** Training a Simple TensorFlow Lite for Microcontrollers model **¶
This notebook demonstrates the process of training a 2.5 kB model using TensorFlow and converting it for use with TensorFlow Lite for Microcontrollers.
Deep learning networks learn to model patterns in underlying data. Here, we're going to train a network to model data generated by a sine function. This will result in a model that can take a value, x, and predict its sine, y.
The model created in this notebook is used in the hello_world example for TensorFlow Lite for MicroControllers.
Configure Defaults¶
In [1]:
# Define paths to model files
import os
MODELS_DIR = 'models/'
if not os.path.exists(MODELS_DIR):
os.mkdir(MODELS_DIR)
MODEL_TF = MODELS_DIR + 'model'
MODEL_NO_QUANT_TFLITE = MODELS_DIR + 'model_no_quant.tflite'
MODEL_TFLITE = MODELS_DIR + 'model.tflite'
MODEL_TFLITE_MICRO = MODELS_DIR + 'model.cc'
Setupenv ¶
In [2]:
#! pip install tensorflow==2.4.0
In [2]:
# TensorFlow is an open source machine learning library
import tensorflow as tf
# Keras is TensorFlow's high-level API for deep learning
from tensorflow import keras
# Numpy is a math library
import numpy as np
# Pandas is a data manipulation library
import pandas as pd
# Matplotlib is a graphing library
import matplotlib.pyplot as plt
# Math is Python's math library
import math
# Set seed for experiment reproducibility
seed = 1
np.random.seed(seed)
tf.random.set_seed(seed)
In [3]:
# Number of sample datapoints
SAMPLES = 1000
# Generate a uniformly distributed set of random numbers in the range from
# 0 to 2π, which covers a complete sine wave oscillation
x_values = np.random.uniform(
low=0, high=2*math.pi, size=SAMPLES).astype(np.float32)
# Shuffle the values to guarantee they're not in order
np.random.shuffle(x_values)
# Calculate the corresponding sine values
y_values = np.sin(x_values).astype(np.float32)
# Plot our data. The 'b.' argument tells the library to print blue dots.
plt.plot(x_values, y_values, 'b.')
plt.show()
2. Add Noise ¶
Since it was generated directly by the sine function, our data fits a nice, smooth curve.
However, machine learning models are good at extracting underlying meaning from messy, real world data. To demonstrate this, we can add some noise to our data to approximate something more life-like.
In the following cell, we'll add some random noise to each value, then draw a new graph:
In [4]:
# Add a small random number to each y value
y_values += 0.1 * np.random.randn(*y_values.shape)
# Plot our data
plt.plot(x_values, y_values, 'b.')
plt.show()
3. Split the Data ¶
We now have a noisy dataset that approximates real world data. We'll be using this to train our model.
To evaluate the accuracy of the model we train, we'll need to compare its predictions to real data and check how well they match up. This evaluation happens during training (where it is referred to as validation) and after training (referred to as testing) It's important in both cases that we use fresh data that was not already used to train the model.
The data is split as follows:
Training: 60% Validation: 20% Testing: 20% The following code will split our data and then plots each set as a different color:
In [5]:
# We'll use 60% of our data for training and 20% for testing. The remaining 20%
# will be used for validation. Calculate the indices of each section.
TRAIN_SPLIT = int(0.6 * SAMPLES)
TEST_SPLIT = int(0.2 * SAMPLES + TRAIN_SPLIT)
# Use np.split to chop our data into three parts.
# The second argument to np.split is an array of indices where the data will be
# split. We provide two indices, so the data will be divided into three chunks.
x_train, x_test, x_validate = np.split(x_values, [TRAIN_SPLIT, TEST_SPLIT])
y_train, y_test, y_validate = np.split(y_values, [TRAIN_SPLIT, TEST_SPLIT])
# Double check that our splits add up correctly
assert (x_train.size + x_validate.size + x_test.size) == SAMPLES
# Plot the data in each partition in different colors:
plt.plot(x_train, y_train, 'b.', label="Train")
plt.plot(x_test, y_test, 'r.', label="Test")
plt.plot(x_validate, y_validate, 'y.', label="Validate")
plt.legend()
plt.show()
Training ¶
1. Design the Model ¶
We're going to build a simple neural network model that will take an input value (in this case, x) and use it to predict a numeric output value (the sine of x). This type of problem is called a regression. It will use layers of neurons to attempt to learn any patterns underlying the training data, so it can make predictions.
To begin with, we'll define two layers. The first layer takes a single input (our x value) and runs it through 8 neurons. Based on this input, each neuron will become activated to a certain degree based on its internal state (its weight and bias values). A neuron's degree of activation is expressed as a number.
The activation numbers from our first layer will be fed as inputs to our second layer, which is a single neuron. It will apply its own weights and bias to these inputs and calculate its own activation, which will be output as our y value.
Note: To learn more about how neural networks function, you can explore the Learn TensorFlow codelabs.
The code in the following cell defines our model using Keras, TensorFlow's high-level API for creating deep learning networks. Once the network is defined, we compile it, specifying parameters that determine how it will be trained:
In [6]:
# We'll use Keras to create a simple model architecture
model_1 = tf.keras.Sequential()
# First layer takes a scalar input and feeds it through 8 "neurons". The
# neurons decide whether to activate based on the 'relu' activation function.
model_1.add(keras.layers.Dense(8, activation='relu', input_shape=(1,)))
# Final layer is a single neuron, since we want to output a single value
model_1.add(keras.layers.Dense(1))
# Compile the model using the standard 'adam' optimizer and the mean squared error or 'mse' loss function for regression.
model_1.compile(optimizer='adam', loss='mse', metrics=['mae'])
2. Train the Model¶
Once we've defined the model, we can use our data to train it. Training involves passing an x value into the neural network, checking how far the network's output deviates from the expected y value, and adjusting the neurons' weights and biases so that the output is more likely to be correct the next time.
Training runs this process on the full dataset multiple times, and each full run-through is known as an epoch. The number of epochs to run during training is a parameter we can set.
During each epoch, data is run through the network in multiple batches. Each batch, several pieces of data are passed into the network, producing output values. These outputs' correctness is measured in aggregate and the network's weights and biases are adjusted accordingly, once per batch. The batch size is also a parameter we can set.
The code in the following cell uses the x and y values from our training data to train the model. It runs for 500 epochs, with 64 pieces of data in each batch. We also pass in some data for validation. As you will see when you run the cell, training can take a while to complete:
In [7]:
# Train the model on our training data while validating on our validation set
history_1 = model_1.fit(x_train, y_train, epochs=500, batch_size=64,
validation_data=(x_validate, y_validate))
Epoch 1/500 10/10 [==============================] - 16s 156ms/step - loss: 0.6922 - mae: 0.6886 - val_loss: 0.6401 - val_mae: 0.6504 Epoch 2/500 10/10 [==============================] - 0s 39ms/step - loss: 0.5949 - mae: 0.6247 - val_loss: 0.5587 - val_mae: 0.6031 Epoch 3/500 10/10 [==============================] - 0s 39ms/step - loss: 0.5225 - mae: 0.5816 - val_loss: 0.5014 - val_mae: 0.5763 Epoch 4/500 10/10 [==============================] - 0s 38ms/step - loss: 0.4734 - mae: 0.5554 - val_loss: 0.4632 - val_mae: 0.5615 Epoch 5/500 10/10 [==============================] - 0s 41ms/step - loss: 0.4395 - mae: 0.5390 - val_loss: 0.4386 - val_mae: 0.5536 Epoch 6/500 10/10 [==============================] - 0s 40ms/step - loss: 0.4178 - mae: 0.5305 - val_loss: 0.4227 - val_mae: 0.5490 Epoch 7/500 10/10 [==============================] - 0s 40ms/step - loss: 0.4034 - mae: 0.5245 - val_loss: 0.4125 - val_mae: 0.5464 Epoch 8/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3942 - mae: 0.5224 - val_loss: 0.4060 - val_mae: 0.5452 Epoch 9/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3879 - mae: 0.5215 - val_loss: 0.4014 - val_mae: 0.5440 Epoch 10/500 10/10 [==============================] - 0s 41ms/step - loss: 0.3835 - mae: 0.5216 - val_loss: 0.3979 - val_mae: 0.5426 Epoch 11/500 10/10 [==============================] - 0s 41ms/step - loss: 0.3798 - mae: 0.5208 - val_loss: 0.3950 - val_mae: 0.5412 Epoch 12/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3769 - mae: 0.5199 - val_loss: 0.3921 - val_mae: 0.5399 Epoch 13/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3741 - mae: 0.5191 - val_loss: 0.3893 - val_mae: 0.5386 Epoch 14/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3711 - mae: 0.5178 - val_loss: 0.3865 - val_mae: 0.5371 Epoch 15/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3683 - mae: 0.5161 - val_loss: 0.3837 - val_mae: 0.5354 Epoch 16/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3655 - mae: 0.5144 - val_loss: 0.3808 - val_mae: 0.5336 Epoch 17/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3625 - mae: 0.5125 - val_loss: 0.3778 - val_mae: 0.5318 Epoch 18/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3596 - mae: 0.5104 - val_loss: 0.3747 - val_mae: 0.5297 Epoch 19/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3566 - mae: 0.5080 - val_loss: 0.3716 - val_mae: 0.5275 Epoch 20/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3537 - mae: 0.5060 - val_loss: 0.3686 - val_mae: 0.5258 Epoch 21/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3505 - mae: 0.5038 - val_loss: 0.3654 - val_mae: 0.5235 Epoch 22/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3475 - mae: 0.5014 - val_loss: 0.3622 - val_mae: 0.5214 Epoch 23/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3445 - mae: 0.4991 - val_loss: 0.3589 - val_mae: 0.5191 Epoch 24/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3413 - mae: 0.4968 - val_loss: 0.3558 - val_mae: 0.5172 Epoch 25/500 10/10 [==============================] - 0s 41ms/step - loss: 0.3382 - mae: 0.4947 - val_loss: 0.3524 - val_mae: 0.5150 Epoch 26/500 10/10 [==============================] - 0s 41ms/step - loss: 0.3350 - mae: 0.4922 - val_loss: 0.3492 - val_mae: 0.5128 Epoch 27/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3319 - mae: 0.4897 - val_loss: 0.3459 - val_mae: 0.5106 Epoch 28/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3288 - mae: 0.4876 - val_loss: 0.3427 - val_mae: 0.5086 Epoch 29/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3255 - mae: 0.4854 - val_loss: 0.3395 - val_mae: 0.5065 Epoch 30/500 10/10 [==============================] - 0s 37ms/step - loss: 0.3225 - mae: 0.4833 - val_loss: 0.3362 - val_mae: 0.5043 Epoch 31/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3194 - mae: 0.4811 - val_loss: 0.3330 - val_mae: 0.5022 Epoch 32/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3162 - mae: 0.4784 - val_loss: 0.3297 - val_mae: 0.4996 Epoch 33/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3132 - mae: 0.4758 - val_loss: 0.3264 - val_mae: 0.4972 Epoch 34/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3102 - mae: 0.4738 - val_loss: 0.3233 - val_mae: 0.4954 Epoch 35/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3070 - mae: 0.4713 - val_loss: 0.3201 - val_mae: 0.4929 Epoch 36/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3040 - mae: 0.4688 - val_loss: 0.3170 - val_mae: 0.4905 Epoch 37/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3010 - mae: 0.4666 - val_loss: 0.3139 - val_mae: 0.4885 Epoch 38/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2979 - mae: 0.4644 - val_loss: 0.3108 - val_mae: 0.4863 Epoch 39/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2949 - mae: 0.4623 - val_loss: 0.3078 - val_mae: 0.4843 Epoch 40/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2919 - mae: 0.4602 - val_loss: 0.3047 - val_mae: 0.4823 Epoch 41/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2889 - mae: 0.4584 - val_loss: 0.3017 - val_mae: 0.4804 Epoch 42/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2861 - mae: 0.4567 - val_loss: 0.2988 - val_mae: 0.4787 Epoch 43/500 10/10 [==============================] - 0s 37ms/step - loss: 0.2830 - mae: 0.4546 - val_loss: 0.2960 - val_mae: 0.4769 Epoch 44/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2802 - mae: 0.4527 - val_loss: 0.2931 - val_mae: 0.4748 Epoch 45/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2773 - mae: 0.4505 - val_loss: 0.2902 - val_mae: 0.4727 Epoch 46/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2747 - mae: 0.4483 - val_loss: 0.2872 - val_mae: 0.4697 Epoch 47/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2719 - mae: 0.4460 - val_loss: 0.2845 - val_mae: 0.4680 Epoch 48/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2691 - mae: 0.4441 - val_loss: 0.2818 - val_mae: 0.4661 Epoch 49/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2665 - mae: 0.4424 - val_loss: 0.2793 - val_mae: 0.4647 Epoch 50/500 10/10 [==============================] - 0s 37ms/step - loss: 0.2638 - mae: 0.4405 - val_loss: 0.2767 - val_mae: 0.4628 Epoch 51/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2613 - mae: 0.4385 - val_loss: 0.2740 - val_mae: 0.4604 Epoch 52/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2587 - mae: 0.4365 - val_loss: 0.2714 - val_mae: 0.4583 Epoch 53/500 10/10 [==============================] - 0s 37ms/step - loss: 0.2561 - mae: 0.4347 - val_loss: 0.2690 - val_mae: 0.4568 Epoch 54/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2536 - mae: 0.4330 - val_loss: 0.2664 - val_mae: 0.4545 Epoch 55/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2511 - mae: 0.4311 - val_loss: 0.2639 - val_mae: 0.4525 Epoch 56/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2487 - mae: 0.4295 - val_loss: 0.2614 - val_mae: 0.4507 Epoch 57/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2462 - mae: 0.4275 - val_loss: 0.2589 - val_mae: 0.4485 Epoch 58/500 10/10 [==============================] - 0s 37ms/step - loss: 0.2441 - mae: 0.4256 - val_loss: 0.2564 - val_mae: 0.4460 Epoch 59/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2416 - mae: 0.4237 - val_loss: 0.2542 - val_mae: 0.4446 Epoch 60/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2396 - mae: 0.4227 - val_loss: 0.2522 - val_mae: 0.4437 Epoch 61/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2371 - mae: 0.4209 - val_loss: 0.2500 - val_mae: 0.4418 Epoch 62/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2349 - mae: 0.4190 - val_loss: 0.2478 - val_mae: 0.4400 Epoch 63/500 10/10 [==============================] - 0s 42ms/step - loss: 0.2327 - mae: 0.4173 - val_loss: 0.2456 - val_mae: 0.4381 Epoch 64/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2308 - mae: 0.4157 - val_loss: 0.2433 - val_mae: 0.4360 Epoch 65/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2284 - mae: 0.4139 - val_loss: 0.2415 - val_mae: 0.4348 Epoch 66/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2265 - mae: 0.4124 - val_loss: 0.2393 - val_mae: 0.4329 Epoch 67/500 10/10 [==============================] - 0s 41ms/step - loss: 0.2245 - mae: 0.4108 - val_loss: 0.2373 - val_mae: 0.4312 Epoch 68/500 10/10 [==============================] - 0s 41ms/step - loss: 0.2226 - mae: 0.4093 - val_loss: 0.2353 - val_mae: 0.4295 Epoch 69/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2206 - mae: 0.4076 - val_loss: 0.2334 - val_mae: 0.4277 Epoch 70/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2188 - mae: 0.4061 - val_loss: 0.2316 - val_mae: 0.4261 Epoch 71/500 10/10 [==============================] - 0s 37ms/step - loss: 0.2170 - mae: 0.4046 - val_loss: 0.2298 - val_mae: 0.4244 Epoch 72/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2153 - mae: 0.4032 - val_loss: 0.2280 - val_mae: 0.4227 Epoch 73/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2136 - mae: 0.4020 - val_loss: 0.2265 - val_mae: 0.4219 Epoch 74/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2120 - mae: 0.4006 - val_loss: 0.2249 - val_mae: 0.4205 Epoch 75/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2104 - mae: 0.3995 - val_loss: 0.2235 - val_mae: 0.4198 Epoch 76/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2089 - mae: 0.3982 - val_loss: 0.2216 - val_mae: 0.4175 Epoch 77/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2072 - mae: 0.3965 - val_loss: 0.2200 - val_mae: 0.4157 Epoch 78/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2058 - mae: 0.3953 - val_loss: 0.2186 - val_mae: 0.4144 Epoch 79/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2043 - mae: 0.3938 - val_loss: 0.2171 - val_mae: 0.4128 Epoch 80/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2029 - mae: 0.3926 - val_loss: 0.2158 - val_mae: 0.4117 Epoch 81/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2016 - mae: 0.3915 - val_loss: 0.2146 - val_mae: 0.4109 Epoch 82/500 10/10 [==============================] - 0s 38ms/step - loss: 0.2004 - mae: 0.3907 - val_loss: 0.2135 - val_mae: 0.4104 Epoch 83/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1991 - mae: 0.3894 - val_loss: 0.2120 - val_mae: 0.4085 Epoch 84/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1979 - mae: 0.3882 - val_loss: 0.2109 - val_mae: 0.4074 Epoch 85/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1965 - mae: 0.3869 - val_loss: 0.2096 - val_mae: 0.4057 Epoch 86/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1955 - mae: 0.3856 - val_loss: 0.2083 - val_mae: 0.4042 Epoch 87/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1943 - mae: 0.3846 - val_loss: 0.2073 - val_mae: 0.4036 Epoch 88/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1933 - mae: 0.3835 - val_loss: 0.2062 - val_mae: 0.4023 Epoch 89/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1923 - mae: 0.3825 - val_loss: 0.2052 - val_mae: 0.4010 Epoch 90/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1912 - mae: 0.3815 - val_loss: 0.2042 - val_mae: 0.4001 Epoch 91/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1903 - mae: 0.3807 - val_loss: 0.2038 - val_mae: 0.4009 Epoch 92/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1896 - mae: 0.3802 - val_loss: 0.2030 - val_mae: 0.4003 Epoch 93/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1885 - mae: 0.3790 - val_loss: 0.2018 - val_mae: 0.3981 Epoch 94/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1876 - mae: 0.3779 - val_loss: 0.2008 - val_mae: 0.3963 Epoch 95/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1869 - mae: 0.3770 - val_loss: 0.2001 - val_mae: 0.3960 Epoch 96/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1862 - mae: 0.3763 - val_loss: 0.1994 - val_mae: 0.3953 Epoch 97/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1852 - mae: 0.3754 - val_loss: 0.1987 - val_mae: 0.3949 Epoch 98/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1845 - mae: 0.3745 - val_loss: 0.1979 - val_mae: 0.3935 Epoch 99/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1838 - mae: 0.3736 - val_loss: 0.1970 - val_mae: 0.3915 Epoch 100/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1833 - mae: 0.3729 - val_loss: 0.1964 - val_mae: 0.3908 Epoch 101/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1826 - mae: 0.3720 - val_loss: 0.1958 - val_mae: 0.3897 Epoch 102/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1820 - mae: 0.3712 - val_loss: 0.1953 - val_mae: 0.3897 Epoch 103/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1814 - mae: 0.3709 - val_loss: 0.1952 - val_mae: 0.3910 Epoch 104/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1807 - mae: 0.3702 - val_loss: 0.1945 - val_mae: 0.3898 Epoch 105/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1803 - mae: 0.3693 - val_loss: 0.1937 - val_mae: 0.3869 Epoch 106/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1798 - mae: 0.3686 - val_loss: 0.1932 - val_mae: 0.3869 Epoch 107/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1792 - mae: 0.3679 - val_loss: 0.1928 - val_mae: 0.3859 Epoch 108/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1788 - mae: 0.3674 - val_loss: 0.1925 - val_mae: 0.3863 Epoch 109/500 10/10 [==============================] - 0s 47ms/step - loss: 0.1783 - mae: 0.3669 - val_loss: 0.1922 - val_mae: 0.3862 Epoch 110/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1778 - mae: 0.3664 - val_loss: 0.1917 - val_mae: 0.3853 Epoch 111/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1774 - mae: 0.3658 - val_loss: 0.1913 - val_mae: 0.3846 Epoch 112/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1769 - mae: 0.3651 - val_loss: 0.1908 - val_mae: 0.3831 Epoch 113/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1766 - mae: 0.3644 - val_loss: 0.1904 - val_mae: 0.3823 Epoch 114/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1763 - mae: 0.3638 - val_loss: 0.1901 - val_mae: 0.3818 Epoch 115/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1758 - mae: 0.3634 - val_loss: 0.1899 - val_mae: 0.3820 Epoch 116/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1755 - mae: 0.3631 - val_loss: 0.1896 - val_mae: 0.3815 Epoch 117/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1754 - mae: 0.3629 - val_loss: 0.1894 - val_mae: 0.3816 Epoch 118/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1748 - mae: 0.3621 - val_loss: 0.1890 - val_mae: 0.3804 Epoch 119/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1747 - mae: 0.3616 - val_loss: 0.1887 - val_mae: 0.3792 Epoch 120/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1746 - mae: 0.3610 - val_loss: 0.1884 - val_mae: 0.3779 Epoch 121/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1742 - mae: 0.3607 - val_loss: 0.1883 - val_mae: 0.3789 Epoch 122/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1739 - mae: 0.3603 - val_loss: 0.1881 - val_mae: 0.3787 Epoch 123/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1739 - mae: 0.3604 - val_loss: 0.1881 - val_mae: 0.3794 Epoch 124/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1737 - mae: 0.3599 - val_loss: 0.1876 - val_mae: 0.3775 Epoch 125/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1731 - mae: 0.3591 - val_loss: 0.1877 - val_mae: 0.3784 Epoch 126/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1730 - mae: 0.3590 - val_loss: 0.1878 - val_mae: 0.3791 Epoch 127/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1729 - mae: 0.3590 - val_loss: 0.1877 - val_mae: 0.3789 Epoch 128/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1727 - mae: 0.3586 - val_loss: 0.1872 - val_mae: 0.3773 Epoch 129/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1724 - mae: 0.3578 - val_loss: 0.1869 - val_mae: 0.3758 Epoch 130/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1724 - mae: 0.3575 - val_loss: 0.1867 - val_mae: 0.3750 Epoch 131/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1723 - mae: 0.3572 - val_loss: 0.1867 - val_mae: 0.3760 Epoch 132/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1721 - mae: 0.3571 - val_loss: 0.1865 - val_mae: 0.3754 Epoch 133/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1722 - mae: 0.3573 - val_loss: 0.1869 - val_mae: 0.3772 Epoch 134/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1719 - mae: 0.3568 - val_loss: 0.1863 - val_mae: 0.3751 Epoch 135/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1716 - mae: 0.3563 - val_loss: 0.1862 - val_mae: 0.3744 Epoch 136/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1717 - mae: 0.3560 - val_loss: 0.1861 - val_mae: 0.3737 Epoch 137/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1715 - mae: 0.3556 - val_loss: 0.1860 - val_mae: 0.3739 Epoch 138/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1713 - mae: 0.3556 - val_loss: 0.1861 - val_mae: 0.3748 Epoch 139/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1715 - mae: 0.3559 - val_loss: 0.1863 - val_mae: 0.3759 Epoch 140/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1715 - mae: 0.3558 - val_loss: 0.1858 - val_mae: 0.3739 Epoch 141/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1711 - mae: 0.3552 - val_loss: 0.1859 - val_mae: 0.3743 Epoch 142/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1711 - mae: 0.3554 - val_loss: 0.1861 - val_mae: 0.3754 Epoch 143/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1712 - mae: 0.3551 - val_loss: 0.1857 - val_mae: 0.3734 Epoch 144/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1709 - mae: 0.3546 - val_loss: 0.1855 - val_mae: 0.3728 Epoch 145/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1708 - mae: 0.3545 - val_loss: 0.1856 - val_mae: 0.3737 Epoch 146/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1707 - mae: 0.3544 - val_loss: 0.1856 - val_mae: 0.3736 Epoch 147/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1707 - mae: 0.3544 - val_loss: 0.1856 - val_mae: 0.3738 Epoch 148/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1706 - mae: 0.3541 - val_loss: 0.1853 - val_mae: 0.3719 Epoch 149/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1706 - mae: 0.3538 - val_loss: 0.1854 - val_mae: 0.3727 Epoch 150/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1706 - mae: 0.3538 - val_loss: 0.1853 - val_mae: 0.3723 Epoch 151/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1705 - mae: 0.3538 - val_loss: 0.1854 - val_mae: 0.3731 Epoch 152/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1705 - mae: 0.3537 - val_loss: 0.1852 - val_mae: 0.3716 Epoch 153/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1704 - mae: 0.3532 - val_loss: 0.1851 - val_mae: 0.3716 Epoch 154/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1703 - mae: 0.3531 - val_loss: 0.1851 - val_mae: 0.3711 Epoch 155/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1703 - mae: 0.3528 - val_loss: 0.1850 - val_mae: 0.3709 Epoch 156/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1706 - mae: 0.3534 - val_loss: 0.1853 - val_mae: 0.3728 Epoch 157/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1703 - mae: 0.3532 - val_loss: 0.1850 - val_mae: 0.3712 Epoch 158/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1702 - mae: 0.3528 - val_loss: 0.1850 - val_mae: 0.3714 Epoch 159/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1702 - mae: 0.3527 - val_loss: 0.1849 - val_mae: 0.3705 Epoch 160/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1702 - mae: 0.3526 - val_loss: 0.1850 - val_mae: 0.3712 Epoch 161/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1702 - mae: 0.3523 - val_loss: 0.1849 - val_mae: 0.3703 Epoch 162/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1701 - mae: 0.3521 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 163/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1704 - mae: 0.3520 - val_loss: 0.1848 - val_mae: 0.3694 Epoch 164/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1704 - mae: 0.3525 - val_loss: 0.1851 - val_mae: 0.3719 Epoch 165/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3521 - val_loss: 0.1848 - val_mae: 0.3698 Epoch 166/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1701 - mae: 0.3518 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 167/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3518 - val_loss: 0.1849 - val_mae: 0.3707 Epoch 168/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1703 - mae: 0.3528 - val_loss: 0.1853 - val_mae: 0.3724 Epoch 169/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1700 - mae: 0.3524 - val_loss: 0.1850 - val_mae: 0.3712 Epoch 170/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3517 - val_loss: 0.1847 - val_mae: 0.3693 Epoch 171/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3514 - val_loss: 0.1847 - val_mae: 0.3693 Epoch 172/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1700 - mae: 0.3518 - val_loss: 0.1849 - val_mae: 0.3710 Epoch 173/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3518 - val_loss: 0.1848 - val_mae: 0.3706 Epoch 174/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3517 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 175/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3512 - val_loss: 0.1846 - val_mae: 0.3680 Epoch 176/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1699 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 177/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1700 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 178/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3518 - val_loss: 0.1850 - val_mae: 0.3712 Epoch 179/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3518 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 180/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3513 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 181/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 182/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 183/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1847 - val_mae: 0.3698 Epoch 184/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1847 - val_mae: 0.3695 Epoch 185/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 186/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 187/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3513 - val_loss: 0.1849 - val_mae: 0.3705 Epoch 188/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3515 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 189/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 190/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 191/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3688 Epoch 192/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3699 Epoch 193/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 194/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3694 Epoch 195/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1849 - val_mae: 0.3708 Epoch 196/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3515 - val_loss: 0.1849 - val_mae: 0.3706 Epoch 197/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3514 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 198/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1845 - val_mae: 0.3690 Epoch 199/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1699 - mae: 0.3508 - val_loss: 0.1844 - val_mae: 0.3677 Epoch 200/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3684 Epoch 201/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3684 Epoch 202/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3688 Epoch 203/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3677 Epoch 204/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 205/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 206/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 207/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 208/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 209/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 210/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3509 - val_loss: 0.1845 - val_mae: 0.3684 Epoch 211/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3686 Epoch 212/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 213/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 214/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1844 - val_mae: 0.3673 Epoch 215/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1702 - mae: 0.3509 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 216/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3686 Epoch 217/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 218/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 219/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 220/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3509 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 221/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1695 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3686 Epoch 222/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 223/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3682 Epoch 224/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 225/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3514 - val_loss: 0.1852 - val_mae: 0.3715 Epoch 226/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3516 - val_loss: 0.1848 - val_mae: 0.3701 Epoch 227/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1847 - val_mae: 0.3699 Epoch 228/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3694 Epoch 229/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3505 - val_loss: 0.1844 - val_mae: 0.3671 Epoch 230/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3503 - val_loss: 0.1845 - val_mae: 0.3682 Epoch 231/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 232/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 233/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 234/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 235/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 236/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3694 Epoch 237/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 238/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 239/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3687 Epoch 240/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 241/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1849 - val_mae: 0.3703 Epoch 242/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 243/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 244/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3688 Epoch 245/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3507 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 246/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3514 - val_loss: 0.1849 - val_mae: 0.3706 Epoch 247/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1695 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3682 Epoch 248/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3500 - val_loss: 0.1844 - val_mae: 0.3671 Epoch 249/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3688 Epoch 250/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1844 - val_mae: 0.3680 Epoch 251/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1844 - val_mae: 0.3676 Epoch 252/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 253/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1844 - val_mae: 0.3675 Epoch 254/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 255/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3510 - val_loss: 0.1847 - val_mae: 0.3695 Epoch 256/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 257/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3681 Epoch 258/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1694 - mae: 0.3506 - val_loss: 0.1849 - val_mae: 0.3702 Epoch 259/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1699 - mae: 0.3514 - val_loss: 0.1850 - val_mae: 0.3707 Epoch 260/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 261/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 262/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3507 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 263/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3514 - val_loss: 0.1850 - val_mae: 0.3709 Epoch 264/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3514 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 265/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 266/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 267/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3502 - val_loss: 0.1844 - val_mae: 0.3669 Epoch 268/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3499 - val_loss: 0.1844 - val_mae: 0.3671 Epoch 269/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3693 Epoch 270/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 271/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 272/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 273/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 274/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 275/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3503 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 276/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3677 Epoch 277/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3500 - val_loss: 0.1845 - val_mae: 0.3672 Epoch 278/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3500 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 279/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3505 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 280/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 281/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 282/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 283/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 284/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1701 - mae: 0.3516 - val_loss: 0.1849 - val_mae: 0.3705 Epoch 285/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1698 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3674 Epoch 286/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3680 Epoch 287/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 288/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1704 - mae: 0.3520 - val_loss: 0.1856 - val_mae: 0.3726 Epoch 289/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3518 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 290/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1702 - mae: 0.3510 - val_loss: 0.1844 - val_mae: 0.3675 Epoch 291/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 292/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1846 - val_mae: 0.3694 Epoch 293/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 294/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 295/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3507 - val_loss: 0.1844 - val_mae: 0.3678 Epoch 296/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 297/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 298/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3513 - val_loss: 0.1848 - val_mae: 0.3704 Epoch 299/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1846 - val_mae: 0.3695 Epoch 300/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3505 - val_loss: 0.1844 - val_mae: 0.3675 Epoch 301/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3500 - val_loss: 0.1844 - val_mae: 0.3669 Epoch 302/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1703 - mae: 0.3499 - val_loss: 0.1846 - val_mae: 0.3655 Epoch 303/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 304/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3698 Epoch 305/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 306/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 307/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3673 Epoch 308/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1694 - mae: 0.3504 - val_loss: 0.1849 - val_mae: 0.3703 Epoch 309/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3518 - val_loss: 0.1853 - val_mae: 0.3718 Epoch 310/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3512 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 311/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 312/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3687 Epoch 313/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 314/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 315/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 316/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1851 - val_mae: 0.3712 Epoch 317/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1699 - mae: 0.3517 - val_loss: 0.1851 - val_mae: 0.3712 Epoch 318/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 319/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 320/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 321/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3506 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 322/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1700 - mae: 0.3518 - val_loss: 0.1854 - val_mae: 0.3720 Epoch 323/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1707 - mae: 0.3520 - val_loss: 0.1845 - val_mae: 0.3680 Epoch 324/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3684 Epoch 325/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1694 - mae: 0.3509 - val_loss: 0.1850 - val_mae: 0.3707 Epoch 326/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3514 - val_loss: 0.1849 - val_mae: 0.3706 Epoch 327/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3515 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 328/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 329/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3512 - val_loss: 0.1848 - val_mae: 0.3704 Epoch 330/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3514 - val_loss: 0.1848 - val_mae: 0.3704 Epoch 331/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1846 - val_mae: 0.3695 Epoch 332/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 333/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 334/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 335/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 336/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 337/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3503 - val_loss: 0.1846 - val_mae: 0.3685 Epoch 338/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 339/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 340/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3509 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 341/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 342/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1695 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 343/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 344/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3675 Epoch 345/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3681 Epoch 346/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3503 - val_loss: 0.1846 - val_mae: 0.3685 Epoch 347/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 348/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 349/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3501 - val_loss: 0.1846 - val_mae: 0.3683 Epoch 350/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1847 - val_mae: 0.3686 Epoch 351/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3678 Epoch 352/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3501 - val_loss: 0.1846 - val_mae: 0.3674 Epoch 353/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1695 - mae: 0.3503 - val_loss: 0.1848 - val_mae: 0.3692 Epoch 354/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3512 - val_loss: 0.1850 - val_mae: 0.3703 Epoch 355/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3683 Epoch 356/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1695 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3679 Epoch 357/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3500 - val_loss: 0.1846 - val_mae: 0.3678 Epoch 358/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3503 - val_loss: 0.1848 - val_mae: 0.3694 Epoch 359/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3516 - val_loss: 0.1855 - val_mae: 0.3718 Epoch 360/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3513 - val_loss: 0.1848 - val_mae: 0.3694 Epoch 361/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3503 - val_loss: 0.1846 - val_mae: 0.3681 Epoch 362/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3501 - val_loss: 0.1846 - val_mae: 0.3673 Epoch 363/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3503 - val_loss: 0.1848 - val_mae: 0.3693 Epoch 364/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3688 Epoch 365/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3691 Epoch 366/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3691 Epoch 367/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 368/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3691 Epoch 369/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1847 - val_mae: 0.3695 Epoch 370/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1695 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3684 Epoch 371/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 372/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1847 - val_mae: 0.3689 Epoch 373/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1694 - mae: 0.3507 - val_loss: 0.1850 - val_mae: 0.3707 Epoch 374/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3515 - val_loss: 0.1851 - val_mae: 0.3711 Epoch 375/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3513 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 376/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 377/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 378/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1700 - mae: 0.3516 - val_loss: 0.1850 - val_mae: 0.3709 Epoch 379/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3685 Epoch 380/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 381/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 382/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 383/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3513 - val_loss: 0.1848 - val_mae: 0.3701 Epoch 384/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3514 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 385/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3514 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 386/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3696 Epoch 387/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3513 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 388/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3510 - val_loss: 0.1846 - val_mae: 0.3684 Epoch 389/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 390/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 391/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1700 - mae: 0.3515 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 392/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3688 Epoch 393/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3510 - val_loss: 0.1847 - val_mae: 0.3695 Epoch 394/500 10/10 [==============================] - 0s 42ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 395/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 396/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1845 - val_mae: 0.3686 Epoch 397/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 398/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3503 - val_loss: 0.1845 - val_mae: 0.3678 Epoch 399/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 400/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3691 Epoch 401/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3688 Epoch 402/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1848 - val_mae: 0.3701 Epoch 403/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 404/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3508 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 405/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1695 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 406/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1700 - mae: 0.3514 - val_loss: 0.1851 - val_mae: 0.3710 Epoch 407/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3510 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 408/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3503 - val_loss: 0.1845 - val_mae: 0.3671 Epoch 409/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3689 Epoch 410/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3511 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 411/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1848 - val_mae: 0.3703 Epoch 412/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3512 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 413/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 414/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 415/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3512 - val_loss: 0.1849 - val_mae: 0.3704 Epoch 416/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3513 - val_loss: 0.1847 - val_mae: 0.3698 Epoch 417/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3681 Epoch 418/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3502 - val_loss: 0.1845 - val_mae: 0.3674 Epoch 419/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3503 - val_loss: 0.1847 - val_mae: 0.3695 Epoch 420/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3510 - val_loss: 0.1847 - val_mae: 0.3698 Epoch 421/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3513 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 422/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3681 Epoch 423/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3686 Epoch 424/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1845 - val_mae: 0.3684 Epoch 425/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3693 Epoch 426/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 427/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3680 Epoch 428/500 10/10 [==============================] - 0s 42ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 429/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3500 - val_loss: 0.1845 - val_mae: 0.3667 Epoch 430/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1703 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3670 Epoch 431/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1695 - mae: 0.3500 - val_loss: 0.1849 - val_mae: 0.3703 Epoch 432/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3519 - val_loss: 0.1851 - val_mae: 0.3712 Epoch 433/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3515 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 434/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3669 Epoch 435/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 436/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 437/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1847 - val_mae: 0.3697 Epoch 438/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1845 - val_mae: 0.3685 Epoch 439/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 440/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3511 - val_loss: 0.1848 - val_mae: 0.3699 Epoch 441/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1845 - val_mae: 0.3682 Epoch 442/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3500 - val_loss: 0.1845 - val_mae: 0.3666 Epoch 443/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3498 - val_loss: 0.1845 - val_mae: 0.3670 Epoch 444/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3501 - val_loss: 0.1845 - val_mae: 0.3682 Epoch 445/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1848 - val_mae: 0.3695 Epoch 446/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1848 - val_mae: 0.3700 Epoch 447/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3510 - val_loss: 0.1848 - val_mae: 0.3698 Epoch 448/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3508 - val_loss: 0.1846 - val_mae: 0.3687 Epoch 449/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3685 Epoch 450/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3694 Epoch 451/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1848 - val_mae: 0.3697 Epoch 452/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3511 - val_loss: 0.1848 - val_mae: 0.3699 Epoch 453/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3509 - val_loss: 0.1845 - val_mae: 0.3679 Epoch 454/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3692 Epoch 455/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1846 - val_mae: 0.3690 Epoch 456/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3506 - val_loss: 0.1846 - val_mae: 0.3688 Epoch 457/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3693 Epoch 458/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3505 - val_loss: 0.1845 - val_mae: 0.3683 Epoch 459/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3684 Epoch 460/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3688 Epoch 461/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3671 Epoch 462/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3500 - val_loss: 0.1846 - val_mae: 0.3682 Epoch 463/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1848 - val_mae: 0.3698 Epoch 464/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1847 - val_mae: 0.3691 Epoch 465/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1845 - val_mae: 0.3676 Epoch 466/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1698 - mae: 0.3500 - val_loss: 0.1845 - val_mae: 0.3670 Epoch 467/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3500 - val_loss: 0.1846 - val_mae: 0.3684 Epoch 468/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1852 - val_mae: 0.3714 Epoch 469/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3518 - val_loss: 0.1852 - val_mae: 0.3712 Epoch 470/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3511 - val_loss: 0.1846 - val_mae: 0.3686 Epoch 471/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1695 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3685 Epoch 472/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3683 Epoch 473/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3671 Epoch 474/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1698 - mae: 0.3503 - val_loss: 0.1848 - val_mae: 0.3693 Epoch 475/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1846 - val_mae: 0.3683 Epoch 476/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 477/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1701 - mae: 0.3504 - val_loss: 0.1846 - val_mae: 0.3674 Epoch 478/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1703 - mae: 0.3515 - val_loss: 0.1851 - val_mae: 0.3707 Epoch 479/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3687 Epoch 480/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3503 - val_loss: 0.1846 - val_mae: 0.3677 Epoch 481/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3499 - val_loss: 0.1846 - val_mae: 0.3677 Epoch 482/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3507 - val_loss: 0.1850 - val_mae: 0.3701 Epoch 483/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1699 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3692 Epoch 484/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1695 - mae: 0.3505 - val_loss: 0.1847 - val_mae: 0.3686 Epoch 485/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1697 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3675 Epoch 486/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1701 - mae: 0.3510 - val_loss: 0.1849 - val_mae: 0.3699 Epoch 487/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1697 - mae: 0.3511 - val_loss: 0.1849 - val_mae: 0.3699 Epoch 488/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1695 - mae: 0.3503 - val_loss: 0.1846 - val_mae: 0.3678 Epoch 489/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1699 - mae: 0.3505 - val_loss: 0.1847 - val_mae: 0.3686 Epoch 490/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1698 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3674 Epoch 491/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1702 - mae: 0.3509 - val_loss: 0.1847 - val_mae: 0.3689 Epoch 492/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1700 - mae: 0.3499 - val_loss: 0.1846 - val_mae: 0.3664 Epoch 493/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1697 - mae: 0.3498 - val_loss: 0.1847 - val_mae: 0.3685 Epoch 494/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1696 - mae: 0.3506 - val_loss: 0.1849 - val_mae: 0.3697 Epoch 495/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1848 - val_mae: 0.3695 Epoch 496/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3508 - val_loss: 0.1848 - val_mae: 0.3695 Epoch 497/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3507 - val_loss: 0.1847 - val_mae: 0.3687 Epoch 498/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3505 - val_loss: 0.1847 - val_mae: 0.3688 Epoch 499/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1696 - mae: 0.3504 - val_loss: 0.1847 - val_mae: 0.3685 Epoch 500/500 10/10 [==============================] - 0s 37ms/step - loss: 0.1696 - mae: 0.3502 - val_loss: 0.1846 - val_mae: 0.3680
In [10]:
gpus = tf.config.experimental.list_physical_devices('GPU')
if gpus:
try:
tf.config.experimental.set_virtual_device_configuration(
gpus[0],[tf.config.experimental.VirtualDeviceConfiguration(memory_limit=5120)])
except RuntimeError as e:
print(e)
Virtual devices cannot be modified after being initialized
3. Plot Metrics¶
Each training epoch, the model prints out its loss and mean absolute error for training and validation. You can read this in the output above (note that your exact numbers may differ):
Epoch 500/500 10/10 [==============================] - 0s 10ms/step - loss: 0.0121 - mae: 0.0882 - val_loss: 0.0115 - val_mae: 0.0865 You can see that we've already got a huge improvement - validation loss has dropped from 0.15 to 0.01, and validation MAE has dropped from 0.33 to 0.08.
The following cell will print the same graphs we used to evaluate our original model, but showing our new training history:
In [21]:
# Draw a graph of the loss, which is the distance between
# the predicted and actual values during training and validation.
train_loss = history_1.history['loss']
val_loss = history_1.history['val_loss']
epochs = range(1, len(train_loss) + 1)
plt.plot(epochs, train_loss, 'g.', label='Training loss')
plt.plot(epochs, val_loss, 'b', label='Validation loss')
plt.title('Training and validation loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.show()
The graph shows the loss (or the difference between the model's predictions and the actual data) for each epoch. There are several ways to calculate loss, and the method we have used is mean squared error. There is a distinct loss value given for the training and the validation data.
As we can see, the amount of loss rapidly decreases over the first 25 epochs, before flattening out. This means that the model is improving and producing more accurate predictions!
Our goal is to stop training when either the model is no longer improving, or when the training loss is less than the validation loss, which would mean that the model has learned to predict the training data so well that it can no longer generalize to new data.
To make the flatter part of the graph more readable, let's skip the first 50 epochs:
In [22]:
# Exclude the first few epochs so the graph is easier to read
SKIP = 50
plt.plot(epochs[SKIP:], train_loss[SKIP:], 'g.', label='Training loss')
plt.plot(epochs[SKIP:], val_loss[SKIP:], 'b.', label='Validation loss')
plt.title('Training and validation loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.show()
From the plot, we can see that loss continues to reduce until around 200 epochs, at which point it is mostly stable. This means that there's no need to train our network beyond 200 epochs.
However, we can also see that the lowest loss value is still around 0.155. This means that our network's predictions are off by an average of ~15%. In addition, the validation loss values jump around a lot, and is sometimes even higher.
2. Mean Absolute Error¶
To gain more insight into our model's performance we can plot some more data. This time, we'll plot the mean absolute error, which is another way of measuring how far the network's predictions are from the actual numbers:
In [23]:
plt.clf()
# Draw a graph of mean absolute error, which is another way of
# measuring the amount of error in the prediction.
train_mae = history_1.history['mae']
val_mae = history_1.history['val_mae']
plt.plot(epochs[SKIP:], train_mae[SKIP:], 'g.', label='Training MAE')
plt.plot(epochs[SKIP:], val_mae[SKIP:], 'b.', label='Validation MAE')
plt.title('Training and validation mean absolute error')
plt.xlabel('Epochs')
plt.ylabel('MAE')
plt.legend()
plt.show()
This graph of mean absolute error tells another story. We can see that training data shows consistently lower error than validation data, which means that the network may have overfit, or learned the training data so rigidly that it can't make effective predictions about new data.
In addition, the mean absolute error values are quite high, ~0.305 at best, which means some of the model's predictions are at least 30% off. A 30% error means we are very far from accurately modelling the sine wave function.
3. Actual vs Predicted Outputs¶
To get more insight into what is happening, let's check its predictions against the test dataset we set aside earlier:
In [24]:
# Calculate and print the loss on our test dataset
test_loss, test_mae = model_1.evaluate(x_test, y_test)
# Make predictions based on our test dataset
y_test_pred = model_1.predict(x_test)
# Graph the predictions against the actual values
plt.clf()
plt.title('Comparison of predictions and actual values')
plt.plot(x_test, y_test, 'b.', label='Actual values')
plt.plot(x_test, y_test_pred, 'r.', label='TF predictions')
plt.legend()
plt.show()
7/7 [==============================] - 0s 13ms/step - loss: 0.1627 - mae: 0.3434
Oh dear! The graph makes it clear that our network has learned to approximate the sine function in a very limited way.
The rigidity of this fit suggests that the model does not have enough capacity to learn the full complexity of the sine wave function, so it's only able to approximate it in an overly simplistic way. By making our model bigger, we should be able to improve its performance.
Training a Larger Model
1. Design the Model¶
To make our model bigger, let's add an additional layer of neurons. The following cell redefines our model in the same way as earlier, but with 16 neurons in the first layer and an additional layer of 16 neurons in the middle:¶
Generate a TensorFlow Lite Model¶
1. Generate Models with or without Quantization¶
We now have an acceptably accurate model. We'll use the TensorFlow Lite Converter to convert the model into a special, space-efficient format for use on memory-constrained devices.
Since this model is going to be deployed on a microcontroller, we want it to be as tiny as possible! One technique for reducing the size of a model is called quantization. It reduces the precision of the model's weights, and possibly the activations (output of each layer) as well, which saves memory, often without much impact on accuracy. Quantized models also run faster, since the calculations required are simpler.
In the following cell, we'll convert the model twice: once with quantization, once without.
In [25]:
model = tf.keras.Sequential()
# First layer takes a scalar input and feeds it through 16 "neurons". The
# neurons decide whether to activate based on the 'relu' activation function.
model.add(keras.layers.Dense(16, activation='relu', input_shape=(1,)))
# The new second and third layer will help the network learn more complex representations
model.add(keras.layers.Dense(16, activation='relu'))
# Final layer is a single neuron, since we want to output a single value
model.add(keras.layers.Dense(1))
# Compile the model using the standard 'adam' optimizer and the mean squared error or 'mse' loss function for regression.
model.compile(optimizer='adam', loss="mse", metrics=["mae"])
2. Train the Model¶
We'll now train and save the new model.
