [==============================] - 0s 6ms/step - loss: 21.6066 - mae: 22.1004 Epoch 61/100 33/33 [==============================] - 0s 6ms/step - loss: 14.8527 - mae: 15.3486 Epoch 62/100 33/33 [==============================] - 0s 6ms/step - loss: 17.4853 - mae: 17.9803 Epoch 63/100 33/33 [==============================] - 0s 6ms/step - loss: 13.0441 - mae: 13.5368 Epoch 64/100 33/33 [==============================] - 0s 6ms/step - loss: 9.5840 - mae: 10.0718 Epoch 65/100 33/33 [==============================] - 0s 6ms/step - loss: 13.4697 - mae: 13.9598 Epoch 66/100 33/33 [==============================] - 0s 7ms/step - loss: 12.2081 - mae: 12.7018 Epoch 67/100 33/33 [==============================] - 0s 6ms/step - loss: 16.0403 - mae: 16.5327 Epoch 68/100 33/33 [==============================] - 0s 6ms/step - loss: 12.5217 - mae: 13.0125 Epoch 69/100 33/33 [==============================] - 0s 6ms/step - loss: 16.6670 - mae: 17.1614 Epoch 70/100 33/33 [==============================] - 0s 6ms/step - loss: 17.0632 - mae: 17.5551 Epoch 71/100 33/33 [==============================] - 0s 6ms/step - loss: 12.4808 - mae: 12.9731 Epoch 72/100 33/33 [==============================] - 0s 6ms/step - loss: 9.7367 - mae: 10.2256 Epoch 73/100 33/33 [==============================] - 0s 6ms/step - loss: 27.4516 - mae: 27.9494 Epoch 74/100 33/33 [==============================] - 0s 6ms/step - loss: 17.3039 - mae: 17.7997 Epoch 75/100 33/33 [==============================] - 0s 6ms/step - loss: 16.8151 - mae: 17.3107 Epoch 76/100 33/33 [==============================] - 0s 6ms/step - loss: 10.9015 - mae: 11.3919 Epoch 77/100 33/33 [==============================] - 0s 6ms/step - loss: 10.8313 - mae: 11.3188 Epoch 78/100 33/33 [==============================] - 0s 6ms/step - loss: 18.3577 - mae: 18.8528 Epoch 79/100 33/33 [==============================] - 0s 6ms/step - loss: 17.3953 - mae: 17.8918 Epoch 80/100 33/33 [==============================] - 0s 6ms/step - loss: 14.3315 - mae: 14.8260 Epoch 81/100 33/33 [==============================] - 0s 6ms/step - loss: 11.0187 - mae: 11.5131 Epoch 82/100 33/33 [==============================] - 0s 6ms/step - loss: 13.6761 - mae: 14.1616 Epoch 83/100 33/33 [==============================] - 0s 6ms/step - loss: 11.8563 - mae: 12.3520 Epoch 84/100 33/33 [==============================] - 0s 6ms/step - loss: 16.8495 - mae: 17.3432 Epoch 85/100 33/33 [==============================] - 0s 6ms/step - loss: 20.7220 - mae: 21.2175 Epoch 86/100 33/33 [==============================] - 0s 6ms/step - loss: 23.8672 - mae: 24.3624 Epoch 87/100 33/33 [==============================] - 0s 6ms/step - loss: 31.4446 - mae: 31.9437 Epoch 88/100 33/33 [==============================] - 0s 7ms/step - loss: 27.5935 - mae: 28.0884 Epoch 89/100 33/33 [==============================] - 0s 7ms/step - loss: 25.3679 - mae: 25.8586 Epoch 90/100 33/33 [==============================] - 0s 6ms/step - loss: 26.0335 - mae: 26.5311 Epoch 91/100 33/33 [==============================] - 0s 6ms/step - loss: 40.2144 - mae: 40.7131 Epoch 92/100 33/33 [==============================] - 0s 6ms/step - loss: 33.3154 - mae: 33.8143 Epoch 93/100 33/33 [==============================] - 0s 6ms/step - loss: 33.2166 - mae: 33.7144 Epoch 94/100 33/33 [==============================] - 0s 6ms/step - loss: 53.2613 - mae: 53.7586 Epoch 95/100 33/33 [==============================] - 0s 6ms/step - loss: 58.7900 - mae: 59.2888 Epoch 96/100 33/33 [==============================] - 0s 6ms/step - loss: 39.9028 - mae: 40.4008 Epoch 97/100 33/33 [==============================] - 0s 6ms/step - loss: 44.4015 - mae: 44.9006 Epoch 98/100 33/33 [==============================] - 0s 6ms/step - loss: 51.3196 - mae: 51.8179 Epoch 99/100 33/33 [==============================] - 0s 6ms/step - loss: 73.1629 - mae: 73.6617 Epoch 100/100 33/33 [==============================] - 0s 6ms/step - loss: 54.0318 - mae: 54.5270
In [7]:
plt.semilogx(history.history["lr"], history.history["loss"])
plt.axis([1e-8, 1e-4, 0, 30])
Out[7]:
(1e-08, 0.0001, 0.0, 30.0)
In [8]:
keras.backend.clear_session()
tf.random.set_seed(42)
np.random.seed(42)
window_size = 30
train_set = sequential_window_dataset(x_train, window_size)
valid_set = sequential_window_dataset(x_valid, window_size)
model = keras.models.Sequential([
keras.layers.LSTM(100, return_sequences=True, stateful=True,
batch_input_shape=[1, None, 1]),
keras.layers.LSTM(100, return_sequences=True, stateful=True),
keras.layers.Dense(1),
keras.layers.Lambda(lambda x: x * 200.0)
])
optimizer = keras.optimizers.SGD(lr=5e-7, momentum=0.9)
model.compile(loss=keras.losses.Huber(),
optimizer=optimizer,
metrics=["mae"])
reset_states = ResetStatesCallback()
model_checkpoint = keras.callbacks.ModelCheckpoint(
"my_checkpoint.h5", save_best_only=True)
early_stopping = keras.callbacks.EarlyStopping(patience=50)
model.fit(train_set, epochs=500,
validation_data=valid_set,
callbacks=[early_stopping, model_checkpoint, reset_states])
Epoch 1/500 33/33 [==============================] - 1s 28ms/step - loss: 30.0122 - mae: 30.5050 - val_loss: 13.0810 - val_mae: 13.5746 Epoch 2/500 33/33 [==============================] - 0s 9ms/step - loss: 15.3618 - mae: 15.8570 - val_loss: 10.0991 - val_mae: 10.5941 Epoch 3/500 33/33 [==============================] - 0s 8ms/step - loss: 10.9284 - mae: 11.4173 - val_loss: 11.2192 - val_mae: 11.7083 Epoch 4/500 33/33 [==============================] - 0s 8ms/step - loss: 9.8165 - mae: 10.3052 - val_loss: 12.2806 - val_mae: 12.7757 Epoch 5/500 33/33 [==============================] - 0s 9ms/step - loss: 7.8424 - mae: 8.3317 - val_loss: 9.6437 - val_mae: 10.1336 Epoch 6/500 33/33 [==============================] - 0s 9ms/step - loss: 7.1708 - mae: 7.6561 - val_loss: 10.6124 - val_mae: 11.1036 Epoch 7/500 33/33 [==============================] - 0s 8ms/step - loss: 5.9200 - mae: 6.4041 - val_loss: 10.1595 - val_mae: 10.6522 Epoch 8/500 33/33 [==============================] - 0s 9ms/step - loss: 5.6843 - mae: 6.1652 - val_loss: 9.6997 - val_mae: 10.1896 Epoch 9/500 33/33 [==============================] - 0s 9ms/step - loss: 5.6694 - mae: 6.1486 - val_loss: 10.0423 - val_mae: 10.5330 Epoch 10/500 33/33 [==============================] - 0s 8ms/step - loss: 5.7491 - mae: 6.2312 - val_loss: 9.8009 - val_mae: 10.2917 Epoch 11/500 33/33 [==============================] - 0s 9ms/step - loss: 5.8786 - mae: 6.3598 - val_loss: 9.6807 - val_mae: 10.1714 Epoch 12/500 33/33 [==============================] - 0s 9ms/step - loss: 5.9175 - mae: 6.3984 - val_loss: 9.4422 - val_mae: 9.9319 Epoch 13/500 33/33 [==============================] - 0s 9ms/step - loss: 5.8339 - mae: 6.3116 - val_loss: 9.1527 - val_mae: 9.6419 Epoch 14/500 33/33 [==============================] - 0s 8ms/step - loss: 5.7167 - mae: 6.1943 - val_loss: 9.1645 - val_mae: 9.6542 Epoch 15/500 33/33 [==============================] - 0s 9ms/step - loss: 5.5860 - mae: 6.0623 - val_loss: 9.4113 - val_mae: 9.9008 Epoch 16/500 33/33 [==============================] - 0s 9ms/step - loss: 5.4922 - mae: 5.9689 - val_loss: 9.6087 - val_mae: 10.0986 Epoch 17/500 33/33 [==============================] - 0s 8ms/step - loss: 5.4966 - mae: 5.9752 - val_loss: 9.6966 - val_mae: 10.1863 Epoch 18/500 33/33 [==============================] - 0s 8ms/step - loss: 5.5476 - mae: 6.0263 - val_loss: 9.6846 - val_mae: 10.1746 Epoch 19/500 33/33 [==============================] - 0s 9ms/step - loss: 