In [26]:
# Train the model
history = model.fit(x_train, y_train, epochs=500, batch_size=64,
validation_data=(x_validate, y_validate))
# Save the model to disk
model.save(MODEL_TF)
Epoch 1/500 10/10 [==============================] - 4s 145ms/step - loss: 0.4227 - mae: 0.5504 - val_loss: 0.4316 - val_mae: 0.5685 Epoch 2/500 10/10 [==============================] - 0s 41ms/step - loss: 0.4077 - mae: 0.5508 - val_loss: 0.4157 - val_mae: 0.5581 Epoch 3/500 10/10 [==============================] - 0s 42ms/step - loss: 0.3908 - mae: 0.5330 - val_loss: 0.3988 - val_mae: 0.5444 Epoch 4/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3760 - mae: 0.5198 - val_loss: 0.3834 - val_mae: 0.5350 Epoch 5/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3606 - mae: 0.5097 - val_loss: 0.3684 - val_mae: 0.5257 Epoch 6/500 10/10 [==============================] - 0s 38ms/step - loss: 0.3457 - mae: 0.5008 - val_loss: 0.3532 - val_mae: 0.5166 Epoch 7/500 10/10 [==============================] - 0s 39ms/step - loss: 0.3303 - mae: 0.4896 - val_loss: 0.3369 - val_mae: 0.5055 Epoch 8/500 10/10 [==============================] - 0s 40ms/step - loss: 0.3142 - mae: 0.4768 - val_loss: 0.3203 - val_mae: 0.4940 Epoch 9/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2978 - mae: 0.4653 - val_loss: 0.3040 - val_mae: 0.4833 Epoch 10/500 10/10 [==============================] - 0s 41ms/step - loss: 0.2811 - mae: 0.4525 - val_loss: 0.2867 - val_mae: 0.4699 Epoch 11/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2648 - mae: 0.4384 - val_loss: 0.2698 - val_mae: 0.4557 Epoch 12/500 10/10 [==============================] - 0s 39ms/step - loss: 0.2475 - mae: 0.4250 - val_loss: 0.2534 - val_mae: 0.4436 Epoch 13/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2322 - mae: 0.4128 - val_loss: 0.2381 - val_mae: 0.4319 Epoch 14/500 10/10 [==============================] - 0s 41ms/step - loss: 0.2157 - mae: 0.3977 - val_loss: 0.2220 - val_mae: 0.4151 Epoch 15/500 10/10 [==============================] - 0s 40ms/step - loss: 0.2009 - mae: 0.3834 - val_loss: 0.2095 - val_mae: 0.4041 Epoch 16/500 10/10 [==============================] - 0s 43ms/step - loss: 0.1886 - mae: 0.3714 - val_loss: 0.1968 - val_mae: 0.3902 Epoch 17/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1768 - mae: 0.3580 - val_loss: 0.1867 - val_mae: 0.3789 Epoch 18/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1679 - mae: 0.3460 - val_loss: 0.1781 - val_mae: 0.3677 Epoch 19/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1582 - mae: 0.3346 - val_loss: 0.1713 - val_mae: 0.3575 Epoch 20/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1525 - mae: 0.3256 - val_loss: 0.1653 - val_mae: 0.3469 Epoch 21/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1465 - mae: 0.3158 - val_loss: 0.1613 - val_mae: 0.3390 Epoch 22/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1420 - mae: 0.3086 - val_loss: 0.1577 - val_mae: 0.3306 Epoch 23/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1395 - mae: 0.3023 - val_loss: 0.1552 - val_mae: 0.3235 Epoch 24/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1381 - mae: 0.2988 - val_loss: 0.1533 - val_mae: 0.3184 Epoch 25/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1365 - mae: 0.2927 - val_loss: 0.1517 - val_mae: 0.3122 Epoch 26/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1327 - mae: 0.2894 - val_loss: 0.1517 - val_mae: 0.3128 Epoch 27/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1322 - mae: 0.2873 - val_loss: 0.1494 - val_mae: 0.3057 Epoch 28/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1306 - mae: 0.2841 - val_loss: 0.1491 - val_mae: 0.3050 Epoch 29/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1302 - mae: 0.2813 - val_loss: 0.1479 - val_mae: 0.3010 Epoch 30/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1292 - mae: 0.2798 - val_loss: 0.1472 - val_mae: 0.2991 Epoch 31/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1277 - mae: 0.2775 - val_loss: 0.1467 - val_mae: 0.2990 Epoch 32/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1285 - mae: 0.2771 - val_loss: 0.1460 - val_mae: 0.2976 Epoch 33/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1265 - mae: 0.2742 - val_loss: 0.1453 - val_mae: 0.2954 Epoch 34/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1261 - mae: 0.2739 - val_loss: 0.1443 - val_mae: 0.2921 Epoch 35/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1261 - mae: 0.2719 - val_loss: 0.1441 - val_mae: 0.2951 Epoch 36/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1248 - mae: 0.2710 - val_loss: 0.1427 - val_mae: 0.2905 Epoch 37/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1238 - mae: 0.2689 - val_loss: 0.1431 - val_mae: 0.2931 Epoch 38/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1232 - mae: 0.2678 - val_loss: 0.1412 - val_mae: 0.2870 Epoch 39/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1222 - mae: 0.2668 - val_loss: 0.1407 - val_mae: 0.2890 Epoch 40/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1218 - mae: 0.2658 - val_loss: 0.1398 - val_mae: 0.2867 Epoch 41/500 10/10 [==============================] - 0s 42ms/step - loss: 0.1204 - mae: 0.2634 - val_loss: 0.1390 - val_mae: 0.2835 Epoch 42/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1214 - mae: 0.2646 - val_loss: 0.1380 - val_mae: 0.2832 Epoch 43/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1196 - mae: 0.2633 - val_loss: 0.1373 - val_mae: 0.2850 Epoch 44/500 10/10 [==============================] - 0s 42ms/step - loss: 0.1182 - mae: 0.2613 - val_loss: 0.1362 - val_mae: 0.2814 Epoch 45/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1185 - mae: 0.2616 - val_loss: 0.1353 - val_mae: 0.2806 Epoch 46/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1228 - mae: 0.2606 - val_loss: 0.1350 - val_mae: 0.2815 Epoch 47/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1217 - mae: 0.2638 - val_loss: 0.1339 - val_mae: 0.2814 Epoch 48/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1168 - mae: 0.2576 - val_loss: 0.1325 - val_mae: 0.2775 Epoch 49/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1173 - mae: 0.2608 - val_loss: 0.1318 - val_mae: 0.2781 Epoch 50/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1128 - mae: 0.2544 - val_loss: 0.1312 - val_mae: 0.2714 Epoch 51/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1119 - mae: 0.2514 - val_loss: 0.1319 - val_mae: 0.2800 Epoch 52/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1143 - mae: 0.2553 - val_loss: 0.1289 - val_mae: 0.2725 Epoch 53/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1115 - mae: 0.2524 - val_loss: 0.1290 - val_mae: 0.2759 Epoch 54/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1102 - mae: 0.2495 - val_loss: 0.1270 - val_mae: 0.2695 Epoch 55/500 10/10 [==============================] - 0s 38ms/step - loss: 0.1093 - mae: 0.2480 - val_loss: 0.1269 - val_mae: 0.2727 Epoch 56/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1079 - mae: 0.2471 - val_loss: 0.1254 - val_mae: 0.2654 Epoch 57/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1087 - mae: 0.2446 - val_loss: 0.1245 - val_mae: 0.2668 Epoch 58/500 10/10 [==============================] - 0s 39ms/step - loss: 0.1072 - mae: 0.2438 - val_loss: 0.1235 - val_mae: 0.2644 Epoch 59/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1058 - mae: 0.2433 - val_loss: 0.1231 - val_mae: 0.2663 Epoch 60/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1056 - mae: 0.2437 - val_loss: 0.1217 - val_mae: 0.2619 Epoch 61/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1044 - mae: 0.2406 - val_loss: 0.1207 - val_mae: 0.2607 Epoch 62/500 10/10 [==============================] - 0s 46ms/step - loss: 0.1034 - mae: 0.2396 - val_loss: 0.1217 - val_mae: 0.2663 Epoch 63/500 10/10 [==============================] - 0s 46ms/step - loss: 0.1032 - mae: 0.2384 - val_loss: 0.1189 - val_mae: 0.2564 Epoch 64/500 10/10 [==============================] - 0s 41ms/step - loss: 0.1017 - mae: 0.2359 - val_loss: 0.1185 - val_mae: 0.2593 Epoch 65/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1016 - mae: 0.2378 - val_loss: 0.1175 - val_mae: 0.2598 Epoch 66/500 10/10 [==============================] - 0s 40ms/step - loss: 0.1005 - mae: 0.2349 - val_loss: 0.1161 - val_mae: 0.2539 Epoch 67/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0989 - mae: 0.2324 - val_loss: 0.1158 - val_mae: 0.2554 Epoch 68/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0986 - mae: 0.2327 - val_loss: 0.1144 - val_mae: 0.2510 Epoch 69/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0974 - mae: 0.2295 - val_loss: 0.1136 - val_mae: 0.2527 Epoch 70/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0967 - mae: 0.2299 - val_loss: 0.1126 - val_mae: 0.2516 Epoch 71/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0956 - mae: 0.2287 - val_loss: 0.1116 - val_mae: 0.2472 Epoch 72/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0953 - mae: 0.2271 - val_loss: 0.1107 - val_mae: 0.2469 Epoch 73/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0964 - mae: 0.2290 - val_loss: 0.1097 - val_mae: 0.2461 Epoch 74/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0940 - mae: 0.2242 - val_loss: 0.1090 - val_mae: 0.2464 Epoch 75/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0932 - mae: 0.2261 - val_loss: 0.1080 - val_mae: 0.2454 Epoch 76/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0935 - mae: 0.2234 - val_loss: 0.1070 - val_mae: 0.2415 Epoch 77/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0912 - mae: 0.2222 - val_loss: 0.1064 - val_mae: 0.2431 Epoch 78/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0900 - mae: 0.2199 - val_loss: 0.1056 - val_mae: 0.2369 Epoch 79/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0896 - mae: 0.2176 - val_loss: 0.1049 - val_mae: 0.2413 Epoch 80/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0886 - mae: 0.2183 - val_loss: 0.1036 - val_mae: 0.2366 Epoch 81/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0880 - mae: 0.2173 - val_loss: 0.1027 - val_mae: 0.2380 Epoch 82/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0879 - mae: 0.2187 - val_loss: 0.1017 - val_mae: 0.2358 Epoch 83/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0874 - mae: 0.2138 - val_loss: 0.1012 - val_mae: 0.2346 Epoch 84/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0870 - mae: 0.2160 - val_loss: 0.1001 - val_mae: 0.2333 Epoch 85/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0849 - mae: 0.2117 - val_loss: 0.0992 - val_mae: 0.2310 Epoch 86/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0841 - mae: 0.2100 - val_loss: 0.0990 - val_mae: 0.2330 Epoch 87/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0836 - mae: 0.2106 - val_loss: 0.0977 - val_mae: 0.2314 Epoch 88/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0830 - mae: 0.2088 - val_loss: 0.0969 - val_mae: 0.2288 Epoch 89/500 10/10 [==============================] - 0s 44ms/step - loss: 0.0819 - mae: 0.2077 - val_loss: 0.0961 - val_mae: 0.2278 Epoch 90/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0812 - mae: 0.2064 - val_loss: 0.0953 - val_mae: 0.2268 Epoch 91/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0810 - mae: 0.2083 - val_loss: 0.0948 - val_mae: 0.2287 Epoch 92/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0799 - mae: 0.2057 - val_loss: 0.0939 - val_mae: 0.2217 Epoch 93/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0791 - mae: 0.2024 - val_loss: 0.0933 - val_mae: 0.2254 Epoch 94/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0783 - mae: 0.2028 - val_loss: 0.0920 - val_mae: 0.2217 Epoch 95/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0781 - mae: 0.2025 - val_loss: 0.0912 - val_mae: 0.2205 Epoch 96/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0772 - mae: 0.2007 - val_loss: 0.0905 - val_mae: 0.2186 Epoch 97/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0763 - mae: 0.1999 - val_loss: 0.0898 - val_mae: 0.2202 Epoch 98/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0764 - mae: 0.1985 - val_loss: 0.0890 - val_mae: 0.2184 Epoch 99/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0753 - mae: 0.1973 - val_loss: 0.0885 - val_mae: 0.2133 Epoch 100/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0750 - mae: 0.1971 - val_loss: 0.0876 - val_mae: 0.2138 Epoch 101/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0739 - mae: 0.1941 - val_loss: 0.0868 - val_mae: 0.2148 Epoch 102/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0739 - mae: 0.1957 - val_loss: 0.0861 - val_mae: 0.2131 Epoch 103/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0733 - mae: 0.1970 - val_loss: 0.0855 - val_mae: 0.2151 Epoch 104/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0726 - mae: 0.1938 - val_loss: 0.0845 - val_mae: 0.2111 Epoch 105/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0720 - mae: 0.1903 - val_loss: 0.0845 - val_mae: 0.2106 Epoch 106/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0714 - mae: 0.1940 - val_loss: 0.0832 - val_mae: 0.2099 Epoch 107/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0708 - mae: 0.1893 - val_loss: 0.0825 - val_mae: 0.2089 Epoch 108/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0715 - mae: 0.1945 - val_loss: 0.0818 - val_mae: 0.2077 Epoch 109/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0695 - mae: 0.1882 - val_loss: 0.0811 - val_mae: 0.2073 Epoch 110/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0684 - mae: 0.1885 - val_loss: 0.0805 - val_mae: 0.2051 Epoch 111/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0677 - mae: 0.1864 - val_loss: 0.0799 - val_mae: 0.2030 Epoch 112/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0674 - mae: 0.1849 - val_loss: 0.0792 - val_mae: 0.2032 Epoch 113/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0668 - mae: 0.1854 - val_loss: 0.0788 - val_mae: 0.2022 Epoch 114/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0664 - mae: 0.1825 - val_loss: 0.0780 - val_mae: 0.2016 Epoch 115/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0657 - mae: 0.1846 - val_loss: 0.0774 - val_mae: 0.2016 Epoch 116/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0651 - mae: 0.1819 - val_loss: 0.0768 - val_mae: 0.1984 Epoch 117/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0664 - mae: 0.1860 - val_loss: 0.0761 - val_mae: 0.1991 Epoch 118/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0646 - mae: 0.1802 - val_loss: 0.0756 - val_mae: 0.1974 Epoch 119/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0638 - mae: 0.1803 - val_loss: 0.0750 - val_mae: 0.1985 Epoch 120/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0638 - mae: 0.1787 - val_loss: 0.0742 - val_mae: 0.1972 Epoch 121/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0628 - mae: 0.1809 - val_loss: 0.0737 - val_mae: 0.1968 Epoch 122/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0632 - mae: 0.1769 - val_loss: 0.0733 - val_mae: 0.1954 Epoch 123/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0632 - mae: 0.1828 - val_loss: 0.0726 - val_mae: 0.1934 Epoch 124/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0633 - mae: 0.1760 - val_loss: 0.0726 - val_mae: 0.1966 Epoch 125/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0621 - mae: 0.1820 - val_loss: 0.0715 - val_mae: 0.1949 Epoch 126/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0602 - mae: 0.1755 - val_loss: 0.0709 - val_mae: 0.1939 Epoch 127/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0596 - mae: 0.1756 - val_loss: 0.0703 - val_mae: 0.1914 Epoch 128/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0596 - mae: 0.1723 - val_loss: 0.0698 - val_mae: 0.1900 Epoch 129/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0588 - mae: 0.1732 - val_loss: 0.0692 - val_mae: 0.1897 Epoch 130/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0584 - mae: 0.1712 - val_loss: 0.0688 - val_mae: 0.1912 Epoch 131/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0582 - mae: 0.1735 - val_loss: 0.0681 - val_mae: 0.1878 Epoch 132/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0592 - mae: 0.1704 - val_loss: 0.0686 - val_mae: 0.1927 Epoch 133/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0595 - mae: 0.1781 - val_loss: 0.0675 - val_mae: 0.1874 Epoch 134/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0589 - mae: 0.1725 - val_loss: 0.0670 - val_mae: 0.1898 Epoch 135/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0573 - mae: 0.1738 - val_loss: 0.0663 - val_mae: 0.1831 Epoch 136/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0572 - mae: 0.1683 - val_loss: 0.0661 - val_mae: 0.1886 Epoch 137/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0554 - mae: 0.1695 - val_loss: 0.0650 - val_mae: 0.1849 Epoch 138/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0547 - mae: 0.1665 - val_loss: 0.0645 - val_mae: 0.1840 Epoch 139/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0548 - mae: 0.1679 - val_loss: 0.0640 - val_mae: 0.1837 Epoch 140/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0544 - mae: 0.1650 - val_loss: 0.0639 - val_mae: 0.1853 Epoch 141/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0542 - mae: 0.1671 - val_loss: 0.0629 - val_mae: 0.1823 Epoch 142/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0531 - mae: 0.1653 - val_loss: 0.0625 - val_mae: 0.1831 Epoch 143/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0533 - mae: 0.1630 - val_loss: 0.0624 - val_mae: 0.1822 Epoch 144/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0524 - mae: 0.1636 - val_loss: 0.0615 - val_mae: 0.1790 Epoch 145/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0516 - mae: 0.1623 - val_loss: 0.0610 - val_mae: 0.1809 Epoch 146/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0513 - mae: 0.1616 - val_loss: 0.0605 - val_mae: 0.1785 Epoch 147/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0508 - mae: 0.1600 - val_loss: 0.0601 - val_mae: 0.1770 Epoch 148/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0504 - mae: 0.1589 - val_loss: 0.0597 - val_mae: 0.1778 Epoch 149/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0504 - mae: 0.1613 - val_loss: 0.0591 - val_mae: 0.1779 Epoch 150/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0499 - mae: 0.1585 - val_loss: 0.0589 - val_mae: 0.1764 Epoch 151/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0499 - mae: 0.1600 - val_loss: 0.0584 - val_mae: 0.1740 Epoch 152/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0495 - mae: 0.1566 - val_loss: 0.0579 - val_mae: 0.1758 Epoch 153/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0486 - mae: 0.1574 - val_loss: 0.0576 - val_mae: 0.1712 Epoch 154/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0483 - mae: 0.1556 - val_loss: 0.0569 - val_mae: 0.1755 Epoch 155/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0481 - mae: 0.1566 - val_loss: 0.0563 - val_mae: 0.1721 Epoch 156/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0482 - mae: 0.1575 - val_loss: 0.0558 - val_mae: 0.1719 Epoch 157/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0471 - mae: 0.1535 - val_loss: 0.0555 - val_mae: 0.1709 Epoch 158/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0471 - mae: 0.1555 - val_loss: 0.0550 - val_mae: 0.1700 Epoch 159/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0463 - mae: 0.1525 - val_loss: 0.0546 - val_mae: 0.1706 Epoch 160/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0460 - mae: 0.1534 - val_loss: 0.0540 - val_mae: 0.1685 Epoch 161/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0459 - mae: 0.1514 - val_loss: 0.0539 - val_mae: 0.1718 Epoch 162/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0453 - mae: 0.1519 - val_loss: 0.0534 - val_mae: 0.1663 Epoch 163/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0455 - mae: 0.1506 - val_loss: 0.0540 - val_mae: 0.1734 Epoch 164/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0454 - mae: 0.1556 - val_loss: 0.0537 - val_mae: 0.1695 Epoch 165/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0466 - mae: 0.1548 - val_loss: 0.0531 - val_mae: 0.1711 Epoch 166/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0446 - mae: 0.1517 - val_loss: 0.0516 - val_mae: 0.1642 Epoch 167/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0435 - mae: 0.1490 - val_loss: 0.0511 - val_mae: 0.1663 Epoch 168/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0436 - mae: 0.1497 - val_loss: 0.0510 - val_mae: 0.1634 Epoch 169/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0431 - mae: 0.1468 - val_loss: 0.0505 - val_mae: 0.1663 Epoch 170/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0427 - mae: 0.1475 - val_loss: 0.0501 - val_mae: 0.1610 Epoch 171/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0421 - mae: 0.1466 - val_loss: 0.0495 - val_mae: 0.1645 Epoch 172/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0419 - mae: 0.1472 - val_loss: 0.0491 - val_mae: 0.1625 Epoch 173/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0416 - mae: 0.1454 - val_loss: 0.0489 - val_mae: 0.1633 Epoch 174/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0413 - mae: 0.1454 - val_loss: 0.0486 - val_mae: 0.1590 Epoch 175/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0409 - mae: 0.1434 - val_loss: 0.0479 - val_mae: 0.1604 Epoch 176/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0405 - mae: 0.1433 - val_loss: 0.0473 - val_mae: 0.1598 Epoch 177/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0404 - mae: 0.1433 - val_loss: 0.0472 - val_mae: 0.1618 Epoch 178/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0397 - mae: 0.1437 - val_loss: 0.0466 - val_mae: 0.1581 Epoch 179/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0396 - mae: 0.1419 - val_loss: 0.0462 - val_mae: 0.1580 Epoch 180/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0394 - mae: 0.1433 - val_loss: 0.0461 - val_mae: 0.1567 Epoch 181/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0389 - mae: 0.1404 - val_loss: 0.0457 - val_mae: 0.1591 Epoch 182/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0390 - mae: 0.1426 - val_loss: 0.0451 - val_mae: 0.1558 Epoch 183/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0383 - mae: 0.1402 - val_loss: 0.0448 - val_mae: 0.1564 Epoch 184/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0379 - mae: 0.1391 - val_loss: 0.0442 - val_mae: 0.1550 Epoch 185/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0375 - mae: 0.1392 - val_loss: 0.0437 - val_mae: 0.1540 Epoch 186/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0372 - mae: 0.1391 - val_loss: 0.0434 - val_mae: 0.1540 Epoch 187/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0371 - mae: 0.1382 - val_loss: 0.0431 - val_mae: 0.1537 Epoch 188/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0369 - mae: 0.1389 - val_loss: 0.0427 - val_mae: 0.1513 Epoch 189/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0364 - mae: 0.1371 - val_loss: 0.0422 - val_mae: 0.1512 Epoch 190/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0365 - mae: 0.1366 - val_loss: 0.0425 - val_mae: 0.1542 Epoch 191/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0371 - mae: 0.1394 - val_loss: 0.0418 - val_mae: 0.1500 Epoch 192/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0355 - mae: 0.1357 - val_loss: 0.0412 - val_mae: 0.1517 Epoch 193/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0356 - mae: 0.1351 - val_loss: 0.0408 - val_mae: 0.1506 Epoch 194/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0355 - mae: 0.1361 - val_loss: 0.0404 - val_mae: 0.1481 Epoch 195/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0352 - mae: 0.1364 - val_loss: 0.0401 - val_mae: 0.1488 Epoch 196/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0344 - mae: 0.1338 - val_loss: 0.0403 - val_mae: 0.1512 Epoch 197/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0343 - mae: 0.1349 - val_loss: 0.0396 - val_mae: 0.1463 Epoch 198/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0339 - mae: 0.1333 - val_loss: 0.0393 - val_mae: 0.1485 Epoch 199/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0345 - mae: 0.1328 - val_loss: 0.0387 - val_mae: 0.1445 Epoch 200/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0337 - mae: 0.1322 - val_loss: 0.0383 - val_mae: 0.1445 Epoch 201/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0337 - mae: 0.1309 - val_loss: 0.0394 - val_mae: 0.1510 Epoch 202/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0328 - mae: 0.1327 - val_loss: 0.0388 - val_mae: 0.1460 Epoch 203/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0328 - mae: 0.1311 - val_loss: 0.0387 - val_mae: 0.1488 Epoch 204/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0336 - mae: 0.1351 - val_loss: 0.0372 - val_mae: 0.1438 Epoch 205/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0323 - mae: 0.1311 - val_loss: 0.0368 - val_mae: 0.1415 Epoch 206/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0321 - mae: 0.1290 - val_loss: 0.0365 - val_mae: 0.1424 Epoch 207/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0315 - mae: 0.1280 - val_loss: 0.0359 - val_mae: 0.1406 Epoch 208/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0312 - mae: 0.1278 - val_loss: 0.0356 - val_mae: 0.1413 Epoch 209/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0312 - mae: 0.1284 - val_loss: 0.0360 - val_mae: 0.1399 Epoch 210/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0308 - mae: 0.1274 - val_loss: 0.0355 - val_mae: 0.1430 Epoch 211/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0320 - mae: 0.1292 - val_loss: 0.0347 - val_mae: 0.1395 Epoch 212/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0308 - mae: 0.1292 - val_loss: 0.0348 - val_mae: 0.1388 Epoch 213/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0304 - mae: 0.1267 - val_loss: 0.0341 - val_mae: 0.1382 Epoch 214/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0302 - mae: 0.1252 - val_loss: 0.0340 - val_mae: 0.1377 Epoch 215/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0305 - mae: 0.1274 - val_loss: 0.0343 - val_mae: 0.1394 Epoch 216/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0297 - mae: 0.1264 - val_loss: 0.0330 - val_mae: 0.1364 Epoch 217/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0292 - mae: 0.1235 - val_loss: 0.0328 - val_mae: 0.1348 Epoch 218/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0288 - mae: 0.1239 - val_loss: 0.0325 - val_mae: 0.1366 Epoch 219/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0288 - mae: 0.1246 - val_loss: 0.0327 - val_mae: 0.1340 Epoch 220/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0290 - mae: 0.1247 - val_loss: 0.0318 - val_mae: 0.1335 Epoch 221/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0287 - mae: 0.1237 - val_loss: 0.0317 - val_mae: 0.1356 Epoch 222/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0279 - mae: 0.1229 - val_loss: 0.0326 - val_mae: 0.1347 Epoch 223/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0289 - mae: 0.1237 - val_loss: 0.0319 - val_mae: 0.1377 Epoch 224/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0280 - mae: 0.1218 - val_loss: 0.0306 - val_mae: 0.1318 Epoch 225/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0280 - mae: 0.1235 - val_loss: 0.0313 - val_mae: 0.1332 Epoch 226/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0277 - mae: 0.1221 - val_loss: 0.0309 - val_mae: 0.1352 Epoch 227/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0270 - mae: 0.1214 - val_loss: 0.0319 - val_mae: 0.1341 Epoch 228/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0285 - mae: 0.1226 - val_loss: 0.0294 - val_mae: 0.1291 Epoch 229/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0281 - mae: 0.1212 - val_loss: 0.0316 - val_mae: 0.1385 Epoch 230/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0271 - mae: 0.1230 - val_loss: 0.0293 - val_mae: 0.1299 Epoch 231/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0259 - mae: 0.1184 - val_loss: 0.0286 - val_mae: 0.1284 Epoch 232/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0257 - mae: 0.1177 - val_loss: 0.0290 - val_mae: 0.1269 Epoch 233/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0255 - mae: 0.1169 - val_loss: 0.0281 - val_mae: 0.1286 Epoch 234/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0255 - mae: 0.1178 - val_loss: 0.0276 - val_mae: 0.1267 Epoch 235/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0252 - mae: 0.1167 - val_loss: 0.0282 - val_mae: 0.1251 Epoch 236/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0254 - mae: 0.1168 - val_loss: 0.0272 - val_mae: 0.1249 Epoch 237/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0253 - mae: 0.1156 - val_loss: 0.0273 - val_mae: 0.1277 Epoch 238/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0245 - mae: 0.1155 - val_loss: 0.0269 - val_mae: 0.1241 Epoch 239/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0245 - mae: 0.1151 - val_loss: 0.0263 - val_mae: 0.1238 Epoch 240/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0241 - mae: 0.1140 - val_loss: 0.0260 - val_mae: 0.1243 Epoch 241/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0241 - mae: 0.1153 - val_loss: 0.0263 - val_mae: 0.1230 Epoch 242/500 10/10 [==============================] - 0s 45ms/step - loss: 0.0237 - mae: 0.1130 - val_loss: 0.0256 - val_mae: 0.1229 Epoch 243/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0234 - mae: 0.1119 - val_loss: 0.0253 - val_mae: 0.1232 Epoch 244/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0233 - mae: 0.1129 - val_loss: 0.0249 - val_mae: 0.1222 Epoch 245/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0233 - mae: 0.1122 - val_loss: 0.0246 - val_mae: 0.1213 Epoch 246/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0226 - mae: 0.1105 - val_loss: 0.0244 - val_mae: 0.1209 Epoch 247/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0225 - mae: 0.1101 - val_loss: 0.0244 - val_mae: 0.1203 Epoch 248/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0223 - mae: 0.1100 - val_loss: 0.0240 - val_mae: 0.1203 Epoch 249/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0223 - mae: 0.1107 - val_loss: 0.0237 - val_mae: 0.1200 Epoch 250/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0220 - mae: 0.1098 - val_loss: 0.0242 - val_mae: 0.1183 Epoch 251/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0222 - mae: 0.1100 - val_loss: 0.0234 - val_mae: 0.1194 Epoch 252/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0217 - mae: 0.1096 - val_loss: 0.0237 - val_mae: 0.1200 Epoch 253/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0217 - mae: 0.1095 - val_loss: 0.0235 - val_mae: 0.1189 Epoch 254/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0213 - mae: 0.1082 - val_loss: 0.0230 - val_mae: 0.1179 Epoch 255/500 10/10 [==============================] - 0s 37ms/step - loss: 0.0219 - mae: 0.1090 - val_loss: 0.0226 - val_mae: 0.1174 Epoch 256/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0215 - mae: 0.1083 - val_loss: 0.0235 - val_mae: 0.1206 Epoch 257/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0219 - mae: 0.1096 - val_loss: 0.0220 - val_mae: 0.1165 Epoch 258/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0210 - mae: 0.1084 - val_loss: 0.0218 - val_mae: 0.1165 Epoch 259/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0207 - mae: 0.1075 - val_loss: 0.0219 - val_mae: 0.1153 Epoch 260/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0204 - mae: 0.1059 - val_loss: 0.0222 - val_mae: 0.1177 Epoch 261/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0211 - mae: 0.1091 - val_loss: 0.0215 - val_mae: 0.1159 Epoch 262/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0208 - mae: 0.1087 - val_loss: 0.0227 - val_mae: 0.1146 Epoch 263/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0201 - mae: 0.1060 - val_loss: 0.0208 - val_mae: 0.1143 Epoch 264/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0199 - mae: 0.1054 - val_loss: 0.0207 - val_mae: 0.1136 Epoch 265/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0196 - mae: 0.1049 - val_loss: 0.0209 - val_mae: 0.1131 Epoch 266/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0198 - mae: 0.1057 - val_loss: 0.0208 - val_mae: 0.1127 Epoch 267/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0199 - mae: 0.1047 - val_loss: 0.0216 - val_mae: 0.1166 Epoch 268/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0195 - mae: 0.1051 - val_loss: 0.0205 - val_mae: 0.1120 Epoch 269/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0193 - mae: 0.1049 - val_loss: 0.0199 - val_mae: 0.1117 Epoch 270/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0191 - mae: 0.1040 - val_loss: 0.0197 - val_mae: 0.1115 Epoch 271/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0188 - mae: 0.1032 - val_loss: 0.0196 - val_mae: 0.1108 Epoch 272/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0188 - mae: 0.1034 - val_loss: 0.0194 - val_mae: 0.1107 Epoch 273/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0186 - mae: 0.1028 - val_loss: 0.0194 - val_mae: 0.1107 Epoch 274/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0185 - mae: 0.1032 - val_loss: 0.0196 - val_mae: 0.1106 Epoch 275/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0192 - mae: 0.1058 - val_loss: 0.0189 - val_mae: 0.1098 Epoch 276/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0196 - mae: 0.1058 - val_loss: 0.0204 - val_mae: 0.1140 Epoch 277/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0193 - mae: 0.1072 - val_loss: 0.0199 - val_mae: 0.1127 Epoch 278/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0199 - mae: 0.1074 - val_loss: 0.0191 - val_mae: 0.1091 Epoch 279/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0181 - mae: 0.1027 - val_loss: 0.0184 - val_mae: 0.1082 Epoch 280/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0180 - mae: 0.1022 - val_loss: 0.0183 - val_mae: 0.1077 Epoch 281/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0178 - mae: 0.1016 - val_loss: 0.0182 - val_mae: 0.1081 Epoch 282/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0181 - mae: 0.1026 - val_loss: 0.0183 - val_mae: 0.1071 Epoch 283/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0177 - mae: 0.1016 - val_loss: 0.0178 - val_mae: 0.1069 Epoch 284/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0182 - mae: 0.1037 - val_loss: 0.0192 - val_mae: 0.1090 Epoch 285/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0177 - mae: 0.1026 - val_loss: 0.0177 - val_mae: 0.1060 Epoch 286/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0177 - mae: 0.1021 - val_loss: 0.0174 - val_mae: 0.1060 Epoch 287/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0173 - mae: 0.1009 - val_loss: 0.0174 - val_mae: 0.1055 Epoch 288/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0173 - mae: 0.1011 - val_loss: 0.0181 - val_mae: 0.1067 Epoch 289/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0176 - mae: 0.1027 - val_loss: 0.0177 - val_mae: 0.1049 Epoch 290/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0167 - mae: 0.0999 - val_loss: 0.0176 - val_mae: 0.1068 Epoch 291/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0174 - mae: 0.1014 - val_loss: 0.0168 - val_mae: 0.1044 Epoch 292/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0169 - mae: 0.1007 - val_loss: 0.0171 - val_mae: 0.1045 Epoch 293/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0164 - mae: 0.0993 - val_loss: 0.0166 - val_mae: 0.1037 Epoch 294/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0165 - mae: 0.0995 - val_loss: 0.0164 - val_mae: 0.1028 Epoch 295/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0162 - mae: 0.0986 - val_loss: 0.0163 - val_mae: 0.1025 Epoch 296/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0166 - mae: 0.0996 - val_loss: 0.0164 - val_mae: 0.1027 Epoch 297/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0162 - mae: 0.0991 - val_loss: 0.0162 - val_mae: 0.1020 Epoch 298/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0166 - mae: 0.0993 - val_loss: 0.0161 - val_mae: 0.1023 Epoch 299/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0165 - mae: 0.0998 - val_loss: 0.0166 - val_mae: 0.1029 Epoch 300/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0171 - mae: 0.1004 - val_loss: 0.0156 - val_mae: 0.1006 Epoch 301/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0160 - mae: 0.0987 - val_loss: 0.0157 - val_mae: 0.1006 Epoch 302/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0157 - mae: 0.0973 - val_loss: 0.0163 - val_mae: 0.1030 Epoch 303/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0159 - mae: 0.0991 - val_loss: 0.0171 - val_mae: 0.1027 Epoch 304/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0165 - mae: 0.1004 - val_loss: 0.0161 - val_mae: 0.1009 Epoch 305/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0163 - mae: 0.0990 - val_loss: 0.0157 - val_mae: 0.1005 Epoch 306/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0163 - mae: 0.0997 - val_loss: 0.0167 - val_mae: 0.1038 Epoch 307/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0160 - mae: 0.0982 - val_loss: 0.0164 - val_mae: 0.1027 Epoch 308/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0177 - mae: 0.1039 - val_loss: 0.0148 - val_mae: 0.0985 Epoch 309/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0154 - mae: 0.0989 - val_loss: 0.0187 - val_mae: 0.1060 Epoch 310/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0163 - mae: 0.1004 - val_loss: 0.0152 - val_mae: 0.0998 Epoch 311/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0164 - mae: 0.1007 - val_loss: 0.0180 - val_mae: 0.1054 Epoch 312/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0162 - mae: 0.1005 - val_loss: 0.0150 - val_mae: 0.0983 Epoch 313/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0152 - mae: 0.0970 - val_loss: 0.0157 - val_mae: 0.0995 Epoch 314/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0150 - mae: 0.0961 - val_loss: 0.0148 - val_mae: 0.0984 Epoch 315/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0152 - mae: 0.0966 - val_loss: 0.0151 - val_mae: 0.0992 Epoch 316/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0156 - mae: 0.0986 - val_loss: 0.0146 - val_mae: 0.0979 Epoch 317/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0150 - mae: 0.0967 - val_loss: 0.0152 - val_mae: 0.0973 Epoch 318/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0148 - mae: 0.0956 - val_loss: 0.0143 - val_mae: 0.0964 Epoch 319/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0147 - mae: 0.0953 - val_loss: 0.0142 - val_mae: 0.0966 Epoch 320/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0150 - mae: 0.0962 - val_loss: 0.0144 - val_mae: 0.0969 Epoch 321/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0149 - mae: 0.0967 - val_loss: 0.0143 - val_mae: 0.0972 Epoch 322/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0147 - mae: 0.0962 - val_loss: 0.0149 - val_mae: 0.0968 Epoch 323/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0151 - mae: 0.0964 - val_loss: 0.0147 - val_mae: 0.0976 Epoch 324/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0149 - mae: 0.0958 - val_loss: 0.0152 - val_mae: 0.0992 Epoch 325/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0147 - mae: 0.0955 - val_loss: 0.0138 - val_mae: 0.0948 Epoch 326/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0141 - mae: 0.0937 - val_loss: 0.0137 - val_mae: 0.0945 Epoch 327/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0140 - mae: 0.0932 - val_loss: 0.0143 - val_mae: 0.0964 Epoch 328/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0144 - mae: 0.0953 - val_loss: 0.0135 - val_mae: 0.0938 Epoch 329/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0146 - mae: 0.0953 - val_loss: 0.0146 - val_mae: 0.0961 Epoch 330/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0149 - mae: 0.0964 - val_loss: 0.0163 - val_mae: 0.1019 Epoch 331/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0148 - mae: 0.0962 - val_loss: 0.0135 - val_mae: 0.0939 Epoch 332/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0138 - mae: 0.0927 - val_loss: 0.0134 - val_mae: 0.0934 Epoch 333/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0137 - mae: 0.0924 - val_loss: 0.0142 - val_mae: 0.0957 Epoch 334/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0138 - mae: 0.0930 - val_loss: 0.0132 - val_mae: 0.0930 Epoch 335/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0137 - mae: 0.0929 - val_loss: 0.0145 - val_mae: 0.0959 Epoch 336/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0138 - mae: 0.0925 - val_loss: 0.0136 - val_mae: 0.0938 Epoch 337/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0135 - mae: 0.0923 - val_loss: 0.0130 - val_mae: 0.0922 Epoch 338/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0137 - mae: 0.0927 - val_loss: 0.0130 - val_mae: 0.0919 Epoch 339/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0138 - mae: 0.0929 - val_loss: 0.0131 - val_mae: 0.0919 Epoch 340/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0136 - mae: 0.0923 - val_loss: 0.0134 - val_mae: 0.0928 Epoch 341/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0136 - mae: 0.0932 - val_loss: 0.0137 - val_mae: 0.0933 Epoch 342/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0137 - mae: 0.0929 - val_loss: 0.0131 - val_mae: 0.0921 Epoch 343/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0142 - mae: 0.0941 - val_loss: 0.0144 - val_mae: 0.0957 Epoch 344/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0142 - mae: 0.0945 - val_loss: 0.0136 - val_mae: 0.0931 Epoch 345/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0134 - mae: 0.0918 - val_loss: 0.0127 - val_mae: 0.0906 Epoch 346/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0133 - mae: 0.0921 - val_loss: 0.0141 - val_mae: 0.0950 Epoch 347/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0140 - mae: 0.0937 - val_loss: 0.0130 - val_mae: 0.0916 Epoch 348/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0147 - mae: 0.0960 - val_loss: 0.0134 - val_mae: 0.0923 Epoch 349/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0138 - mae: 0.0919 - val_loss: 0.0140 - val_mae: 0.0938 Epoch 350/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0132 - mae: 0.0918 - val_loss: 0.0129 - val_mae: 0.0915 Epoch 351/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0133 - mae: 0.0915 - val_loss: 0.0132 - val_mae: 0.0921 Epoch 352/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0132 - mae: 0.0908 - val_loss: 0.0128 - val_mae: 0.0909 Epoch 353/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0128 - mae: 0.0905 - val_loss: 0.0135 - val_mae: 0.0928 Epoch 354/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0135 - mae: 0.0928 - val_loss: 0.0126 - val_mae: 0.0903 Epoch 355/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0130 - mae: 0.0909 - val_loss: 0.0136 - val_mae: 0.0936 Epoch 356/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0135 - mae: 0.0921 - val_loss: 0.0138 - val_mae: 0.0946 Epoch 357/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0136 - mae: 0.0940 - val_loss: 0.0127 - val_mae: 0.0904 Epoch 358/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0128 - mae: 0.0900 - val_loss: 0.0125 - val_mae: 0.0894 Epoch 359/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0128 - mae: 0.0901 - val_loss: 0.0125 - val_mae: 0.0900 Epoch 360/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0130 - mae: 0.0915 - val_loss: 0.0135 - val_mae: 0.0936 Epoch 361/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0131 - mae: 0.0914 - val_loss: 0.0124 - val_mae: 0.0894 Epoch 362/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0133 - mae: 0.0918 - val_loss: 0.0125 - val_mae: 0.0894 Epoch 363/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0126 - mae: 0.0895 - val_loss: 0.0127 - val_mae: 0.0907 Epoch 364/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0129 - mae: 0.0908 - val_loss: 0.0124 - val_mae: 0.0896 Epoch 365/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0128 - mae: 0.0900 - val_loss: 0.0123 - val_mae: 0.0893 Epoch 366/500 10/10 [==============================] - 0s 45ms/step - loss: 0.0127 - mae: 0.0900 - val_loss: 0.0123 - val_mae: 0.0895 Epoch 367/500 10/10 [==============================] - 0s 46ms/step - loss: 0.0127 - mae: 0.0892 - val_loss: 0.0121 - val_mae: 0.0882 Epoch 368/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0127 - mae: 0.0900 - val_loss: 0.0126 - val_mae: 0.0903 Epoch 369/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0129 - mae: 0.0910 - val_loss: 0.0122 - val_mae: 0.0887 Epoch 370/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0127 - mae: 0.0897 - val_loss: 0.0120 - val_mae: 0.0882 Epoch 371/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0131 - mae: 0.0906 - val_loss: 0.0128 - val_mae: 0.0904 Epoch 372/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0138 - mae: 0.0932 - val_loss: 0.0131 - val_mae: 0.0908 Epoch 373/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0136 - mae: 0.0930 - val_loss: 0.0125 - val_mae: 0.0890 Epoch 374/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0130 - mae: 0.0912 - val_loss: 0.0125 - val_mae: 0.0900 Epoch 375/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0124 - mae: 0.0890 - val_loss: 0.0121 - val_mae: 0.0879 Epoch 376/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0893 - val_loss: 0.0121 - val_mae: 0.0878 Epoch 377/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0897 - val_loss: 0.0125 - val_mae: 0.0890 Epoch 378/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0127 - mae: 0.0896 - val_loss: 0.0122 - val_mae: 0.0881 Epoch 379/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0124 - mae: 0.0888 - val_loss: 0.0127 - val_mae: 0.0909 Epoch 380/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0127 - mae: 0.0903 - val_loss: 0.0119 - val_mae: 0.0878 Epoch 381/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0126 - mae: 0.0894 - val_loss: 0.0123 - val_mae: 0.0894 Epoch 382/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0895 - val_loss: 0.0119 - val_mae: 0.0872 Epoch 383/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0125 - mae: 0.0891 - val_loss: 0.0119 - val_mae: 0.0872 Epoch 384/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0904 - val_loss: 0.0120 - val_mae: 0.0878 Epoch 385/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0890 - val_loss: 0.0121 - val_mae: 0.0886 Epoch 386/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0886 - val_loss: 0.0120 - val_mae: 0.0880 Epoch 387/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0885 - val_loss: 0.0118 - val_mae: 0.0873 Epoch 388/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0887 - val_loss: 0.0133 - val_mae: 0.0914 Epoch 389/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0132 - mae: 0.0921 - val_loss: 0.0126 - val_mae: 0.0894 Epoch 390/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0131 - mae: 0.0912 - val_loss: 0.0131 - val_mae: 0.0909 Epoch 391/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0130 - mae: 0.0912 - val_loss: 0.0126 - val_mae: 0.0892 Epoch 392/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0127 - mae: 0.0904 - val_loss: 0.0121 - val_mae: 0.0878 Epoch 393/500 10/10 [==============================] - 0s 44ms/step - loss: 0.0133 - mae: 0.0915 - val_loss: 0.0121 - val_mae: 0.0880 Epoch 394/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0895 - val_loss: 0.0117 - val_mae: 0.0871 Epoch 395/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0889 - val_loss: 0.0120 - val_mae: 0.0881 Epoch 396/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0125 - mae: 0.0886 - val_loss: 0.0116 - val_mae: 0.0866 Epoch 397/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0124 - mae: 0.0891 - val_loss: 0.0121 - val_mae: 0.0883 Epoch 398/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0885 - val_loss: 0.0120 - val_mae: 0.0883 Epoch 399/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0899 - val_loss: 0.0122 - val_mae: 0.0893 Epoch 400/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0124 - mae: 0.0894 - val_loss: 0.0120 - val_mae: 0.0879 Epoch 401/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0124 - mae: 0.0887 - val_loss: 0.0118 - val_mae: 0.0869 Epoch 402/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0889 - val_loss: 0.0119 - val_mae: 0.0878 Epoch 403/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0876 - val_loss: 0.0125 - val_mae: 0.0900 Epoch 404/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0124 - mae: 0.0887 - val_loss: 0.0118 - val_mae: 0.0876 Epoch 405/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0123 - mae: 0.0886 - val_loss: 0.0121 - val_mae: 0.0880 Epoch 406/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0121 - mae: 0.0878 - val_loss: 0.0115 - val_mae: 0.0863 Epoch 407/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0877 - val_loss: 0.0122 - val_mae: 0.0881 Epoch 408/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0882 - val_loss: 0.0121 - val_mae: 0.0889 Epoch 409/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0880 - val_loss: 0.0127 - val_mae: 0.0910 Epoch 410/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0124 - mae: 0.0890 - val_loss: 0.0120 - val_mae: 0.0877 Epoch 411/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0894 - val_loss: 0.0125 - val_mae: 0.0889 Epoch 412/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0891 - val_loss: 0.0116 - val_mae: 0.0867 Epoch 413/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0886 - val_loss: 0.0127 - val_mae: 0.0894 Epoch 414/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0121 - mae: 0.0884 - val_loss: 0.0128 - val_mae: 0.0898 Epoch 415/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0147 - mae: 0.0959 - val_loss: 0.0121 - val_mae: 0.0878 Epoch 416/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0140 - mae: 0.0944 - val_loss: 0.0116 - val_mae: 0.0866 Epoch 417/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0889 - val_loss: 0.0120 - val_mae: 0.0880 Epoch 418/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0887 - val_loss: 0.0120 - val_mae: 0.0883 Epoch 419/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0121 - mae: 0.0878 - val_loss: 0.0119 - val_mae: 0.0879 Epoch 420/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0126 - mae: 0.0903 - val_loss: 0.0127 - val_mae: 0.0902 Epoch 421/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0126 - mae: 0.0898 - val_loss: 0.0130 - val_mae: 0.0917 Epoch 422/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0891 - val_loss: 0.0117 - val_mae: 0.0871 Epoch 423/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0880 - val_loss: 0.0119 - val_mae: 0.0875 Epoch 424/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0123 - mae: 0.0888 - val_loss: 0.0117 - val_mae: 0.0867 Epoch 425/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0873 - val_loss: 0.0118 - val_mae: 0.0868 Epoch 426/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0125 - mae: 0.0897 - val_loss: 0.0117 - val_mae: 0.0868 Epoch 427/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0119 - mae: 0.0879 - val_loss: 0.0116 - val_mae: 0.0865 Epoch 428/500 10/10 [==============================] - 0s 42ms/step - loss: 0.0122 - mae: 0.0884 - val_loss: 0.0119 - val_mae: 0.0875 Epoch 429/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0121 - mae: 0.0877 - val_loss: 0.0120 - val_mae: 0.0878 Epoch 430/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0128 - mae: 0.0918 - val_loss: 0.0123 - val_mae: 0.0885 Epoch 431/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0124 - mae: 0.0894 - val_loss: 0.0118 - val_mae: 0.0872 Epoch 432/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0873 - val_loss: 0.0120 - val_mae: 0.0878 Epoch 433/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0879 - val_loss: 0.0116 - val_mae: 0.0862 Epoch 434/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0875 - val_loss: 0.0120 - val_mae: 0.0880 Epoch 435/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0121 - mae: 0.0877 - val_loss: 0.0127 - val_mae: 0.0905 Epoch 436/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0128 - mae: 0.0901 - val_loss: 0.0127 - val_mae: 0.0906 Epoch 437/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0130 - mae: 0.0918 - val_loss: 0.0123 - val_mae: 0.0889 Epoch 438/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0881 - val_loss: 0.0116 - val_mae: 0.0867 Epoch 439/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0879 - val_loss: 0.0131 - val_mae: 0.0907 Epoch 440/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0877 - val_loss: 0.0115 - val_mae: 0.0865 Epoch 441/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0118 - mae: 0.0869 - val_loss: 0.0118 - val_mae: 0.0874 Epoch 442/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0876 - val_loss: 0.0119 - val_mae: 0.0874 Epoch 443/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0881 - val_loss: 0.0124 - val_mae: 0.0893 Epoch 444/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0123 - mae: 0.0883 - val_loss: 0.0121 - val_mae: 0.0883 Epoch 445/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0884 - val_loss: 0.0121 - val_mae: 0.0879 Epoch 446/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0882 - val_loss: 0.0115 - val_mae: 0.0864 Epoch 447/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0119 - mae: 0.0872 - val_loss: 0.0122 - val_mae: 0.0883 Epoch 448/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0122 - mae: 0.0888 - val_loss: 0.0117 - val_mae: 0.0868 Epoch 449/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0120 - mae: 0.0883 - val_loss: 0.0117 - val_mae: 0.0864 Epoch 450/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0117 - mae: 0.0873 - val_loss: 0.0130 - val_mae: 0.0919 Epoch 451/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0130 - mae: 0.0919 - val_loss: 0.0120 - val_mae: 0.0882 Epoch 452/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0119 - mae: 0.0872 - val_loss: 0.0115 - val_mae: 0.0861 Epoch 453/500 10/10 [==============================] - 0s 41ms/step - loss: 0.0118 - mae: 0.0869 - val_loss: 0.0117 - val_mae: 0.0867 Epoch 454/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0905 - val_loss: 0.0118 - val_mae: 0.0871 Epoch 455/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0122 - mae: 0.0879 - val_loss: 0.0115 - val_mae: 0.0861 Epoch 456/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0875 - val_loss: 0.0117 - val_mae: 0.0870 Epoch 457/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0118 - mae: 0.0871 - val_loss: 0.0114 - val_mae: 0.0861 Epoch 458/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0121 - mae: 0.0883 - val_loss: 0.0117 - val_mae: 0.0869 Epoch 459/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0884 - val_loss: 0.0116 - val_mae: 0.0865 Epoch 460/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0878 - val_loss: 0.0119 - val_mae: 0.0876 Epoch 461/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0875 - val_loss: 0.0115 - val_mae: 0.0863 Epoch 462/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0874 - val_loss: 0.0117 - val_mae: 0.0872 Epoch 463/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0881 - val_loss: 0.0118 - val_mae: 0.0873 Epoch 464/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0877 - val_loss: 0.0116 - val_mae: 0.0864 Epoch 465/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0118 - mae: 0.0867 - val_loss: 0.0115 - val_mae: 0.0864 Epoch 466/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0118 - mae: 0.0874 - val_loss: 0.0114 - val_mae: 0.0860 Epoch 467/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0124 - mae: 0.0892 - val_loss: 0.0116 - val_mae: 0.0868 Epoch 468/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0878 - val_loss: 0.0117 - val_mae: 0.0871 Epoch 469/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0120 - mae: 0.0885 - val_loss: 0.0115 - val_mae: 0.0862 Epoch 470/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0121 - mae: 0.0878 - val_loss: 0.0128 - val_mae: 0.0900 Epoch 471/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0887 - val_loss: 0.0114 - val_mae: 0.0860 Epoch 472/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0119 - mae: 0.0875 - val_loss: 0.0114 - val_mae: 0.0860 Epoch 473/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0122 - mae: 0.0882 - val_loss: 0.0129 - val_mae: 0.0903 Epoch 474/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0137 - mae: 0.0944 - val_loss: 0.0130 - val_mae: 0.0905 Epoch 475/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0129 - mae: 0.0908 - val_loss: 0.0119 - val_mae: 0.0878 Epoch 476/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0904 - val_loss: 0.0114 - val_mae: 0.0860 Epoch 477/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0120 - mae: 0.0881 - val_loss: 0.0117 - val_mae: 0.0873 Epoch 478/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0123 - mae: 0.0887 - val_loss: 0.0113 - val_mae: 0.0858 Epoch 479/500 10/10 [==============================] - 0s 43ms/step - loss: 0.0120 - mae: 0.0875 - val_loss: 0.0115 - val_mae: 0.0863 Epoch 480/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0117 - mae: 0.0869 - val_loss: 0.0117 - val_mae: 0.0871 Epoch 481/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0119 - mae: 0.0871 - val_loss: 0.0114 - val_mae: 0.0861 Epoch 482/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0119 - mae: 0.0875 - val_loss: 0.0114 - val_mae: 0.0861 Epoch 483/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0118 - mae: 0.0869 - val_loss: 0.0124 - val_mae: 0.0895 Epoch 484/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0123 - mae: 0.0882 - val_loss: 0.0116 - val_mae: 0.0870 Epoch 485/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0118 - mae: 0.0869 - val_loss: 0.0123 - val_mae: 0.0886 Epoch 486/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0876 - val_loss: 0.0114 - val_mae: 0.0859 Epoch 487/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0117 - mae: 0.0866 - val_loss: 0.0118 - val_mae: 0.0875 Epoch 488/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0874 - val_loss: 0.0114 - val_mae: 0.0859 Epoch 489/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0877 - val_loss: 0.0130 - val_mae: 0.0917 Epoch 490/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0127 - mae: 0.0903 - val_loss: 0.0131 - val_mae: 0.0921 Epoch 491/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0129 - mae: 0.0911 - val_loss: 0.0120 - val_mae: 0.0884 Epoch 492/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0122 - mae: 0.0885 - val_loss: 0.0114 - val_mae: 0.0861 Epoch 493/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0879 - val_loss: 0.0114 - val_mae: 0.0862 Epoch 494/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0118 - mae: 0.0870 - val_loss: 0.0116 - val_mae: 0.0866 Epoch 495/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0121 - mae: 0.0876 - val_loss: 0.0113 - val_mae: 0.0858 Epoch 496/500 10/10 [==============================] - 0s 38ms/step - loss: 0.0119 - mae: 0.0875 - val_loss: 0.0114 - val_mae: 0.0862 Epoch 497/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0870 - val_loss: 0.0115 - val_mae: 0.0863 Epoch 498/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0120 - mae: 0.0886 - val_loss: 0.0114 - val_mae: 0.0860 Epoch 499/500 10/10 [==============================] - 0s 40ms/step - loss: 0.0119 - mae: 0.0877 - val_loss: 0.0113 - val_mae: 0.0857 Epoch 500/500 10/10 [==============================] - 0s 39ms/step - loss: 0.0120 - mae: 0.0877 - val_loss: 0.0115 - val_mae: 0.0865
2022-03-18 13:31:56.675996: W tensorflow/python/util/util.cc:368] Sets are not currently considered sequences, but this may change in the future, so consider avoiding using them.