5.6006 - mae: 6.0809 - val_loss: 9.6537 - val_mae: 10.1434 Epoch 20/500 33/33 [==============================] - 0s 8ms/step - loss: 5.6103 - mae: 6.0913 - val_loss: 9.5965 - val_mae: 10.0860 Epoch 21/500 33/33 [==============================] - 0s 8ms/step - loss: 5.5743 - mae: 6.0525 - val_loss: 9.5047 - val_mae: 9.9941 Epoch 22/500 33/33 [==============================] - 0s 8ms/step - loss: 5.5159 - mae: 5.9908 - val_loss: 9.4526 - val_mae: 9.9416 Epoch 23/500 33/33 [==============================] - 0s 8ms/step - loss: 5.4661 - mae: 5.9414 - val_loss: 9.4918 - val_mae: 9.9810 Epoch 24/500 33/33 [==============================] - 0s 8ms/step - loss: 5.4316 - mae: 5.9093 - val_loss: 9.5130 - val_mae: 10.0043 Epoch 25/500 33/33 [==============================] - 0s 9ms/step - loss: 5.4204 - mae: 5.8999 - val_loss: 9.5088 - val_mae: 10.0010 Epoch 26/500 33/33 [==============================] - 0s 8ms/step - loss: 5.4186 - mae: 5.8987 - val_loss: 9.4891 - val_mae: 9.9818 Epoch 27/500 33/33 [==============================] - 0s 8ms/step - loss: 5.4135 - mae: 5.8932 - val_loss: 9.4548 - val_mae: 9.9476 Epoch 28/500 33/33 [==============================] - 0s 9ms/step - loss: 5.4015 - mae: 5.8806 - val_loss: 9.4135 - val_mae: 9.9062 Epoch 29/500 33/33 [==============================] - 0s 9ms/step - loss: 5.3869 - mae: 5.8655 - val_loss: 9.3735 - val_mae: 9.8662 Epoch 30/500 33/33 [==============================] - 0s 8ms/step - loss: 5.3712 - mae: 5.8498 - val_loss: 9.3322 - val_mae: 9.8248 Epoch 31/500 33/33 [==============================] - 0s 9ms/step - loss: 5.3564 - mae: 5.8350 - val_loss: 9.2882 - val_mae: 9.7808 Epoch 32/500 33/33 [==============================] - 0s 9ms/step - loss: 5.3417 - mae: 5.8204 - val_loss: 9.2416 - val_mae: 9.7340 Epoch 33/500 33/33 [==============================] - 0s 9ms/step - loss: 5.3267 - mae: 5.8054 - val_loss: 9.1931 - val_mae: 9.6855 Epoch 34/500 33/33 [==============================] - 0s 9ms/step - loss: 5.3104 - mae: 5.7891 - val_loss: 9.1413 - val_mae: 9.6336 Epoch 35/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2948 - mae: 5.7736 - val_loss: 9.0912 - val_mae: 9.5834 Epoch 36/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2808 - mae: 5.7595 - val_loss: 9.0401 - val_mae: 9.5321 Epoch 37/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2688 - mae: 5.7475 - val_loss: 8.9865 - val_mae: 9.4782 Epoch 38/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2582 - mae: 5.7368 - val_loss: 8.9292 - val_mae: 9.4204 Epoch 39/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2478 - mae: 5.7263 - val_loss: 8.8691 - val_mae: 9.3598 Epoch 40/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2369 - mae: 5.7155 - val_loss: 8.8074 - val_mae: 9.2974 Epoch 41/500 33/33 [==============================] - 0s 10ms/step - loss: 5.2250 - mae: 5.7036 - val_loss: 8.7454 - val_mae: 9.2347 Epoch 42/500 33/33 [==============================] - 0s 9ms/step - loss: 5.2118 - mae: 5.6905 - val_loss: 8.6859 - val_mae: 9.1747 Epoch 43/500 33/33 [==============================] - 0s 9ms/step - loss: 5.1972 - mae: 5.6763 - val_loss: 8.6281 - val_mae: 9.1162 Epoch 44/500 33/33 [==============================] - 0s 10ms/step - loss: 5.1813 - mae: 5.6606 - val_loss: 8.5700 - val_mae: 9.0573 Epoch 45/500 33/33 [==============================] - 0s 9ms/step - loss: 5.1642 - mae: 5.6438 - val_loss: 8.5138 - val_mae: 9.0003 Epoch 46/500 33/33 [==============================] - 0s 9ms/step - loss: 5.1463 - mae: 5.6260 - val_loss: 8.4536 - val_mae: 8.9391 Epoch 47/500 33/33 [==============================] - 0s 10ms/step - loss: 5.1282 - mae: 5.6080 - val_loss: 8.3915 - val_mae: 8.8757 Epoch 48/500 33/33 [==============================] - 0s 9ms/step - loss: 5.1110 - mae: 5.5906 - val_loss: 8.3288 - val_mae: 8.8123 Epoch 49/500 33/33 [==============================] - 0s 9ms/step - loss: 5.0950 - mae: 5.5745 - val_loss: 8.2669 - val_mae: 8.7504 Epoch 50/500 33/33 [==============================] - 0s 10ms/step - loss: 5.0798 - mae: 5.5591 - val_loss: 8.2066 - val_mae: 8.6900 Epoch 51/500 33/33 [==============================] - 0s 9ms/step - loss: 5.0644 - mae: 5.5435 - val_loss: 8.1468 - val_mae: 8.6307 Epoch 52/500 33/33 [==============================] - 0s 9ms/step - loss: 5.0486 - mae: 5.5274 - val_loss: 8.0832 - val_mae: 8.5679 Epoch 53/500 33/33 [==============================] - 0s 10ms/step - loss: 5.0323 - mae: 5.5108 - val_loss: 8.0271 - val_mae: 8.5125 Epoch 54/500 33/33 [==============================] - 0s 9ms/step - loss: 5.0146 - mae: 5.4930 - val_loss: 7.9815 - val_mae: 8.4674 Epoch 55/500 33/33 [==============================] - 0s 10ms/step - loss: 4.9958 - mae: 5.4740 - val_loss: 7.9396 - val_mae: 8.4258 Epoch 56/500 33/33 [==============================] - 0s 9ms/step - loss: 4.9777 - mae: 5.4556 - val_loss: 7.9008 - val_mae: 8.3874 Epoch 57/500 33/33 [==============================] - 0s 9ms/step - loss: 4.9600 - mae: 5.4378 - val_loss: 7.8673 - val_mae: 8.3541 Epoch 58/500 33/33 [==============================] - 0s 9ms/step - loss: 4.9414 - mae: 5.4198 - val_loss: 7.8364 - val_mae: 8.3233 Epoch 59/500 33/33 [==============================] - 0s 10ms/step - loss: 4.9229 - mae: 5.4017 - val_loss: 7.8052 - val_mae: 8.2919 Epoch 60/500 33/33 [==============================] - 0s 10ms/step - loss: 4.9055 - mae: 5.3841 - val_loss: 7.7852 - val_mae: 8.2718 Epoch 61/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8906 - mae: 5.3689 - val_loss: 7.7705 - val_mae: 8.2575 Epoch 62/500 33/33 [==============================] - 0s 10ms/step - loss: 4.8800 - mae: 5.3581 - val_loss: 7.7570 - val_mae: 8.2442 Epoch 63/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8741 - mae: 5.3520 - val_loss: 7.7333 - val_mae: 8.2205 Epoch 64/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8722 - mae: 5.3498 - val_loss: 7.6966 - val_mae: 8.1836 Epoch 65/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8715 - mae: 5.3490 - val_loss: 7.6494 - val_mae: 8.1360 Epoch 66/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8703 - mae: 5.3477 - val_loss: 7.5974 - val_mae: 8.0836 Epoch 67/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8671 - mae: 5.3447 - val_loss: 7.5479 - val_mae: 8.0336 Epoch 68/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8605 - mae: 5.3386 - val_loss: 7.5021 - val_mae: 7.9872 Epoch 69/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8501 - mae: 5.3287 - val_loss: 7.4629 - val_mae: 7.9475 Epoch 70/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8373 - mae: 5.3167 - val_loss: 7.4356 - val_mae: 7.9198 Epoch 71/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8242 - mae: 5.3040 - val_loss: 7.4154 - val_mae: 7.8996 Epoch 72/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8123 - mae: 5.2921 - val_loss: 7.4003 - val_mae: 7.8847 Epoch 73/500 33/33 [==============================] - 0s 9ms/step - loss: 4.8031 - mae: 5.2826 - val_loss: 7.3807 - val_mae: 7.8652 Epoch 74/500 33/33 [==============================] - 0s 10ms/step - loss: 4.7962 - mae: 5.2754 - val_loss: 7.3570 - val_mae: 7.8416 Epoch 75/500 33/33 [==============================] - 0s 10ms/step - loss: 4.7908 - mae: 5.2698 - val_loss: 7.3274 - val_mae: 7.8120 Epoch 76/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7861 - mae: 5.2651 - val_loss: 7.2960 - val_mae: 7.7805 Epoch 77/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7807 - mae: 5.2598 - val_loss: 7.2653 - val_mae: 7.7499 Epoch 78/500 33/33 [==============================] - 0s 10ms/step - loss: 4.7738 - mae: 5.2531 - val_loss: 7.2375 - val_mae: 7.7222 Epoch 79/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7650 - mae: 5.2446 - val_loss: 7.2146 - val_mae: 7.6995 Epoch 80/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7540 - mae: 5.2336 - val_loss: 7.2061 - val_mae: 7.6912 Epoch 81/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7409 - mae: 5.2206 - val_loss: 7.2074 - val_mae: 7.6927 Epoch 82/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7278 - mae: 5.2074 - val_loss: 7.2109 - val_mae: 7.6965 Epoch 83/500 33/33 [==============================] - 0s 8ms/step - loss: 4.7155 - mae: 5.1949 - val_loss: 7.2163 - val_mae: 7.7019 Epoch 84/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7051 - mae: 5.1843 - val_loss: 7.2190 - val_mae: 7.7044 Epoch 85/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6969 - mae: 5.1757 - val_loss: 7.2145 - val_mae: 7.6999 Epoch 86/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6917 - mae: 5.1703 - val_loss: 7.1949 - val_mae: 7.6802 Epoch 87/500 33/33 [==============================] - 0s 10ms/step - loss: 4.6904 - mae: 5.1688 - val_loss: 7.1590 - val_mae: 7.6443 Epoch 88/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6946 - mae: 5.1731 - val_loss: 7.0969 - val_mae: 7.5826 Epoch 89/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7069 - mae: 5.1860 - val_loss: 6.9948 - val_mae: 7.4810 Epoch 90/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7216 - mae: 5.2014 - val_loss: 6.9047 - val_mae: 7.3906 Epoch 91/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7094 - mae: 5.1872 - val_loss: 6.9057 - val_mae: 7.3916 Epoch 92/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6958 - mae: 5.1747 - val_loss: 6.9038 - val_mae: 7.3895 Epoch 93/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6750 - mae: 5.1530 - val_loss: 6.9370 - val_mae: 7.4225 Epoch 94/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6657 - mae: 5.1438 - val_loss: 6.9525 - val_mae: 7.4376 Epoch 95/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6477 - mae: 5.1251 - val_loss: 6.9946 - val_mae: 7.4794 Epoch 96/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6498 - mae: 5.1272 - val_loss: 7.0029 - val_mae: 7.4876 Epoch 97/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6275 - mae: 5.1044 - val_loss: 7.0682 - val_mae: 7.5530 Epoch 98/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6444 - mae: 5.1213 - val_loss: 7.0426 - val_mae: 7.5274 Epoch 99/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6084 - mae: 5.0841 - val_loss: 7.1595 - val_mae: 7.6452 Epoch 100/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6195 - mae: 5.0964 - val_loss: 7.0713 - val_mae: 7.5561 Epoch 101/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6172 - mae: 5.0927 - val_loss: 7.0731 - val_mae: 7.5580 Epoch 102/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6493 - mae: 5.1257 - val_loss: 6.9755 - val_mae: 7.4603 Epoch 103/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6196 - mae: 5.0946 - val_loss: 7.0608 - val_mae: 7.5457 Epoch 104/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6380 - mae: 5.1137 - val_loss: 6.8513 - val_mae: 7.3364 Epoch 105/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6720 - mae: 5.1471 - val_loss: 6.7294 - val_mae: 7.2154 Epoch 106/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7122 - mae: 5.1874 - val_loss: 6.6637 - val_mae: 7.1498 Epoch 107/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7030 - mae: 5.1816 - val_loss: 6.5765 - val_mae: 7.0622 Epoch 108/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7216 - mae: 5.2018 - val_loss: 6.4016 - val_mae: 6.8836 Epoch 109/500 33/33 [==============================] - 0s 9ms/step - loss: 4.7133 - mae: 5.1939 - val_loss: 6.3716 - val_mae: 6.8527 Epoch 110/500 33/33 [==============================] - 0s 10ms/step - loss: 4.6982 - mae: 5.1778 - val_loss: 6.3100 - val_mae: 6.7905 Epoch 111/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6685 - mae: 5.1485 - val_loss: 6.3300 - val_mae: 6.8107 Epoch 112/500 33/33 [==============================] - 0s 8ms/step - loss: 4.6339 - mae: 5.1145 - val_loss: 6.3569 - val_mae: 6.8386 Epoch 113/500 33/33 [==============================] - 0s 9ms/step - loss: 4.6043 - mae: 5.0852 - val_loss: 6.4010 - val_mae: 6.8836 Epoch 114/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5846 - mae: 5.0641 - val_loss: 6.4428 - val_mae: 6.9257 Epoch 115/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5718 - mae: 5.0500 - val_loss: 6.4661 - val_mae: 6.9491 Epoch 116/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5643 - mae: 5.0413 - val_loss: 6.4823 - val_mae: 6.9654 Epoch 117/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5600 - mae: 5.0360 - val_loss: 6.4904 - val_mae: 6.9733 Epoch 118/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5581 - mae: 5.0337 - val_loss: 6.4945 - val_mae: 6.9773 Epoch 119/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5579 - mae: 5.0335 - val_loss: 6.4901 - val_mae: 6.9728 Epoch 120/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5590 - mae: 5.0347 - val_loss: 6.4773 - val_mae: 6.9598 Epoch 121/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5624 - mae: 5.0382 - val_loss: 6.4616 - val_mae: 6.9439 Epoch 122/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5668 - mae: 5.0430 - val_loss: 6.4371 - val_mae: 6.9190 Epoch 123/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5730 - mae: 5.0498 - val_loss: 6.3981 - val_mae: 6.8795 Epoch 124/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5805 - mae: 5.0580 - val_loss: 6.3478 - val_mae: 6.8286 Epoch 125/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5904 - mae: 5.0684 - val_loss: 6.2879 - val_mae: 6.7684 Epoch 126/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5976 - mae: 5.0764 - val_loss: 6.2282 - val_mae: 6.7077 Epoch 127/500 33/33 [==============================] - 0s 10ms/step - loss: 4.6012 - mae: 5.0806 - val_loss: 6.1627 - val_mae: 6.6411 Epoch 128/500 33/33 [==============================] - 0s 10ms/step - loss: 4.6009 - mae: 5.0804 - val_loss: 6.1094 - val_mae: 6.5875 Epoch 129/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5965 - mae: 5.0763 - val_loss: 6.0683 - val_mae: 6.5465 Epoch 130/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5875 - mae: 5.0677 - val_loss: 6.0413 - val_mae: 6.5196 Epoch 131/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5753 - mae: 5.0564 - val_loss: 6.0272 - val_mae: 6.5055 Epoch 132/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5618 - mae: 5.0434 - val_loss: 6.0175 - val_mae: 6.4959 Epoch 133/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5496 - mae: 5.0313 - val_loss: 6.0180 - val_mae: 6.4965 Epoch 134/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5383 - mae: 5.0194 - val_loss: 6.0228 - val_mae: 6.5011 Epoch 135/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5299 - mae: 5.0103 - val_loss: 6.0303 - val_mae: 6.5088 Epoch 136/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5239 - mae: 5.0036 - val_loss: 6.0382 - val_mae: 6.5167 Epoch 137/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5194 - mae: 4.9986 - val_loss: 6.0417 - val_mae: 6.5203 Epoch 138/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5162 - mae: 4.9952 - val_loss: 6.0427 - val_mae: 6.5214 Epoch 139/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5142 - mae: 4.9928 - val_loss: 6.0470 - val_mae: 6.5257 Epoch 140/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5127 - mae: 4.9912 - val_loss: 6.0519 - val_mae: 6.5306 Epoch 141/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5115 - mae: 4.9899 - val_loss: 6.0559 - val_mae: 6.5347 Epoch 142/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5107 - mae: 4.9889 - val_loss: 6.0606 - val_mae: 6.5394 Epoch 143/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5103 - mae: 4.9883 - val_loss: 6.0643 - val_mae: 6.5432 Epoch 144/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5103 - mae: 4.9881 - val_loss: 6.0674 - val_mae: 6.5463 Epoch 145/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5104 - mae: 4.9883 - val_loss: 6.0692 - val_mae: 6.5482 Epoch 146/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5109 - mae: 4.9887 - val_loss: 6.0706 - val_mae: 6.5496 Epoch 147/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5115 - mae: 4.9893 - val_loss: 6.0712 - val_mae: 6.5502 Epoch 148/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5122 - mae: 4.9901 - val_loss: 6.0713 - val_mae: 6.5503 Epoch 149/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5131 - mae: 4.9911 - val_loss: 6.0721 - val_mae: 6.5511 Epoch 150/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5140 - mae: 4.9919 - val_loss: 6.0723 - val_mae: 6.5513 Epoch 151/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5148 - mae: 4.9928 - val_loss: 6.0716 - val_mae: 6.5506 Epoch 152/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5156 - mae: 4.9936 - val_loss: 6.0700 - val_mae: 6.5490 Epoch 153/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5162 - mae: 4.9942 - val_loss: 6.0678 - val_mae: 6.5468 Epoch 154/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5165 - mae: 4.9945 - val_loss: 6.0646 - val_mae: 6.5435 Epoch 155/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5166 - mae: 4.9946 - val_loss: 6.0612 - val_mae: 6.5401 Epoch 156/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5165 - mae: 4.9945 - val_loss: 6.0580 - val_mae: 6.5369 Epoch 157/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5162 - mae: 4.9943 - val_loss: 6.0549 - val_mae: 6.5338 Epoch 158/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5157 - mae: 4.9938 - val_loss: 6.0521 - val_mae: 6.5310 Epoch 159/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5150 - mae: 4.9931 - val_loss: 6.0492 - val_mae: 6.5281 Epoch 160/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5140 - mae: 4.9922 - val_loss: 6.0467 - val_mae: 6.5255 Epoch 161/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5128 - mae: 4.9909 - val_loss: 6.0446 - val_mae: 6.5234 Epoch 162/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5114 - mae: 4.9895 - val_loss: 6.0428 - val_mae: 6.5216 Epoch 163/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5098 - mae: 4.9879 - val_loss: 6.0417 - val_mae: 6.5205 Epoch 164/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5084 - mae: 4.9863 - val_loss: 6.0412 - val_mae: 6.5200 Epoch 165/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5068 - mae: 4.9847 - val_loss: 6.0411 - val_mae: 6.5199 Epoch 166/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5052 - mae: 4.9831 - val_loss: 6.0406 - val_mae: 6.5194 Epoch 167/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5038 - mae: 4.9817 - val_loss: 6.0396 - val_mae: 6.5183 Epoch 168/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5027 - mae: 4.9806 - val_loss: 6.0382 - val_mae: 6.5169 Epoch 169/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5019 - mae: 4.9798 - val_loss: 6.0358 - val_mae: 6.5145 Epoch 170/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5014 - mae: 4.9793 - val_loss: 6.0329 - val_mae: 6.5115 Epoch 171/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5013 - mae: 4.9792 - val_loss: 6.0294 - val_mae: 6.5080 Epoch 172/500 33/33 [==============================] - 0s 8ms/step - loss: 4.5015 - mae: 4.9795 - val_loss: 6.0242 - val_mae: 6.5028 Epoch 173/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5020 - mae: 4.9799 - val_loss: 6.0175 - val_mae: 6.4960 Epoch 174/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5026 - mae: 4.9805 - val_loss: 6.0088 - val_mae: 6.4873 Epoch 175/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5032 - mae: 4.9811 - val_loss: 5.9979 - val_mae: 6.4764 Epoch 176/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5040 - mae: 4.9818 - val_loss: 5.9853 - val_mae: 6.4636 Epoch 177/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5044 - mae: 4.9823 - val_loss: 5.9701 - val_mae: 6.4482 Epoch 178/500 33/33 [==============================] - 0s 9ms/step - loss: 4.5040 - mae: 4.9819 - val_loss: 5.9533 - val_mae: 6.4311 Epoch 179/500 33/33 [==============================] - 0s 10ms/step - loss: 4.5027 - mae: 4.9806 - val_loss: 5.9359 - val_mae: 6.4136 Epoch 180/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4995 - mae: 4.9776 - val_loss: 5.9199 - val_mae: 6.3974 Epoch 181/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4951 - mae: 4.9733 - val_loss: 5.9059 - val_mae: 6.3832 Epoch 182/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4896 - mae: 4.9680 - val_loss: 5.8927 - val_mae: 6.3698 Epoch 183/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4837 - mae: 4.9621 - val_loss: 5.8809 - val_mae: 6.3578 Epoch 184/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4777 - mae: 4.9560 - val_loss: 5.8717 - val_mae: 6.3484 Epoch 185/500 33/33 [==============================] - 0s 10ms/step - loss: 4.4720 - mae: 4.9500 - val_loss: 5.8624 - val_mae: 6.3390 Epoch 186/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4672 - mae: 4.9448 - val_loss: 5.8582 - val_mae: 6.3347 Epoch 187/500 33/33 [==============================] - 0s 10ms/step - loss: 4.4631 - mae: 4.9402 - val_loss: 5.8564 - val_mae: 6.3328 Epoch 188/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4601 - mae: 4.9368 - val_loss: 5.8563 - val_mae: 6.3327 Epoch 189/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4579 - mae: 4.9344 - val_loss: 5.8564 - val_mae: 6.3327 Epoch 190/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4564 - mae: 4.9328 - val_loss: 5.8560 - val_mae: 6.3323 Epoch 191/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4556 - mae: 4.9318 - val_loss: 5.8548 - val_mae: 6.3312 Epoch 192/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4551 - mae: 4.9313 - val_loss: 5.8525 - val_mae: 6.3288 Epoch 193/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4548 - mae: 4.9310 - val_loss: 5.8491 - val_mae: 6.3254 Epoch 194/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4544 - mae: 4.9306 - val_loss: 5.8446 - val_mae: 6.3208 Epoch 195/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4537 - mae: 4.9300 - val_loss: 5.8391 - val_mae: 6.3153 Epoch 196/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4526 - mae: 4.9290 - val_loss: 5.8333 - val_mae: 6.3094 Epoch 197/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4511 - mae: 4.9275 - val_loss: 5.8270 - val_mae: 6.3032 Epoch 198/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4492 - mae: 4.9257 - val_loss: 5.8213 - val_mae: 6.2974 Epoch 199/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4469 - mae: 4.9235 - val_loss: 5.8163 - val_mae: 6.2924 Epoch 200/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4443 - mae: 4.9209 - val_loss: 5.8118 - val_mae: 6.2880 Epoch 201/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4416 - mae: 4.9183 - val_loss: 5.8083 - val_mae: 6.2844 Epoch 202/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4390 - mae: 4.9157 - val_loss: 5.8055 - val_mae: 6.2816 Epoch 203/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4364 - mae: 4.9132 - val_loss: 5.8034 - val_mae: 6.2795 Epoch 204/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4339 - mae: 4.9108 - val_loss: 5.8022 - val_mae: 6.2783 Epoch 205/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4318 - mae: 4.9087 - val_loss: 5.8016 - val_mae: 6.2779 Epoch 206/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4299 - mae: 4.9069 - val_loss: 5.8016 - val_mae: 6.2778 Epoch 207/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4285 - mae: 4.9055 - val_loss: 5.8018 - val_mae: 6.2780 Epoch 208/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4274 - mae: 4.9044 - val_loss: 5.8021 - val_mae: 6.2784 Epoch 209/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4264 - mae: 4.9036 - val_loss: 5.8025 - val_mae: 6.2788 Epoch 210/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4256 - mae: 4.9028 - val_loss: 5.8029 - val_mae: 6.2792 Epoch 211/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4248 - mae: 4.9020 - val_loss: 5.8032 - val_mae: 6.2795 Epoch 212/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4241 - mae: 4.9013 - val_loss: 5.8038 - val_mae: 6.2801 Epoch 213/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4233 - mae: 4.9006 - val_loss: 5.8029 - val_mae: 6.2793 Epoch 214/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4226 - mae: 4.8999 - val_loss: 5.8050 - val_mae: 6.2813 Epoch 215/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4218 - mae: 4.8990 - val_loss: 5.7986 - val_mae: 6.2750 Epoch 216/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4212 - mae: 4.8985 - val_loss: 5.8157 - val_mae: 6.2919 Epoch 217/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4201 - mae: 4.8972 - val_loss: 5.7727 - val_mae: 6.2493 Epoch 218/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4212 - mae: 4.8983 - val_loss: 5.9284 - val_mae: 6.4064 Epoch 219/500 33/33 [==============================] - 0s 10ms/step - loss: 4.4185 - mae: 4.8952 - val_loss: 5.7362 - val_mae: 6.2127 Epoch 220/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4328 - mae: 4.9084 - val_loss: 5.8902 - val_mae: 6.3679 Epoch 221/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4235 - mae: 4.9010 - val_loss: 5.7129 - val_mae: 6.1900 Epoch 222/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4311 - mae: 4.9073 - val_loss: 6.0098 - val_mae: 6.4888 Epoch 223/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4156 - mae: 4.8929 - val_loss: 5.6693 - val_mae: 6.1468 Epoch 224/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4395 - mae: 4.9144 - val_loss: 5.7900 - val_mae: 6.2664 Epoch 225/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4231 - mae: 4.9005 - val_loss: 5.7993 - val_mae: 6.2760 Epoch 226/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4134 - mae: 4.8913 - val_loss: 5.6931 - val_mae: 6.1707 Epoch 227/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4151 - mae: 4.8917 - val_loss: 5.9471 - val_mae: 6.4258 Epoch 228/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4038 - mae: 4.8809 - val_loss: 5.6575 - val_mae: 6.1351 Epoch 229/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4280 - mae: 4.9032 - val_loss: 5.7819 - val_mae: 6.2583 Epoch 230/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4150 - mae: 4.8922 - val_loss: 5.7141 - val_mae: 6.1917 Epoch 231/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4116 - mae: 4.8888 - val_loss: 5.8809 - val_mae: 6.3589 Epoch 232/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4051 - mae: 4.8821 - val_loss: 5.6579 - val_mae: 6.1358 Epoch 233/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4193 - mae: 4.8951 - val_loss: 5.9148 - val_mae: 6.3928 Epoch 234/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4078 - mae: 4.8845 - val_loss: 5.6474 - val_mae: 6.1255 Epoch 235/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4232 - mae: 4.8992 - val_loss: 5.9270 - val_mae: 6.4054 Epoch 236/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4089 - mae: 4.8861 - val_loss: 5.6281 - val_mae: 6.1062 Epoch 237/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4231 - mae: 4.8992 - val_loss: 5.9298 - val_mae: 6.4085 Epoch 238/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4068 - mae: 4.8840 - val_loss: 5.6122 - val_mae: 6.0904 Epoch 239/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4212 - mae: 4.8975 - val_loss: 5.9023 - val_mae: 6.3803 Epoch 240/500 33/33 [==============================] - 0s 10ms/step - loss: 4.4049 - mae: 4.8820 - val_loss: 5.6073 - val_mae: 6.0855 Epoch 241/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4176 - mae: 4.8938 - val_loss: 5.9148 - val_mae: 6.3934 Epoch 242/500 33/33 [==============================] - 0s 10ms/step - loss: 4.4013 - mae: 4.8785 - val_loss: 5.5935 - val_mae: 6.0719 Epoch 243/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4166 - mae: 4.8931 - val_loss: 5.8631 - val_mae: 6.3408 Epoch 244/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4010 - mae: 4.8778 - val_loss: 5.5987 - val_mae: 6.0770 Epoch 245/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4129 - mae: 4.8891 - val_loss: 5.9199 - val_mae: 6.3987 Epoch 246/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3963 - mae: 4.8736 - val_loss: 5.5665 - val_mae: 6.0452 Epoch 247/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4155 - mae: 4.8922 - val_loss: 5.7719 - val_mae: 6.2486 Epoch 248/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3992 - mae: 4.8760 - val_loss: 5.5998 - val_mae: 6.0786 Epoch 249/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4054 - mae: 4.8818 - val_loss: 5.8961 - val_mae: 6.3743 Epoch 250/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3910 - mae: 4.8680 - val_loss: 5.5418 - val_mae: 6.0208 Epoch 251/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4153 - mae: 4.8921 - val_loss: 5.6909 - val_mae: 6.1687 Epoch 252/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3985 - mae: 4.8753 - val_loss: 5.7813 - val_mae: 6.2587 Epoch 253/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3902 - mae: 4.8669 - val_loss: 5.5799 - val_mae: 6.0586 Epoch 254/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4008 - mae: 4.8772 - val_loss: 5.8577 - val_mae: 6.3348 Epoch 255/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3838 - mae: 4.8605 - val_loss: 5.5146 - val_mae: 5.9941 Epoch 256/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4143 - mae: 4.8913 - val_loss: 5.6538 - val_mae: 6.1319 Epoch 257/500 33/33 [==============================] - 0s 8ms/step - loss: 4.4005 - mae: 4.8769 - val_loss: 5.8697 - val_mae: 6.3477 Epoch 258/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3891 - mae: 4.8664 - val_loss: 5.5332 - val_mae: 6.0121 Epoch 259/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4036 - mae: 4.8807 - val_loss: 5.8062 - val_mae: 6.2833 Epoch 260/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3877 - mae: 4.8640 - val_loss: 5.5666 - val_mae: 6.0453 Epoch 261/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4003 - mae: 4.8769 - val_loss: 5.8523 - val_mae: 6.3293 Epoch 262/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3829 - mae: 4.8599 - val_loss: 5.4939 - val_mae: 5.9740 Epoch 263/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4093 - mae: 4.8868 - val_loss: 5.6404 - val_mae: 6.1187 Epoch 264/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3950 - mae: 4.8714 - val_loss: 5.8611 - val_mae: 6.3390 Epoch 265/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3833 - mae: 4.8604 - val_loss: 5.5161 - val_mae: 5.9955 Epoch 266/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3982 - mae: 4.8754 - val_loss: 5.7901 - val_mae: 6.2672 Epoch 267/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3826 - mae: 4.8593 - val_loss: 5.5579 - val_mae: 6.0365 Epoch 268/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3953 - mae: 4.8719 - val_loss: 5.8217 - val_mae: 6.2986 Epoch 269/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3785 - mae: 4.8553 - val_loss: 5.4729 - val_mae: 5.9536 Epoch 270/500 33/33 [==============================] - 0s 9ms/step - loss: 4.4056 - mae: 4.8833 - val_loss: 5.6249 - val_mae: 6.1032 Epoch 271/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3925 - mae: 4.8688 - val_loss: 5.8538 - val_mae: 6.3316 Epoch 272/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3798 - mae: 4.8570 - val_loss: 5.4911 - val_mae: 5.9712 Epoch 273/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3963 - mae: 4.8739 - val_loss: 5.6973 - val_mae: 6.1757 Epoch 274/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3792 - mae: 4.8561 - val_loss: 5.6411 - val_mae: 6.1204 Epoch 275/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3756 - mae: 4.8511 - val_loss: 5.8305 - val_mae: 6.3083 Epoch 276/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3705 - mae: 4.8467 - val_loss: 5.5182 - val_mae: 5.9975 Epoch 277/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3883 - mae: 4.8648 - val_loss: 5.8787 - val_mae: 6.3569 Epoch 278/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3715 - mae: 4.8485 - val_loss: 5.4936 - val_mae: 5.9738 Epoch 279/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3958 - mae: 4.8736 - val_loss: 5.7129 - val_mae: 6.1907 Epoch 280/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3797 - mae: 4.8570 - val_loss: 5.7567 - val_mae: 6.2349 Epoch 281/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3726 - mae: 4.8494 - val_loss: 5.6258 - val_mae: 6.1051 Epoch 282/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3766 - mae: 4.8520 - val_loss: 5.8099 - val_mae: 6.2881 Epoch 283/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3697 - mae: 4.8462 - val_loss: 5.4826 - val_mae: 5.9632 Epoch 284/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3919 - mae: 4.8696 - val_loss: 5.7645 - val_mae: 6.2418 Epoch 285/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3755 - mae: 4.8527 - val_loss: 5.5944 - val_mae: 6.0735 Epoch 286/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3814 - mae: 4.8583 - val_loss: 5.9171 - val_mae: 6.3965 Epoch 287/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3765 - mae: 4.8532 - val_loss: 5.5245 - val_mae: 6.0040 Epoch 288/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3975 - mae: 4.8743 - val_loss: 5.9291 - val_mae: 6.4089 Epoch 289/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3789 - mae: 4.8567 - val_loss: 5.4857 - val_mae: 5.9660 Epoch 290/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3969 - mae: 4.8744 - val_loss: 5.9073 - val_mae: 6.3867 Epoch 291/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3756 - mae: 4.8530 - val_loss: 5.5048 - val_mae: 5.9843 Epoch 292/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3922 - mae: 4.8694 - val_loss: 5.8508 - val_mae: 6.3289 Epoch 293/500 33/33 [==============================] - 0s 10ms/step - loss: 4.3731 - mae: 4.8507 - val_loss: 5.4380 - val_mae: 5.9190 Epoch 294/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3928 - mae: 4.8708 - val_loss: 5.6497 - val_mae: 6.1286 Epoch 295/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3730 - mae: 4.8493 - val_loss: 5.8301 - val_mae: 6.3090 Epoch 296/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3640 - mae: 4.8408 - val_loss: 5.4721 - val_mae: 5.9526 Epoch 297/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3813 - mae: 4.8575 - val_loss: 5.7477 - val_mae: 6.2272 Epoch 298/500 33/33 [==============================] - 0s 10ms/step - loss: 4.3620 - mae: 4.8386 - val_loss: 5.4074 - val_mae: 5.8885 Epoch 299/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3883 - mae: 4.8664 - val_loss: 5.5808 - val_mae: 6.0597 Epoch 300/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3765 - mae: 4.8525 - val_loss: 5.8120 - val_mae: 6.2909 Epoch 301/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3640 - mae: 4.8410 - val_loss: 5.4315 - val_mae: 5.9126 Epoch 302/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3833 - mae: 4.8613 - val_loss: 5.6272 - val_mae: 6.1061 Epoch 303/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3668 - mae: 4.8427 - val_loss: 5.8379 - val_mae: 6.3171 Epoch 304/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3587 - mae: 4.8353 - val_loss: 5.4603 - val_mae: 5.9413 Epoch 305/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3762 - mae: 4.8527 - val_loss: 5.8912 - val_mae: 6.3709 Epoch 306/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3583 - mae: 4.8353 - val_loss: 5.4716 - val_mae: 5.9526 Epoch 307/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3789 - mae: 4.8562 - val_loss: 5.8784 - val_mae: 6.3578 Epoch 308/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3627 - mae: 4.8398 - val_loss: 5.5055 - val_mae: 5.9855 Epoch 309/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3816 - mae: 4.8585 - val_loss: 5.8589 - val_mae: 6.3374 Epoch 310/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3644 - mae: 4.8420 - val_loss: 5.4310 - val_mae: 5.9122 Epoch 311/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3863 - mae: 4.8645 - val_loss: 5.6133 - val_mae: 6.0920 Epoch 312/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3716 - mae: 4.8478 - val_loss: 5.8520 - val_mae: 6.3310 Epoch 313/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3610 - mae: 4.8381 - val_loss: 5.4429 - val_mae: 5.9240 Epoch 314/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3767 - mae: 4.8543 - val_loss: 5.7863 - val_mae: 6.2644 Epoch 315/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3571 - mae: 4.8337 - val_loss: 5.5092 - val_mae: 5.9896 Epoch 316/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3706 - mae: 4.8491 - val_loss: 5.8256 - val_mae: 6.3029 Epoch 317/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3702 - mae: 4.8467 - val_loss: 5.5406 - val_mae: 6.0201 Epoch 318/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3897 - mae: 4.8682 - val_loss: 5.6561 - val_mae: 6.1353 Epoch 319/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3875 - mae: 4.8628 - val_loss: 5.8954 - val_mae: 6.3752 Epoch 320/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3735 - mae: 4.8520 - val_loss: 5.4238 - val_mae: 5.9047 Epoch 321/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3812 - mae: 4.8608 - val_loss: 5.6021 - val_mae: 6.0809 Epoch 322/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3761 - mae: 4.8525 - val_loss: 5.8315 - val_mae: 6.3110 Epoch 323/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3602 - mae: 4.8371 - val_loss: 5.6078 - val_mae: 6.0888 Epoch 324/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3559 - mae: 4.8307 - val_loss: 5.7482 - val_mae: 6.2291 Epoch 325/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3485 - mae: 4.8243 - val_loss: 5.3694 - val_mae: 5.8501 Epoch 326/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3674 - mae: 4.8438 - val_loss: 5.7670 - val_mae: 6.2478 Epoch 327/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3451 - mae: 4.8215 - val_loss: 5.3590 - val_mae: 5.8397 Epoch 328/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3642 - mae: 4.8410 - val_loss: 5.7609 - val_mae: 6.2407 Epoch 329/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3429 - mae: 4.8189 - val_loss: 5.4254 - val_mae: 5.9064 Epoch 330/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3637 - mae: 4.8425 - val_loss: 5.8844 - val_mae: 6.3627 Epoch 331/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3602 - mae: 4.8366 - val_loss: 5.4968 - val_mae: 5.9774 Epoch 332/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3827 - mae: 4.8587 - val_loss: 5.9351 - val_mae: 6.4164 Epoch 333/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3614 - mae: 4.8394 - val_loss: 5.4296 - val_mae: 5.9106 Epoch 334/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3753 - mae: 4.8520 - val_loss: 5.8041 - val_mae: 6.2844 Epoch 335/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3539 - mae: 4.8307 - val_loss: 5.3645 - val_mae: 5.8453 Epoch 336/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3684 - mae: 4.8460 - val_loss: 5.6055 - val_mae: 6.0858 Epoch 337/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3494 - mae: 4.8253 - val_loss: 5.7242 - val_mae: 6.2057 Epoch 338/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3411 - mae: 4.8169 - val_loss: 5.3527 - val_mae: 5.8333 Epoch 339/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3564 - mae: 4.8329 - val_loss: 5.7297 - val_mae: 6.2096 Epoch 340/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3352 - mae: 4.8106 - val_loss: 5.4634 - val_mae: 5.9452 Epoch 341/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3498 - mae: 4.8272 - val_loss: 5.9414 - val_mae: 6.4214 Epoch 342/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3507 - mae: 4.8266 - val_loss: 5.4842 - val_mae: 5.9652 Epoch 343/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3758 - mae: 4.8528 - val_loss: 5.8704 - val_mae: 6.3505 Epoch 344/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3545 - mae: 4.8311 - val_loss: 5.6385 - val_mae: 6.1197 Epoch 345/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3552 - mae: 4.8317 - val_loss: 5.9329 - val_mae: 6.4145 Epoch 346/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3509 - mae: 4.8264 - val_loss: 5.4610 - val_mae: 5.9418 Epoch 347/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3727 - mae: 4.8492 - val_loss: 5.8548 - val_mae: 6.3345 Epoch 348/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3551 - mae: 4.8325 - val_loss: 5.3830 - val_mae: 5.8639 Epoch 349/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3694 - mae: 4.8472 - val_loss: 5.6872 - val_mae: 6.1672 Epoch 350/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3466 - mae: 4.8224 - val_loss: 5.8296 - val_mae: 6.3102 Epoch 351/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3387 - mae: 4.8148 - val_loss: 5.3893 - val_mae: 5.8702 Epoch 352/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3575 - mae: 4.8340 - val_loss: 5.6360 - val_mae: 6.1182 Epoch 353/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3399 - mae: 4.8165 - val_loss: 5.3128 - val_mae: 5.7931 Epoch 354/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3586 - mae: 4.8351 - val_loss: 5.4946 - val_mae: 5.9757 Epoch 355/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3503 - mae: 4.8260 - val_loss: 5.6496 - val_mae: 6.1317 Epoch 356/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3380 - mae: 4.8146 - val_loss: 5.3118 - val_mae: 5.7921 Epoch 357/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3572 - mae: 4.8335 - val_loss: 5.5014 - val_mae: 5.9824 Epoch 358/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3467 - mae: 4.8219 - val_loss: 5.8873 - val_mae: 6.3679 Epoch 359/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3297 - mae: 4.8054 - val_loss: 5.3663 - val_mae: 5.8471 Epoch 360/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3511 - mae: 4.8273 - val_loss: 5.5324 - val_mae: 6.0132 Epoch 361/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3453 - mae: 4.8208 - val_loss: 5.7932 - val_mae: 6.2747 Epoch 362/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3344 - mae: 4.8108 - val_loss: 5.3412 - val_mae: 5.8221 Epoch 363/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3600 - mae: 4.8367 - val_loss: 5.5459 - val_mae: 6.0261 Epoch 364/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3503 - mae: 4.8257 - val_loss: 5.9607 - val_mae: 6.4429 Epoch 365/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3361 - mae: 4.8116 - val_loss: 5.4774 - val_mae: 5.9584 Epoch 366/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3528 - mae: 4.8287 - val_loss: 5.9953 - val_mae: 6.4779 Epoch 367/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3369 - mae: 4.8130 - val_loss: 5.4730 - val_mae: 5.9542 Epoch 368/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3559 - mae: 4.8329 - val_loss: 5.7851 - val_mae: 6.2646 Epoch 369/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3404 - mae: 4.8162 - val_loss: 5.9412 - val_mae: 6.4234 Epoch 370/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3348 - mae: 4.8103 - val_loss: 5.4829 - val_mae: 5.9637 Epoch 371/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3534 - mae: 4.8320 - val_loss: 6.0937 - val_mae: 6.5762 Epoch 372/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3477 - mae: 4.8237 - val_loss: 5.5160 - val_mae: 5.9967 Epoch 373/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3742 - mae: 4.8511 - val_loss: 5.6867 - val_mae: 6.1668 Epoch 374/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3611 - mae: 4.8371 - val_loss: 6.0491 - val_mae: 6.5315 Epoch 375/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3495 - mae: 4.8269 - val_loss: 5.4618 - val_mae: 5.9428 Epoch 376/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3602 - mae: 4.8369 - val_loss: 6.0142 - val_mae: 6.4961 Epoch 377/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3405 - mae: 4.8171 - val_loss: 5.3962 - val_mae: 5.8772 Epoch 378/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3597 - mae: 4.8370 - val_loss: 5.6138 - val_mae: 6.0944 Epoch 379/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3443 - mae: 4.8196 - val_loss: 5.9376 - val_mae: 6.4186 Epoch 380/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3359 - mae: 4.8124 - val_loss: 5.3970 - val_mae: 5.8778 Epoch 381/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3506 - mae: 4.8271 - val_loss: 5.7479 - val_mae: 6.2284 Epoch 382/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3313 - mae: 4.8068 - val_loss: 5.8582 - val_mae: 6.3388 Epoch 383/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3247 - mae: 4.8000 - val_loss: 5.4454 - val_mae: 5.9271 Epoch 384/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3470 - mae: 4.8253 - val_loss: 5.6152 - val_mae: 6.0942 Epoch 385/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3660 - mae: 4.8413 - val_loss: 5.6168 - val_mae: 6.0968 Epoch 386/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3726 - mae: 4.8503 - val_loss: 5.5579 - val_mae: 6.0379 Epoch 387/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3891 - mae: 4.8645 - val_loss: 6.1157 - val_mae: 6.5980 Epoch 388/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3646 - mae: 4.8437 - val_loss: 5.4714 - val_mae: 5.9526 Epoch 389/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3568 - mae: 4.8343 - val_loss: 5.5565 - val_mae: 6.0364 Epoch 390/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3832 - mae: 4.8583 - val_loss: 5.5660 - val_mae: 6.0456 Epoch 391/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3874 - mae: 4.8627 - val_loss: 6.0309 - val_mae: 6.5137 Epoch 392/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3495 - mae: 4.8284 - val_loss: 5.6924 - val_mae: 6.1744 Epoch 393/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3307 - mae: 4.8071 - val_loss: 5.7216 - val_mae: 6.2038 Epoch 394/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3221 - mae: 4.7975 - val_loss: 5.2578 - val_mae: 5.7384 Epoch 395/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3325 - mae: 4.8077 - val_loss: 5.7803 - val_mae: 6.2625 Epoch 396/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3128 - mae: 4.7883 - val_loss: 5.2736 - val_mae: 5.7536 Epoch 397/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3322 - mae: 4.8078 - val_loss: 5.4412 - val_mae: 5.9230 Epoch 398/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3186 - mae: 4.7956 - val_loss: 5.1973 - val_mae: 5.6805 Epoch 399/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3305 - mae: 4.8048 - val_loss: 5.3617 - val_mae: 5.8422 Epoch 400/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3199 - mae: 4.7943 - val_loss: 5.7305 - val_mae: 6.2120 Epoch 401/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3076 - mae: 4.7813 - val_loss: 5.4890 - val_mae: 5.9707 Epoch 402/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3253 - mae: 4.7981 - val_loss: 5.5121 - val_mae: 5.9950 Epoch 403/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3235 - mae: 4.8008 - val_loss: 5.2498 - val_mae: 5.7317 Epoch 404/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3329 - mae: 4.8075 - val_loss: 5.4437 - val_mae: 5.9255 Epoch 405/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3333 - mae: 4.8091 - val_loss: 5.5039 - val_mae: 5.9853 Epoch 406/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3363 - mae: 4.8116 - val_loss: 5.9174 - val_mae: 6.3979 Epoch 407/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3257 - mae: 4.8026 - val_loss: 5.3420 - val_mae: 5.8225 Epoch 408/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3422 - mae: 4.8174 - val_loss: 5.5446 - val_mae: 6.0251 Epoch 409/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3368 - mae: 4.8125 - val_loss: 5.6004 - val_mae: 6.0809 Epoch 410/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3368 - mae: 4.8120 - val_loss: 6.1096 - val_mae: 6.5915 Epoch 411/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3278 - mae: 4.8043 - val_loss: 5.3977 - val_mae: 5.8789 Epoch 412/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3491 - mae: 4.8246 - val_loss: 5.5998 - val_mae: 6.0797 Epoch 413/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3433 - mae: 4.8189 - val_loss: 5.9099 - val_mae: 6.3907 Epoch 414/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3309 - mae: 4.8068 - val_loss: 5.9607 - val_mae: 6.4431 Epoch 415/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3266 - mae: 4.8023 - val_loss: 5.8417 - val_mae: 6.3222 Epoch 416/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3264 - mae: 4.8014 - val_loss: 6.0974 - val_mae: 6.5807 Epoch 417/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3257 - mae: 4.8011 - val_loss: 5.4841 - val_mae: 5.9652 Epoch 418/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3451 - mae: 4.8209 - val_loss: 5.6730 - val_mae: 6.1538 Epoch 419/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3414 - mae: 4.8166 - val_loss: 6.2043 - val_mae: 6.6888 Epoch 420/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3363 - mae: 4.8130 - val_loss: 5.4771 - val_mae: 5.9583 Epoch 421/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3563 - mae: 4.8325 - val_loss: 5.6312 - val_mae: 6.1114 Epoch 422/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3517 - mae: 4.8268 - val_loss: 6.2350 - val_mae: 6.7200 Epoch 423/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3416 - mae: 4.8186 - val_loss: 5.4807 - val_mae: 5.9615 Epoch 424/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3541 - mae: 4.8303 - val_loss: 5.8298 - val_mae: 6.3096 Epoch 425/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3378 - mae: 4.8130 - val_loss: 6.0820 - val_mae: 6.5654 Epoch 426/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3302 - mae: 4.8064 - val_loss: 5.4764 - val_mae: 5.9575 Epoch 427/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3409 - mae: 4.8163 - val_loss: 6.1405 - val_mae: 6.6241 Epoch 428/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3310 - mae: 4.8072 - val_loss: 5.4444 - val_mae: 5.9256 Epoch 429/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3528 - mae: 4.8287 - val_loss: 5.6370 - val_mae: 6.1177 Epoch 430/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3439 - mae: 4.8188 - val_loss: 6.1185 - val_mae: 6.6019 Epoch 431/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3367 - mae: 4.8132 - val_loss: 5.4667 - val_mae: 5.9477 Epoch 432/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3483 - mae: 4.8233 - val_loss: 6.1034 - val_mae: 6.5864 Epoch 433/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3312 - mae: 4.8080 - val_loss: 5.4110 - val_mae: 5.8918 Epoch 434/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3464 - mae: 4.8224 - val_loss: 5.6782 - val_mae: 6.1594 Epoch 435/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3333 - mae: 4.8080 - val_loss: 6.0268 - val_mae: 6.5096 Epoch 436/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3274 - mae: 4.8035 - val_loss: 5.4247 - val_mae: 5.9056 Epoch 437/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3396 - mae: 4.8145 - val_loss: 6.0182 - val_mae: 6.5009 Epoch 438/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3241 - mae: 4.8005 - val_loss: 5.3947 - val_mae: 5.8753 Epoch 439/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3399 - mae: 4.8159 - val_loss: 5.6825 - val_mae: 6.1637 Epoch 440/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3274 - mae: 4.8021 - val_loss: 5.9878 - val_mae: 6.4703 Epoch 441/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3232 - mae: 4.7990 - val_loss: 5.4096 - val_mae: 5.8903 Epoch 442/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3364 - mae: 4.8116 - val_loss: 5.9934 - val_mae: 6.4760 Epoch 443/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3213 - mae: 4.7973 - val_loss: 5.4028 - val_mae: 5.8835 Epoch 444/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3374 - mae: 4.8134 - val_loss: 5.8538 - val_mae: 6.3344 Epoch 445/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3210 - mae: 4.7963 - val_loss: 5.6520 - val_mae: 6.1340 Epoch 446/500 33/33 [==============================] - 0s 9ms/step - loss: 4.3231 - mae: 4.7979 - val_loss: 5.9736 - val_mae: 6.4557 Epoch 447/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3213 - mae: 4.7968 - val_loss: 5.3953 - val_mae: 5.8759 Epoch 448/500 33/33 [==============================] - 0s 8ms/step - loss: 4.3360 - mae: 4.8122 - val_loss: 5.8250 - val_mae: 6.3059
Out[8]:
<tensorflow.python.keras.callbacks.History at 0x7f8a5b9c8f98>
In [9]:
model = keras.models.load_model("my_checkpoint.h5")
In [10]:
rnn_forecast = model.predict(series[np.newaxis, :, np.newaxis])
rnn_forecast = rnn_forecast[0, split_time - 1:-1, 0]
In [11]:
plt.figure(figsize=(10, 6))
plot_series(time_valid, x_valid)
plot_series(time_valid, rnn_forecast)
In [12]:
keras.metrics.mean_absolute_error(x_valid, rnn_forecast).numpy()
Out[12]:
5.788751