INFO:tensorflow:Assets written to: models/model/assets
Convert the model to the TensorFlow Lite format without quantization¶
- Plot Metrics Each training epoch, the model prints out its loss and mean absolute error for training and validation. You can read this in the output above (note that your exact numbers may differ):
Epoch 500/500 10/10 [==============================] - 0s 10ms/step - loss: 0.0121 - mae: 0.0882 - val_loss: 0.0115 - val_mae: 0.0865 You can see that we've already got a huge improvement - validation loss has dropped from 0.15 to 0.01, and validation MAE has dropped from 0.33 to 0.08.
The following cell will print the same graphs we used to evaluate our original model, but showing our new training history:
In [27]:
# Draw a graph of the loss, which is the distance between
# the predicted and actual values during training and validation.
train_loss = history.history['loss']
val_loss = history.history['val_loss']
epochs = range(1, len(train_loss) + 1)
# Exclude the first few epochs so the graph is easier to read
SKIP = 100
plt.figure(figsize=(10, 4))
plt.subplot(1, 2, 1)
plt.plot(epochs[SKIP:], train_loss[SKIP:], 'g.', label='Training loss')
plt.plot(epochs[SKIP:], val_loss[SKIP:], 'b.', label='Validation loss')
plt.title('Training and validation loss')
plt.xlabel('Epochs')
plt.ylabel('Loss')
plt.legend()
plt.subplot(1, 2, 2)
# Draw a graph of mean absolute error, which is another way of
# measuring the amount of error in the prediction.
train_mae = history.history['mae']
val_mae = history.history['val_mae']
plt.plot(epochs[SKIP:], train_mae[SKIP:], 'g.', label='Training MAE')
plt.plot(epochs[SKIP:], val_mae[SKIP:], 'b.', label='Validation MAE')
plt.title('Training and validation mean absolute error')
plt.xlabel('Epochs')
plt.ylabel('MAE')
plt.legend()
plt.tight_layout()
Great results! From these graphs, we can see several exciting things:
The overall loss and MAE are much better than our previous network Metrics are better for validation than training, which means the network is not overfitting The reason the metrics for validation are better than those for training is that validation metrics are calculated at the end of each epoch, while training metrics are calculated throughout the epoch, so validation happens on a model that has been trained slightly longer.
This all means our network seems to be performing well! To confirm, let's check its predictions against the test dataset we set aside earlier:
In [28]:
# Calculate and print the loss on our test dataset
test_loss, test_mae = model.evaluate(x_test, y_test)
# Make predictions based on our test dataset
y_test_pred = model.predict(x_test)
# Graph the predictions against the actual values
plt.clf()
plt.title('Comparison of predictions and actual values')
plt.plot(x_test, y_test, 'b.', label='Actual values')
plt.plot(x_test, y_test_pred, 'r.', label='TF predicted')
plt.legend()
plt.show()
7/7 [==============================] - 0s 14ms/step - loss: 0.0102 - mae: 0.0811
Much better! The evaluation metrics we printed show that the model has a low loss and MAE on the test data, and the predictions line up visually with our data fairly well.
The model isn't perfect; its predictions don't form a smooth sine curve. For instance, the line is almost straight when x is between 4.2 and 5.2. If we wanted to go further, we could try further increasing the capacity of the model, perhaps using some techniques to defend from overfitting.
However, an important part of machine learning is knowing when to stop. This model is good enough for our use case - which is to make some LEDs blink in a pleasing pattern.
Generate a TensorFlow Lite Model ¶
1. Generate Models with or without Quantization ¶
We now have an acceptably accurate model. We'll use the TensorFlow Lite Converter to convert the model into a special, space-efficient format for use on memory-constrained devices.
Since this model is going to be deployed on a microcontroller, we want it to be as tiny as possible! One technique for reducing the size of a model is called quantization. It reduces the precision of the model's weights, and possibly the activations (output of each layer) as well, which saves memory, often without much impact on accuracy. Quantized models also run faster, since the calculations required are simpler.
In the following cell, we'll convert the model twice: once with quantization, once without.
In [29]:
# Convert the model to the TensorFlow Lite format without quantization
converter = tf.lite.TFLiteConverter.from_saved_model(MODEL_TF)
model_no_quant_tflite = converter.convert()
# Save the model to disk
open(MODEL_NO_QUANT_TFLITE, "wb").write(model_no_quant_tflite)
# Convert the model to the TensorFlow Lite format with quantization
def representative_dataset():
for i in range(500):
yield([x_train[i].reshape(1, 1)])
# Set the optimization flag.
converter.optimizations = [tf.lite.Optimize.DEFAULT]
# Enforce integer only quantization
converter.target_spec.supported_ops = [tf.lite.OpsSet.TFLITE_BUILTINS_INT8]
converter.inference_input_type = tf.int8
converter.inference_output_type = tf.int8
# Provide a representative dataset to ensure we quantize correctly.
converter.representative_dataset = representative_dataset
model_tflite = converter.convert()
# Save the model to disk
open(MODEL_TFLITE, "wb").write(model_tflite)
2022-03-18 13:33:52.085501: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:357] Ignored output_format. 2022-03-18 13:33:52.085668: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:360] Ignored drop_control_dependency. WARNING:absl:Buffer deduplication procedure will be skipped when flatbuffer library is not properly loaded 2022-03-18 13:33:53.841980: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:357] Ignored output_format. 2022-03-18 13:33:53.842167: W tensorflow/compiler/mlir/lite/python/tf_tfl_flatbuffer_helpers.cc:360] Ignored drop_control_dependency. fully_quantize: 0, inference_type: 6, input_inference_type: 9, output_inference_type: 9 WARNING:absl:Buffer deduplication procedure will be skipped when flatbuffer library is not properly loaded
Out[29]:
2408
2. Compare Model Performance¶
To prove these models are accurate even after conversion and quantization, we'll compare their predictions and loss on our test dataset.
Helper functions
We define the predict (for predictions) and evaluate (for loss) functions for TFLite models. Note: These are already included in a TF model, but not in a TFLite model.
In [30]:
def predict_tflite(tflite_model, x_test):
# Prepare the test data
x_test_ = x_test.copy()
x_test_ = x_test_.reshape((x_test.size, 1))
x_test_ = x_test_.astype(np.float32)
# Initialize the TFLite interpreter
interpreter = tf.lite.Interpreter(model_content=tflite_model)
interpreter.allocate_tensors()
input_details = interpreter.get_input_details()[0]
output_details = interpreter.get_output_details()[0]
# If required, quantize the input layer (from float to integer)
input_scale, input_zero_point = input_details["quantization"]
if (input_scale, input_zero_point) != (0.0, 0):
x_test_ = x_test_ / input_scale + input_zero_point
x_test_ = x_test_.astype(input_details["dtype"])
# Invoke the interpreter
y_pred = np.empty(x_test_.size, dtype=output_details["dtype"])
for i in range(len(x_test_)):
interpreter.set_tensor(input_details["index"], [x_test_[i]])
interpreter.invoke()
y_pred[i] = interpreter.get_tensor(output_details["index"])[0]
# If required, dequantized the output layer (from integer to float)
output_scale, output_zero_point = output_details["quantization"]
if (output_scale, output_zero_point) != (0.0, 0):
y_pred = y_pred.astype(np.float32)
y_pred = (y_pred - output_zero_point) * output_scale
return y_pred
def evaluate_tflite(tflite_model, x_test, y_true):
global model
y_pred = predict_tflite(tflite_model, x_test)
loss_function = tf.keras.losses.get(model.loss)
loss = loss_function(y_true, y_pred).numpy()
return loss
1. Predictions¶
In [31]:
# Calculate predictions
y_test_pred_tf = model.predict(x_test)
y_test_pred_no_quant_tflite = predict_tflite(model_no_quant_tflite, x_test)
y_test_pred_tflite = predict_tflite(model_tflite, x_test)
INFO: Created TensorFlow Lite XNNPACK delegate for CPU.
In [32]:
# Compare predictions
plt.clf()
plt.title('Comparison of various models against actual values')
plt.plot(x_test, y_test, 'bo', label='Actual values')
plt.plot(x_test, y_test_pred_tf, 'ro', label='TF predictions')
plt.plot(x_test, y_test_pred_no_quant_tflite, 'bx', label='TFLite predictions')
plt.plot(x_test, y_test_pred_tflite, 'gx', label='TFLite quantized predictions')
plt.legend()
plt.show()
2. Loss (MSE/Mean Squared Error)¶
In [33]:
# Calculate loss
loss_tf, _ = model.evaluate(x_test, y_test, verbose=0)
loss_no_quant_tflite = evaluate_tflite(model_no_quant_tflite, x_test, y_test)
loss_tflite = evaluate_tflite(model_tflite, x_test, y_test)
In [34]:
# Compare loss
df = pd.DataFrame.from_records(
[["TensorFlow", loss_tf],
["TensorFlow Lite", loss_no_quant_tflite],
["TensorFlow Lite Quantized", loss_tflite]],
columns = ["Model", "Loss/MSE"], index="Model").round(4)
df
Out[34]:
| Loss/MSE | |
|---|---|
| Model | |
| TensorFlow | 0.0102 |
| TensorFlow Lite | 0.0102 |
| TensorFlow Lite Quantized | 0.0108 |
3. Size¶
In [35]:
# Calculate size
size_tf = os.path.getsize(MODEL_TF)
size_no_quant_tflite = os.path.getsize(MODEL_NO_QUANT_TFLITE)
size_tflite = os.path.getsize(MODEL_TFLITE)
In [36]:
# Compare size
pd.DataFrame.from_records(
[["TensorFlow", f"{size_tf} bytes", ""],
["TensorFlow Lite", f"{size_no_quant_tflite} bytes ", f"(reduced by {size_tf - size_no_quant_tflite} bytes)"],
["TensorFlow Lite Quantized", f"{size_tflite} bytes", f"(reduced by {size_no_quant_tflite - size_tflite} bytes)"]],
columns = ["Model", "Size", ""], index="Model")
Out[36]:
| Size | ||
|---|---|---|
| Model | ||
| TensorFlow | 84 bytes | |
| TensorFlow Lite | 2932 bytes | (reduced by -2848 bytes) |
| TensorFlow Lite Quantized | 2408 bytes | (reduced by 524 bytes) |
Summary
We can see from the predictions (graph) and loss (table) that the original TF model, the TFLite model, and the quantized TFLite model are all close enough to be indistinguishable - even though they differ in size (table). This implies that the quantized (smallest) model is ready to use!
Note: The quantized (integer) TFLite model is just 300 bytes smaller than the original (float) TFLite model - a tiny reduction in size! This is because the model is already so small that quantization has little effect. Complex models with more weights, can have upto a 4x reduction in size!
Generate a TensorFlow Lite for Microcontrollers Model¶
Convert the TensorFlow Lite quantized model into a C source file that can be loaded by TensorFlow Lite for Microcontrollers.
In [38]:
# Install xxd if it is not available
#!apt-get update && apt-get -qq install xxd
# Convert to a C source file, i.e, a TensorFlow Lite for Microcontrollers model
!xxd -i {MODEL_TFLITE} > {MODEL_TFLITE_MICRO}
# Update variable names
REPLACE_TEXT = MODEL_TFLITE.replace('/', '_').replace('.', '_')
!sed -i 's/'{REPLACE_TEXT}'/g_model/g' {MODEL_TFLITE_MICRO}
Deploy to a Microcontroller¶
Follow the instructions in the hello_world README.md for TensorFlow Lite for MicroControllers to deploy this model on a specific microcontroller.
Reference Model: If you have not modified this notebook, you can follow the instructions as is, to deploy the model. Refer to the hello_world/train/models directory to access the models generated in this notebook.
New Model: If you have generated a new model, then update the values assigned to the variables defined in hello_world/model.cc with values displayed after running the following cell.
In [41]:
# Print the C source file
print(MODEL_TFLITE_MICRO)
!cat {MODEL_TFLITE_MICRO}
models/model.cc
unsigned char g_model[] = {
0x20, 0x00, 0x00, 0x00, 0x54, 0x46, 0x4c, 0x33, 0x00, 0x00, 0x00, 0x00,
0x14, 0x00, 0x20, 0x00, 0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x10, 0x00,
0x14, 0x00, 0x00, 0x00, 0x18, 0x00, 0x1c, 0x00, 0x14, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00,
0x28, 0x01, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00, 0xd0, 0x00, 0x00, 0x00,
0x48, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x98, 0x03, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x30, 0x01, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x10, 0x09, 0x00, 0x00, 0x0c, 0x09, 0x00, 0x00, 0xdc, 0x07, 0x00, 0x00,
0x30, 0x07, 0x00, 0x00, 0xbc, 0x05, 0x00, 0x00, 0x18, 0x05, 0x00, 0x00,
0x94, 0x04, 0x00, 0x00, 0x2c, 0x04, 0x00, 0x00, 0xf0, 0x08, 0x00, 0x00,
0xec, 0x08, 0x00, 0x00, 0xe8, 0x08, 0x00, 0x00, 0xbc, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0a, 0x00,
0x10, 0x00, 0x0c, 0x00, 0x08, 0x00, 0x04, 0x00, 0x0a, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x1c, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00,
0x0f, 0x00, 0x00, 0x00, 0x73, 0x65, 0x72, 0x76, 0x69, 0x6e, 0x67, 0x5f,
0x64, 0x65, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x00, 0x01, 0x00, 0x00, 0x00,
0x14, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x24, 0x00, 0x00, 0x00,
0x08, 0x00, 0x0c, 0x00, 0x08, 0x00, 0x04, 0x00, 0x08, 0x00, 0x00, 0x00,
0x09, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00,
0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x34, 0x00, 0xba, 0xf8, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x0d, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73,
0x65, 0x5f, 0x32, 0x5f, 0x69, 0x6e, 0x70, 0x75, 0x74, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00,
0x04, 0x00, 0x08, 0x00, 0x08, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x0b, 0x00, 0x00, 0x00, 0x13, 0x00, 0x00, 0x00, 0x6d, 0x69, 0x6e, 0x5f,
0x72, 0x75, 0x6e, 0x74, 0x69, 0x6d, 0x65, 0x5f, 0x76, 0x65, 0x72, 0x73,
0x69, 0x6f, 0x6e, 0x00, 0x0a, 0xf9, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x31, 0x2e, 0x31, 0x34, 0x2e, 0x30, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0f, 0x00, 0x00, 0x00,
0x4d, 0x4c, 0x49, 0x52, 0x20, 0x43, 0x6f, 0x6e, 0x76, 0x65, 0x72, 0x74,
0x65, 0x64, 0x2e, 0x00, 0x00, 0x00, 0x0e, 0x00, 0x18, 0x00, 0x04, 0x00,
0x08, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x14, 0x00, 0x0e, 0x00, 0x00, 0x00,
0x14, 0x00, 0x00, 0x00, 0x3c, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00,
0x44, 0x00, 0x00, 0x00, 0x50, 0x00, 0x00, 0x00, 0x0a, 0x00, 0x00, 0x00,
0x44, 0x07, 0x00, 0x00, 0xc8, 0x06, 0x00, 0x00, 0x3c, 0x06, 0x00, 0x00,
0x88, 0x05, 0x00, 0x00, 0x24, 0x04, 0x00, 0x00, 0x70, 0x03, 0x00, 0x00,
0xfc, 0x02, 0x00, 0x00, 0x38, 0x02, 0x00, 0x00, 0x1c, 0x01, 0x00, 0x00,
0x60, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0xc0, 0x01, 0x00, 0x00, 0xc8, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x6d, 0x61, 0x69, 0x6e, 0x00, 0x00, 0x00, 0x00,
0x5a, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x08, 0x18, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0xa0, 0xf8, 0xff, 0xff,
0x01, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0x08, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00, 0x06, 0x00, 0x00, 0x00,
0x54, 0xf9, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09, 0x14, 0x00, 0x00, 0x00,
0x0a, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x40, 0x00, 0x00, 0x00,
0x30, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x19, 0x00, 0x00, 0x00, 0x53, 0x74, 0x61, 0x74,
0x65, 0x66, 0x75, 0x6c, 0x50, 0x61, 0x72, 0x74, 0x69, 0x74, 0x69, 0x6f,
0x6e, 0x65, 0x64, 0x43, 0x61, 0x6c, 0x6c, 0x3a, 0x30, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0xff, 0xff, 0xff, 0xff, 0x01, 0x00, 0x00, 0x00,
0x3c, 0xf9, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x6b, 0x53, 0x0b, 0x3c, 0x01, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x0e, 0x00,
0x14, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x07, 0x00, 0x10, 0x00,
0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08, 0x1c, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00, 0x0a, 0xff, 0xff, 0xff,
0x00, 0x00, 0x00, 0x01, 0x01, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x0c, 0xfa, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09,
0x14, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00,
0x78, 0x00, 0x00, 0x00, 0x68, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x52, 0x00, 0x00, 0x00,
0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31,
0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33, 0x2f, 0x4d, 0x61, 0x74,
0x4d, 0x75, 0x6c, 0x3b, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33,
0x2f, 0x52, 0x65, 0x6c, 0x75, 0x3b, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e,
0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65,
0x5f, 0x33, 0x2f, 0x42, 0x69, 0x61, 0x73, 0x41, 0x64, 0x64, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0xff, 0xff, 0xff, 0xff, 0x10, 0x00, 0x00, 0x00,
0x2c, 0xfa, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x17, 0xd7, 0xbe, 0x3b, 0x01, 0x00, 0x00, 0x00,
0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x0e, 0x00, 0x16, 0x00, 0x00, 0x00, 0x08, 0x00, 0x0c, 0x00,
0x07, 0x00, 0x10, 0x00, 0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x08,
0x24, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x00, 0x00, 0x06, 0x00, 0x08, 0x00, 0x07, 0x00, 0x06, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x01, 0x01, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00,
0x03, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x0b, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x04, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x09, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x09, 0x04, 0x00, 0x00, 0x00, 0x24, 0xfb, 0xff, 0xff,
0x00, 0x00, 0x00, 0x09, 0x14, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x18, 0x00, 0x00, 0x00, 0x78, 0x00, 0x00, 0x00, 0x68, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x52, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x3b, 0x73, 0x65, 0x71, 0x75,
0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e,
0x73, 0x65, 0x5f, 0x32, 0x2f, 0x52, 0x65, 0x6c, 0x75, 0x3b, 0x73, 0x65,
0x71, 0x75, 0x65, 0x6e, 0x74, 0x69, 0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64,
0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x2f, 0x42, 0x69, 0x61, 0x73, 0x41,
0x64, 0x64, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0xff, 0xff, 0xff, 0xff,
0x10, 0x00, 0x00, 0x00, 0x44, 0xfb, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x8f, 0xb7, 0x4a, 0x3c,
0x01, 0x00, 0x00, 0x00, 0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0x00, 0x00, 0x00, 0x00, 0x6a, 0xfc, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x80, 0xef, 0xff, 0xff, 0x56, 0xfc, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x10, 0x00, 0x00, 0x00, 0x07, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73,
0x65, 0x5f, 0x34, 0x2f, 0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00,
0xac, 0xfb, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x73, 0x59, 0x4f, 0x38, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xce, 0xfc, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x21, 0xa2, 0x8b, 0xc9,
0x5e, 0x1d, 0xce, 0x42, 0x9c, 0xce, 0x1f, 0xb1, 0xdf, 0x54, 0x2f, 0x81,
0xc6, 0xfc, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09, 0x10, 0x00, 0x00, 0x00,
0x06, 0x00, 0x00, 0x00, 0x14, 0x00, 0x00, 0x00, 0x30, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x1b, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x34,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x00, 0x2c, 0xfc, 0xff, 0xff,
0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xb2, 0x12, 0x0b, 0x3c, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x4e, 0xfd, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x40, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xdf, 0x04, 0x00, 0x00,
0xb5, 0x0a, 0x00, 0x00, 0x16, 0x06, 0x00, 0x00, 0x79, 0xf8, 0xff, 0xff,
0x00, 0x00, 0x00, 0x00, 0xa7, 0x0a, 0x00, 0x00, 0x0c, 0xf8, 0xff, 0xff,
0x1d, 0x05, 0x00, 0x00, 0xd9, 0x09, 0x00, 0x00, 0x45, 0xfe, 0xff, 0xff,
0xf1, 0x04, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xb8, 0xf7, 0xff, 0xff,
0x53, 0xf9, 0xff, 0xff, 0x6e, 0x05, 0x00, 0x00, 0x76, 0xfd, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x10, 0x00, 0x00, 0x00, 0x05, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73,
0x65, 0x5f, 0x33, 0x2f, 0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00,
0xcc, 0xfc, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0xed, 0x35, 0x21, 0x39, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xee, 0xfd, 0xff, 0xff,
0x04, 0x00, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0xee, 0xfc, 0x00, 0xec,
0x05, 0x17, 0xef, 0xeb, 0xe5, 0xf8, 0x03, 0x01, 0x00, 0xfa, 0xf8, 0xf5,
0xdb, 0xe9, 0x27, 0x14, 0xf1, 0xdd, 0xe1, 0xda, 0xf0, 0xde, 0x33, 0x06,
0x02, 0xe4, 0xed, 0xf9, 0x01, 0x17, 0x08, 0xdf, 0xff, 0xff, 0xe8, 0x06,
0xe6, 0xee, 0x81, 0x1c, 0x19, 0xea, 0xc8, 0x01, 0x0f, 0x00, 0xda, 0xf7,
0x0e, 0xeb, 0x13, 0x1f, 0x04, 0x13, 0xb1, 0xe6, 0xfd, 0x06, 0xb8, 0xdf,
0x0d, 0xec, 0xef, 0xde, 0xeb, 0xf7, 0x05, 0x27, 0x1a, 0xe4, 0x72, 0x1a,
0xea, 0x1f, 0x37, 0xde, 0x1a, 0xf3, 0xf0, 0x1a, 0x0f, 0x03, 0x1c, 0xe1,
0xdd, 0x14, 0xf6, 0x1a, 0xff, 0xf6, 0x1b, 0x18, 0xf0, 0x1d, 0xda, 0x1b,
0x1b, 0x21, 0xe5, 0x1a, 0xf5, 0xff, 0x95, 0x0b, 0x00, 0x01, 0xcb, 0xde,
0x0d, 0xf6, 0x16, 0xe3, 0xec, 0xfc, 0x0e, 0xe8, 0xfa, 0xeb, 0x5f, 0xfc,
0x1e, 0x02, 0x5d, 0xe2, 0xe1, 0xf2, 0x15, 0xec, 0xf5, 0x00, 0x14, 0x06,
0xec, 0x0c, 0x1e, 0x15, 0x0e, 0xe6, 0x0b, 0xf4, 0x19, 0x00, 0xd7, 0x05,
0x27, 0x02, 0x15, 0xe9, 0xea, 0x02, 0x99, 0x00, 0x0c, 0xfa, 0xe8, 0xe9,
0xfd, 0x00, 0x15, 0xfc, 0x0b, 0x02, 0xef, 0xee, 0x06, 0xee, 0x01, 0x0d,
0x06, 0xe6, 0xf6, 0x11, 0xf8, 0x0d, 0xf9, 0xf0, 0x28, 0xff, 0x0e, 0xf3,
0xeb, 0x12, 0x27, 0x1d, 0xf3, 0xe9, 0x29, 0x19, 0xdf, 0xfb, 0xf3, 0xf4,
0x05, 0x1d, 0x1e, 0xfb, 0xfd, 0x1e, 0xfc, 0x12, 0xe7, 0x07, 0x09, 0x03,
0x12, 0xf2, 0x36, 0xfa, 0xdc, 0x1c, 0xf9, 0xef, 0xf3, 0xe6, 0x71, 0x0c,
0x1e, 0x00, 0x47, 0xfd, 0x0e, 0xf0, 0x0a, 0x1a, 0x1b, 0xfa, 0xdf, 0x19,
0x1f, 0x14, 0x37, 0x1d, 0x12, 0xeb, 0x3d, 0x0c, 0xb5, 0xc8, 0xe6, 0x13,
0xf9, 0xea, 0xf1, 0x06, 0x1c, 0xfa, 0x19, 0xe4, 0xec, 0xca, 0x0c, 0xf3,
0xd6, 0xfe, 0xff, 0xff, 0x00, 0x00, 0x00, 0x09, 0x10, 0x00, 0x00, 0x00,
0x04, 0x00, 0x00, 0x00, 0x14, 0x00, 0x00, 0x00, 0x30, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00,
0x1b, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x33,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x00, 0x3c, 0xfe, 0xff, 0xff,
0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x7b, 0x95, 0x4b, 0x3c, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x5e, 0xff, 0xff, 0xff, 0x04, 0x00, 0x00, 0x00,
0x40, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00, 0x7c, 0x0b, 0x00, 0x00,
0xa4, 0xf6, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00, 0x42, 0x0c, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xd0, 0x06, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x7f, 0x21, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0xcb, 0x07, 0x00, 0x00, 0x25, 0xf5, 0xff, 0xff,
0x19, 0xf0, 0xff, 0xff, 0x00, 0x00, 0x00, 0x00, 0x86, 0xff, 0xff, 0xff,
0x00, 0x00, 0x00, 0x02, 0x10, 0x00, 0x00, 0x00, 0x03, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x20, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x64, 0x65, 0x6e, 0x73,
0x65, 0x5f, 0x32, 0x2f, 0x62, 0x69, 0x61, 0x73, 0x00, 0x00, 0x00, 0x00,
0xdc, 0xfe, 0xff, 0xff, 0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x28, 0xb3, 0xd9, 0x38, 0x01, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x06, 0x00,
0x08, 0x00, 0x04, 0x00, 0x06, 0x00, 0x00, 0x00, 0x04, 0x00, 0x00, 0x00,
0x10, 0x00, 0x00, 0x00, 0x77, 0x1d, 0x10, 0xe1, 0x0c, 0x81, 0xa5, 0x43,
0xfe, 0xd5, 0xd4, 0xb2, 0x63, 0x76, 0x1a, 0xdf, 0x00, 0x00, 0x0e, 0x00,
0x18, 0x00, 0x08, 0x00, 0x07, 0x00, 0x0c, 0x00, 0x10, 0x00, 0x14, 0x00,
0x0e, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09, 0x10, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x14, 0x00, 0x00, 0x00, 0x30, 0x00, 0x00, 0x00,
0x02, 0x00, 0x00, 0x00, 0x10, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x1b, 0x00, 0x00, 0x00, 0x73, 0x65, 0x71, 0x75, 0x65, 0x6e, 0x74, 0x69,
0x61, 0x6c, 0x5f, 0x31, 0x2f, 0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32,
0x2f, 0x4d, 0x61, 0x74, 0x4d, 0x75, 0x6c, 0x00, 0x74, 0xff, 0xff, 0xff,
0x08, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0xd5, 0x6b, 0x8a, 0x3b, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x14, 0x00, 0x1c, 0x00, 0x08, 0x00, 0x07, 0x00,
0x0c, 0x00, 0x10, 0x00, 0x14, 0x00, 0x00, 0x00, 0x00, 0x00, 0x18, 0x00,
0x14, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x09, 0x14, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x18, 0x00, 0x00, 0x00, 0x50, 0x00, 0x00, 0x00,
0x34, 0x00, 0x00, 0x00, 0x02, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00,
0x01, 0x00, 0x00, 0x00, 0x1f, 0x00, 0x00, 0x00, 0x73, 0x65, 0x72, 0x76,
0x69, 0x6e, 0x67, 0x5f, 0x64, 0x65, 0x66, 0x61, 0x75, 0x6c, 0x74, 0x5f,
0x64, 0x65, 0x6e, 0x73, 0x65, 0x5f, 0x32, 0x5f, 0x69, 0x6e, 0x70, 0x75,
0x74, 0x3a, 0x30, 0x00, 0x02, 0x00, 0x00, 0x00, 0xff, 0xff, 0xff, 0xff,
0x01, 0x00, 0x00, 0x00, 0x0c, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x00, 0x00,
0x04, 0x00, 0x08, 0x00, 0x0c, 0x00, 0x00, 0x00, 0x08, 0x00, 0x00, 0x00,
0x0c, 0x00, 0x00, 0x00, 0x01, 0x00, 0x00, 0x00, 0x5d, 0x4f, 0xc9, 0x3c,
0x01, 0x00, 0x00, 0x00, 0x80, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0x04, 0x00, 0x04, 0x00, 0x04, 0x00, 0x00, 0x00
};
unsigned int g_model_len = 2408;
In [ ]: