{ "cells": [ { "cell_type": "markdown", "metadata": { "id": "FhGuhbZ6M5tl" }, "source": [ "##### Copyright 2018 The TensorFlow Authors.\n" ] }, { "cell_type": "code", "execution_count": 1, "metadata": { "cellView": "form", "id": "AwOEIRJC6Une" }, "outputs": [], "source": [ "# @title Licensed under the Apache License, Version 2.0 (the \"License\");\n", "# you may not use this file except in compliance with the License.\n", "# You may obtain a copy of the License at\n", "#\n", "# https://www.apache.org/licenses/LICENSE-2.0\n", "#\n", "# Unless required by applicable law or agreed to in writing, software\n", "# distributed under the License is distributed on an \"AS IS\" BASIS,\n", "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n", "# See the License for the specific language governing permissions and\n", "# limitations under the License." ] }, { "cell_type": "code", "execution_count": 2, "metadata": { "cellView": "form", "id": "KyPEtTqk6VdG" }, "outputs": [], "source": [ "# @title MIT License\n", "#\n", "# Copyright (c) 2017 François Chollet\n", "#\n", "# Permission is hereby granted, free of charge, to any person obtaining a\n", "# copy of this software and associated documentation files (the \"Software\"),\n", "# to deal in the Software without restriction, including without limitation\n", "# the rights to use, copy, modify, merge, publish, distribute, sublicense,\n", "# and/or sell copies of the Software, and to permit persons to whom the\n", "# Software is furnished to do so, subject to the following conditions:\n", "#\n", "# The above copyright notice and this permission notice shall be included in\n", "# all copies or substantial portions of the Software.\n", "#\n", "# THE SOFTWARE IS PROVIDED \"AS IS\", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR\n", "# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,\n", "# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL\n", "# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER\n", "# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING\n", "# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER\n", "# DEALINGS IN THE SOFTWARE." ] }, { "cell_type": "markdown", "metadata": { "id": "EIdT9iu_Z4Rb" }, "source": [ "# Basic regression: Predict fuel efficiency\n" ] }, { "cell_type": "markdown", "metadata": { "id": "bBIlTPscrIT9" }, "source": [ "\n", " \n", " \n", " \n", " \n", "
\n", " View on TensorFlow.org\n", " \n", " Run in Google Colab\n", " \n", " View source on GitHub\n", " \n", " Download notebook\n", "
\n" ] }, { "cell_type": "markdown", "metadata": { "id": "AHp3M9ZmrIxj" }, "source": [ "In a _regression_ problem, the aim is to predict the output of a continuous value, like a price or a probability. Contrast this with a _classification_ problem, where the aim is to select a class from a list of classes (for example, where a picture contains an apple or an orange, recognizing which fruit is in the picture).\n", "\n", "This tutorial uses the classic [Auto MPG](https://archive.ics.uci.edu/ml/datasets/auto+mpg) dataset and demonstrates how to build models to predict the fuel efficiency of the late-1970s and early 1980s automobiles. To do this, you will provide the models with a description of many automobiles from that time period. This description includes attributes like cylinders, displacement, horsepower, and weight.\n", "\n", "This example uses the Keras API. (Visit the Keras [tutorials](https://www.tensorflow.org/tutorials/keras) and [guides](https://www.tensorflow.org/guide/keras) to learn more.)\n" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "id": "1rRo8oNqZ-Rj" }, "outputs": [], "source": [ "import matplotlib.pyplot as plt\n", "import numpy as np\n", "import pandas as pd\n", "import seaborn as sns\n", "\n", "# Make NumPy printouts easier to read.\n", "np.set_printoptions(precision=3, suppress=True)" ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "id": "9xQKvCJ85kCQ" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2.16.1\n" ] } ], "source": [ "import tensorflow as tf\n", "\n", "from tensorflow import keras\n", "from tensorflow.keras import layers\n", "\n", "print(tf.__version__)" ] }, { "cell_type": "markdown", "metadata": { "id": "F_72b0LCNbjx" }, "source": [ "## The Auto MPG dataset\n", "\n", "The dataset is available from the [UCI Machine Learning Repository](https://archive.ics.uci.edu/ml/).\n" ] }, { "cell_type": "markdown", "metadata": { "id": "gFh9ne3FZ-On" }, "source": [ "### Get the data\n", "\n", "First download and import the dataset using pandas:\n" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "id": "CiX2FI4gZtTt" }, "outputs": [], "source": [ "url = 'http://archive.ics.uci.edu/ml/machine-learning-databases/auto-mpg/auto-mpg.data'\n", "column_names = ['MPG', 'Cylinders', 'Displacement', 'Horsepower', 'Weight',\n", " 'Acceleration', 'Model Year', 'Origin']\n", "\n", "raw_dataset = pd.read_csv(url, names=column_names,\n", " na_values='?', comment='\\t',\n", " sep=' ', skipinitialspace=True)" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "id": "2oY3pMPagJrO" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
MPGCylindersDisplacementHorsepowerWeightAccelerationModel YearOrigin
39327.04140.086.02790.015.6821
39444.0497.052.02130.024.6822
39532.04135.084.02295.011.6821
39628.04120.079.02625.018.6821
39731.04119.082.02720.019.4821
\n", "
" ], "text/plain": [ " MPG Cylinders Displacement Horsepower Weight Acceleration \\\n", "393 27.0 4 140.0 86.0 2790.0 15.6 \n", "394 44.0 4 97.0 52.0 2130.0 24.6 \n", "395 32.0 4 135.0 84.0 2295.0 11.6 \n", "396 28.0 4 120.0 79.0 2625.0 18.6 \n", "397 31.0 4 119.0 82.0 2720.0 19.4 \n", "\n", " Model Year Origin \n", "393 82 1 \n", "394 82 2 \n", "395 82 1 \n", "396 82 1 \n", "397 82 1 " ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dataset = raw_dataset.copy()\n", "dataset.tail()" ] }, { "cell_type": "markdown", "metadata": { "id": "3MWuJTKEDM-f" }, "source": [ "### Clean the data\n", "\n", "The dataset contains a few unknown values:\n" ] }, { "cell_type": "code", "execution_count": 7, "metadata": { "id": "JEJHhN65a2VV" }, "outputs": [ { "data": { "text/plain": [ "MPG 0\n", "Cylinders 0\n", "Displacement 0\n", "Horsepower 6\n", "Weight 0\n", "Acceleration 0\n", "Model Year 0\n", "Origin 0\n", "dtype: int64" ] }, "execution_count": 7, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dataset.isna().sum()" ] }, { "cell_type": "markdown", "metadata": { "id": "9UPN0KBHa_WI" }, "source": [ "Drop those rows to keep this initial tutorial simple:\n" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "id": "4ZUDosChC1UN" }, "outputs": [], "source": [ "dataset = dataset.dropna()" ] }, { "cell_type": "markdown", "metadata": { "id": "8XKitwaH4v8h" }, "source": [ "The `\"Origin\"` column is categorical, not numeric. So the next step is to one-hot encode the values in the column with [pd.get_dummies](https://pandas.pydata.org/docs/reference/api/pandas.get_dummies.html).\n", "\n", "Note: You can set up the `tf.keras.Model` to do this kind of transformation for you but that's beyond the scope of this tutorial. Check out the [Classify structured data using Keras preprocessing layers](../structured_data/preprocessing_layers.ipynb) or [Load CSV data](../load_data/csv.ipynb) tutorials for examples.\n" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "id": "gWNTD2QjBWFJ" }, "outputs": [], "source": [ "dataset['Origin'] = dataset['Origin'].map({1: 'USA', 2: 'Europe', 3: 'Japan'})" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "id": "ulXz4J7PAUzk" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
MPGCylindersDisplacementHorsepowerWeightAccelerationModel YearEuropeJapanUSA
39327.04140.086.02790.015.682FalseFalseTrue
39444.0497.052.02130.024.682TrueFalseFalse
39532.04135.084.02295.011.682FalseFalseTrue
39628.04120.079.02625.018.682FalseFalseTrue
39731.04119.082.02720.019.482FalseFalseTrue
\n", "
" ], "text/plain": [ " MPG Cylinders Displacement Horsepower Weight Acceleration \\\n", "393 27.0 4 140.0 86.0 2790.0 15.6 \n", "394 44.0 4 97.0 52.0 2130.0 24.6 \n", "395 32.0 4 135.0 84.0 2295.0 11.6 \n", "396 28.0 4 120.0 79.0 2625.0 18.6 \n", "397 31.0 4 119.0 82.0 2720.0 19.4 \n", "\n", " Model Year Europe Japan USA \n", "393 82 False False True \n", "394 82 True False False \n", "395 82 False False True \n", "396 82 False False True \n", "397 82 False False True " ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "dataset = pd.get_dummies(dataset, columns=['Origin'], prefix='', prefix_sep='')\n", "dataset.tail()" ] }, { "cell_type": "markdown", "metadata": { "id": "Cuym4yvk76vU" }, "source": [ "### Split the data into training and test sets\n", "\n", "Now, split the dataset into a training set and a test set. You will use the test set in the final evaluation of your models.\n" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "id": "qn-IGhUE7_1H" }, "outputs": [], "source": [ "train_dataset = dataset.sample(frac=0.8, random_state=0)\n", "test_dataset = dataset.drop(train_dataset.index)" ] }, { "cell_type": "markdown", "metadata": { "id": "J4ubs136WLNp" }, "source": [ "### Inspect the data\n", "\n", "Review the joint distribution of a few pairs of columns from the training set.\n", "\n", "The top row suggests that the fuel efficiency (MPG) is a function of all the other parameters. The other rows indicate they are functions of each other.\n" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "id": "oRKO_x8gWKv-" }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "sns.pairplot(\n", " train_dataset[['MPG', 'Cylinders', 'Displacement', 'Weight']], diag_kind='kde')" ] }, { "cell_type": "markdown", "metadata": { "id": "gavKO_6DWRMP" }, "source": [ "Let's also check the overall statistics. Note how each feature covers a very different range:\n" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "id": "yi2FzC3T21jR" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
countmeanstdmin25%50%75%max
MPG314.023.3105107.72865210.017.0022.028.9546.6
Cylinders314.05.4777071.6997883.04.004.08.008.0
Displacement314.0195.318471104.33158968.0105.50151.0265.75455.0
Horsepower314.0104.86942738.09621446.076.2594.5128.00225.0
Weight314.02990.251592843.8985961649.02256.502822.53608.005140.0
Acceleration314.015.5592362.7892308.013.8015.517.2024.8
Model Year314.075.8980893.67564270.073.0076.079.0082.0
\n", "
" ], "text/plain": [ " count mean std min 25% 50% \\\n", "MPG 314.0 23.310510 7.728652 10.0 17.00 22.0 \n", "Cylinders 314.0 5.477707 1.699788 3.0 4.00 4.0 \n", "Displacement 314.0 195.318471 104.331589 68.0 105.50 151.0 \n", "Horsepower 314.0 104.869427 38.096214 46.0 76.25 94.5 \n", "Weight 314.0 2990.251592 843.898596 1649.0 2256.50 2822.5 \n", "Acceleration 314.0 15.559236 2.789230 8.0 13.80 15.5 \n", "Model Year 314.0 75.898089 3.675642 70.0 73.00 76.0 \n", "\n", " 75% max \n", "MPG 28.95 46.6 \n", "Cylinders 8.00 8.0 \n", "Displacement 265.75 455.0 \n", "Horsepower 128.00 225.0 \n", "Weight 3608.00 5140.0 \n", "Acceleration 17.20 24.8 \n", "Model Year 79.00 82.0 " ] }, "execution_count": 13, "metadata": {}, "output_type": "execute_result" } ], "source": [ "train_dataset.describe().transpose()" ] }, { "cell_type": "markdown", "metadata": { "id": "Db7Auq1yXUvh" }, "source": [ "### Split features from labels\n", "\n", "Separate the target value—the \"label\"—from the features. This label is the value that you will train the model to predict.\n" ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "id": "t2sluJdCW7jN" }, "outputs": [], "source": [ "train_features = train_dataset.copy()\n", "test_features = test_dataset.copy()\n", "\n", "train_labels = train_features.pop('MPG')\n", "test_labels = test_features.pop('MPG')" ] }, { "cell_type": "markdown", "metadata": { "id": "mRklxK5s388r" }, "source": [ "## Normalization\n", "\n", "In the table of statistics it's easy to see how different the ranges of each feature are:\n" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "id": "IcmY6lKKbkw8" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meanstd
MPG23.3105107.728652
Cylinders5.4777071.699788
Displacement195.318471104.331589
Horsepower104.86942738.096214
Weight2990.251592843.898596
Acceleration15.5592362.789230
Model Year75.8980893.675642
\n", "
" ], "text/plain": [ " mean std\n", "MPG 23.310510 7.728652\n", "Cylinders 5.477707 1.699788\n", "Displacement 195.318471 104.331589\n", "Horsepower 104.869427 38.096214\n", "Weight 2990.251592 843.898596\n", "Acceleration 15.559236 2.789230\n", "Model Year 75.898089 3.675642" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "train_dataset.describe().transpose()[['mean', 'std']]" ] }, { "cell_type": "markdown", "metadata": { "id": "-ywmerQ6dSox" }, "source": [ "It is good practice to normalize features that use different scales and ranges.\n", "\n", "One reason this is important is because the features are multiplied by the model weights. So, the scale of the outputs and the scale of the gradients are affected by the scale of the inputs.\n", "\n", "Although a model _might_ converge without feature normalization, normalization makes training much more stable.\n", "\n", "Note: There is no advantage to normalizing the one-hot features—it is done here for simplicity. For more details on how to use the preprocessing layers, refer to the [Working with preprocessing layers](https://www.tensorflow.org/guide/keras/preprocessing_layers) guide and the [Classify structured data using Keras preprocessing layers](../structured_data/preprocessing_layers.ipynb) tutorial.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "aFJ6ISropeoo" }, "source": [ "### The Normalization layer\n", "\n", "The `tf.keras.layers.Normalization` is a clean and simple way to add feature normalization into your model.\n", "\n", "The first step is to create the layer:\n" ] }, { "cell_type": "code", "execution_count": 16, "metadata": { "id": "JlC5ooJrgjQF" }, "outputs": [], "source": [ "normalizer = tf.keras.layers.Normalization(axis=-1)" ] }, { "cell_type": "markdown", "metadata": { "id": "XYA2Ap6nVOha" }, "source": [ "Then, fit the state of the preprocessing layer to the data by calling `Normalization.adapt`:\n" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "id": "CrBbbjbwV91f" }, "outputs": [], "source": [ "normalizer.adapt(np.array(train_features))" ] }, { "cell_type": "markdown", "metadata": { "id": "oZccMR5yV9YV" }, "source": [ "Calculate the mean and variance, and store them in the layer:\n" ] }, { "cell_type": "code", "execution_count": 18, "metadata": { "id": "GGn-ukwxSPtx" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "[[ 5.478 195.318 104.869 2990.252 15.559 75.898 0.178 0.197\n", " 0.624]]\n" ] } ], "source": [ "print(normalizer.mean.numpy())" ] }, { "cell_type": "markdown", "metadata": { "id": "oGWKaF9GSRuN" }, "source": [ "When the layer is called, it returns the input data, with each feature independently normalized:\n" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "id": "2l7zFL_XWIRu" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "First example: [[4 90.0 75.0 2125.0 14.5 74 False False True]]\n", "\n", "Normalized: [[-0.87 -1.01 -0.79 -1.03 -0.38 -0.52 -0.47 -0.5 0.78]]\n" ] } ], "source": [ "first = np.array(train_features[:1])\n", "\n", "with np.printoptions(precision=2, suppress=True):\n", " print('First example:', first)\n", " print()\n", " print('Normalized:', normalizer(first).numpy())" ] }, { "cell_type": "markdown", "metadata": { "id": "6o3CrycBXA2s" }, "source": [ "## Linear regression\n", "\n", "Before building a deep neural network model, start with linear regression using one and several variables.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "lFby9n0tnHkw" }, "source": [ "### Linear regression with one variable\n", "\n", "Begin with a single-variable linear regression to predict `'MPG'` from `'Horsepower'`.\n", "\n", "Training a model with `tf.keras` typically starts by defining the model architecture. Use a `tf.keras.Sequential` model, which [represents a sequence of steps](https://www.tensorflow.org/guide/keras/sequential_model).\n", "\n", "There are two steps in your single-variable linear regression model:\n", "\n", "- Normalize the `'Horsepower'` input features using the `tf.keras.layers.Normalization` preprocessing layer.\n", "- Apply a linear transformation ($y = mx+b$) to produce 1 output using a linear layer (`tf.keras.layers.Dense`).\n", "\n", "The number of _inputs_ can either be set by the `input_shape` argument, or automatically when the model is run for the first time.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "Xp3gAFn3TPv8" }, "source": [ "First, create a NumPy array made of the `'Horsepower'` features. Then, instantiate the `tf.keras.layers.Normalization` and fit its state to the `horsepower` data:\n" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "id": "1gJAy0fKs1TS" }, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/Users/diegomaradona/miniforge3/envs/formula-1/lib/python3.11/site-packages/keras/src/layers/preprocessing/normalization.py:99: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", " super().__init__(**kwargs)\n" ] } ], "source": [ "horsepower = np.array(train_features['Horsepower'])\n", "\n", "horsepower_normalizer = layers.Normalization(input_shape=[1,], axis=None)\n", "horsepower_normalizer.adapt(horsepower)" ] }, { "cell_type": "markdown", "metadata": { "id": "4NVlHJY2TWlC" }, "source": [ "Build the Keras Sequential model:\n" ] }, { "cell_type": "code", "execution_count": 21, "metadata": { "id": "c0sXM7qLlKfZ" }, "outputs": [ { "data": { "text/html": [ "
Model: \"sequential\"\n",
       "
\n" ], "text/plain": [ "\u001b[1mModel: \"sequential\"\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃ Layer (type)                    ┃ Output Shape           ┃       Param # ┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ normalization_1 (Normalization) │ (None, 1)              │             3 │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense (Dense)                   │ (None, 1)              │             2 │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "
\n" ], "text/plain": [ "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n", "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n", "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n", "│ normalization_1 (\u001b[38;5;33mNormalization\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m3\u001b[0m │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m2\u001b[0m │\n", "└─────────────────────────────────┴────────────────────────┴───────────────┘\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Total params: 5 (24.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m5\u001b[0m (24.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Trainable params: 2 (8.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m2\u001b[0m (8.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Non-trainable params: 3 (16.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m3\u001b[0m (16.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "horsepower_model = tf.keras.Sequential([\n", " horsepower_normalizer,\n", " layers.Dense(units=1)\n", "])\n", "\n", "horsepower_model.summary()" ] }, { "cell_type": "markdown", "metadata": { "id": "eObQu9fDnXGL" }, "source": [ "This model will predict `'MPG'` from `'Horsepower'`.\n", "\n", "Run the untrained model on the first 10 'Horsepower' values. The output won't be good, but notice that it has the expected shape of `(10, 1)`:\n" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "id": "UfV1HS6bns-s" }, "outputs": [], "source": [ "# horsepower_model.predict(horsepower[:10])" ] }, { "cell_type": "markdown", "metadata": { "id": "CSkanJlmmFBX" }, "source": [ "Once the model is built, configure the training procedure using the Keras `Model.compile` method. The most important arguments to compile are the `loss` and the `optimizer`, since these define what will be optimized (`mean_absolute_error`) and how (using the `tf.keras.optimizers.Adam`).\n" ] }, { "cell_type": "code", "execution_count": 23, "metadata": { "id": "JxA_3lpOm-SK" }, "outputs": [], "source": [ "horsepower_model.compile(\n", " optimizer=tf.keras.optimizers.Adam(learning_rate=0.1),\n", " loss='mean_absolute_error')" ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[PhysicalDevice(name='/physical_device:CPU:0', device_type='CPU')]" ] }, "execution_count": 24, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import tensorflow as tf\n", "tf.config.list_physical_devices()" ] }, { "cell_type": "markdown", "metadata": { "id": "Z3q1I9TwnRSC" }, "source": [ "Use Keras `Model.fit` to execute the training for 100 epochs:\n" ] }, { "cell_type": "code", "execution_count": 25, "metadata": { "id": "-iSrNy59nRAp" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 57ms/step - loss: 23.0730 - val_loss: 22.9327\n", "Epoch 2/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 22.5346 - val_loss: 22.1325\n", "Epoch 3/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 21.4295 - val_loss: 21.3325\n", "Epoch 4/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 20.6460 - val_loss: 20.5327\n", "Epoch 5/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 19.8564 - val_loss: 19.7326\n", "Epoch 6/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 18.8528 - val_loss: 18.9328\n", "Epoch 7/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 18.2561 - val_loss: 18.1326\n", "Epoch 8/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 17.7468 - val_loss: 17.3326\n", "Epoch 9/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 16.6506 - val_loss: 16.5324\n", "Epoch 10/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 15.8016 - val_loss: 15.7326\n", "Epoch 11/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 14.7536 - val_loss: 14.9328\n", "Epoch 12/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 14.2862 - val_loss: 14.1328\n", "Epoch 13/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 13.8565 - val_loss: 13.3326\n", "Epoch 14/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 12.9746 - val_loss: 12.5326\n", "Epoch 15/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 11.4509 - val_loss: 11.7336\n", "Epoch 16/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 10.7102 - val_loss: 10.9477\n", "Epoch 17/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 10.0547 - val_loss: 10.1759\n", "Epoch 18/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 9.5363 - val_loss: 9.3999\n", "Epoch 19/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 8.1100 - val_loss: 8.6601\n", "Epoch 20/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 7.9764 - val_loss: 7.9265\n", "Epoch 21/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 7.3117 - val_loss: 7.2149\n", "Epoch 22/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 6.6744 - val_loss: 6.5776\n", "Epoch 23/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 6.0048 - val_loss: 6.0127\n", "Epoch 24/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.3425 - val_loss: 5.5844\n", "Epoch 25/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 5.1510 - val_loss: 5.2445\n", "Epoch 26/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.8535 - val_loss: 5.0415\n", "Epoch 27/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.7903 - val_loss: 4.8585\n", "Epoch 28/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8860 - val_loss: 4.6968\n", "Epoch 29/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.2123 - val_loss: 4.5549\n", "Epoch 30/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.2328 - val_loss: 4.4754\n", "Epoch 31/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9166 - val_loss: 4.3810\n", "Epoch 32/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.7030 - val_loss: 4.3217\n", "Epoch 33/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8560 - val_loss: 4.2818\n", "Epoch 34/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.1715 - val_loss: 4.2335\n", "Epoch 35/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0006 - val_loss: 4.2207\n", "Epoch 36/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.8899 - val_loss: 4.2311\n", "Epoch 37/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.8019 - val_loss: 4.2415\n", "Epoch 38/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8169 - val_loss: 4.2166\n", "Epoch 39/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 12ms/step - loss: 3.6327 - val_loss: 4.2004\n", "Epoch 40/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7577 - val_loss: 4.2239\n", "Epoch 41/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7890 - val_loss: 4.2036\n", "Epoch 42/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.6212 - val_loss: 4.1861\n", "Epoch 43/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.9229 - val_loss: 4.1913\n", "Epoch 44/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 4.0109 - val_loss: 4.1976\n", "Epoch 45/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.7400 - val_loss: 4.2034\n", "Epoch 46/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8781 - val_loss: 4.2080\n", "Epoch 47/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.6624 - val_loss: 4.1974\n", "Epoch 48/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.0419 - val_loss: 4.1811\n", "Epoch 49/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9899 - val_loss: 4.1871\n", "Epoch 50/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9428 - val_loss: 4.1958\n", "Epoch 51/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7333 - val_loss: 4.2022\n", "Epoch 52/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.8142 - val_loss: 4.1805\n", "Epoch 53/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.8327 - val_loss: 4.1987\n", "Epoch 54/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.7288 - val_loss: 4.1919\n", "Epoch 55/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.5987 - val_loss: 4.1899\n", "Epoch 56/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step - loss: 3.9085 - val_loss: 4.1790\n", "Epoch 57/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.8779 - val_loss: 4.1617\n", "Epoch 58/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.7819 - val_loss: 4.1960\n", "Epoch 59/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.6504 - val_loss: 4.2081\n", "Epoch 60/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 11ms/step - loss: 3.7937 - val_loss: 4.1800\n", "Epoch 61/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.9108 - val_loss: 4.1779\n", "Epoch 62/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8943 - val_loss: 4.1962\n", "Epoch 63/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.7918 - val_loss: 4.1904\n", "Epoch 64/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9330 - val_loss: 4.1968\n", "Epoch 65/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6947 - val_loss: 4.1918\n", "Epoch 66/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8556 - val_loss: 4.1864\n", "Epoch 67/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9801 - val_loss: 4.1951\n", "Epoch 68/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7387 - val_loss: 4.1925\n", "Epoch 69/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9845 - val_loss: 4.1846\n", "Epoch 70/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.6867 - val_loss: 4.2075\n", "Epoch 71/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8190 - val_loss: 4.1868\n", "Epoch 72/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 4.1884 - val_loss: 4.1955\n", "Epoch 73/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8159 - val_loss: 4.2095\n", "Epoch 74/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9562 - val_loss: 4.1710\n", "Epoch 75/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.6341 - val_loss: 4.1875\n", "Epoch 76/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 45ms/step - loss: 3.8470 - val_loss: 4.1941\n", "Epoch 77/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.7104 - val_loss: 4.2076\n", "Epoch 78/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.7653 - val_loss: 4.1907\n", "Epoch 79/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8820 - val_loss: 4.1834\n", "Epoch 80/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 4.0157 - val_loss: 4.1614\n", "Epoch 81/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7353 - val_loss: 4.1801\n", "Epoch 82/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.9267 - val_loss: 4.1826\n", "Epoch 83/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.7905 - val_loss: 4.2104\n", "Epoch 84/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8504 - val_loss: 4.2180\n", "Epoch 85/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.8914 - val_loss: 4.1662\n", "Epoch 86/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.8619 - val_loss: 4.1526\n", "Epoch 87/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.8183 - val_loss: 4.1902\n", "Epoch 88/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7469 - val_loss: 4.1966\n", "Epoch 89/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8581 - val_loss: 4.1926\n", "Epoch 90/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8881 - val_loss: 4.1802\n", "Epoch 91/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.9785 - val_loss: 4.1634\n", "Epoch 92/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.8815 - val_loss: 4.1949\n", "Epoch 93/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8258 - val_loss: 4.2060\n", "Epoch 94/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.4390 - val_loss: 4.1952\n", "Epoch 95/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7804 - val_loss: 4.1843\n", "Epoch 96/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.6523 - val_loss: 4.1944\n", "Epoch 97/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - loss: 3.7563 - val_loss: 4.1825\n", "Epoch 98/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 7ms/step - loss: 3.9667 - val_loss: 4.1760\n", "Epoch 99/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 34ms/step - loss: 3.7305 - val_loss: 4.1878\n", "Epoch 100/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step - loss: 3.6647 - val_loss: 4.1957\n", "CPU times: user 8.09 s, sys: 884 ms, total: 8.98 s\n", "Wall time: 10.6 s\n" ] } ], "source": [ "%%time\n", "history = horsepower_model.fit(\n", " train_features['Horsepower'],\n", " train_labels,\n", " epochs=100,\n", " # Suppress logging.\n", " verbose=1,\n", " # Calculate validation results on 20% of the training data.\n", " validation_split = 0.2)" ] }, { "cell_type": "markdown", "metadata": { "id": "tQm3pc0FYPQB" }, "source": [ "Visualize the model's training progress using the stats stored in the `history` object:\n" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "id": "YCAwD_y4AdC3" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
lossval_lossepoch
953.8066424.19438795
963.8046034.18254796
973.8036874.17596597
983.8041384.18775398
993.8021884.19569899
\n", "
" ], "text/plain": [ " loss val_loss epoch\n", "95 3.806642 4.194387 95\n", "96 3.804603 4.182547 96\n", "97 3.803687 4.175965 97\n", "98 3.804138 4.187753 98\n", "99 3.802188 4.195698 99" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "hist = pd.DataFrame(history.history)\n", "hist['epoch'] = history.epoch\n", "hist.tail()" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "id": "9E54UoZunqhc" }, "outputs": [], "source": [ "def plot_loss(history):\n", " plt.plot(history.history['loss'], label='loss')\n", " plt.plot(history.history['val_loss'], label='val_loss')\n", " plt.ylim([0, 10])\n", " plt.xlabel('Epoch')\n", " plt.ylabel('Error [MPG]')\n", " plt.legend()\n", " plt.grid(True)" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "id": "yYsQYrIZyqjz" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_loss(history)" ] }, { "cell_type": "markdown", "metadata": { "id": "CMNrt8X2ebXd" }, "source": [ "Collect the results on the test set for later:\n" ] }, { "cell_type": "code", "execution_count": 29, "metadata": { "id": "kDZ8EvNYrDtx" }, "outputs": [], "source": [ "test_results = {}\n", "\n", "test_results['horsepower_model'] = horsepower_model.evaluate(\n", " test_features['Horsepower'],\n", " test_labels, verbose=0)" ] }, { "cell_type": "markdown", "metadata": { "id": "F0qutYAKwoda" }, "source": [ "Since this is a single variable regression, it's easy to view the model's predictions as a function of the input:\n" ] }, { "cell_type": "code", "execution_count": 30, "metadata": { "id": "xDS2JEtOn9Jn" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 8ms/step\n" ] } ], "source": [ "x = tf.linspace(0.0, 250, 251)\n", "y = horsepower_model.predict(x)" ] }, { "cell_type": "code", "execution_count": 31, "metadata": { "id": "rttFCTU8czsI" }, "outputs": [], "source": [ "def plot_horsepower(x, y):\n", " plt.scatter(train_features['Horsepower'], train_labels, label='Data')\n", " plt.plot(x, y, color='k', label='Predictions')\n", " plt.xlabel('Horsepower')\n", " plt.ylabel('MPG')\n", " plt.legend()" ] }, { "cell_type": "code", "execution_count": 32, "metadata": { "id": "7l9ZiAOEUNBL" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_horsepower(x, y)" ] }, { "cell_type": "markdown", "metadata": { "id": "Yk2RmlqPoM9u" }, "source": [ "### Linear regression with multiple inputs\n" ] }, { "cell_type": "markdown", "metadata": { "id": "PribnwDHUksC" }, "source": [ "You can use an almost identical setup to make predictions based on multiple inputs. This model still does the same $y = mx+b$ except that $m$ is a matrix and $x$ is a vector.\n", "\n", "Create a two-step Keras Sequential model again with the first layer being `normalizer` (`tf.keras.layers.Normalization(axis=-1)`) you defined earlier and adapted to the whole dataset:\n" ] }, { "cell_type": "code", "execution_count": 33, "metadata": { "id": "ssnVcKg7oMe6" }, "outputs": [], "source": [ "linear_model = tf.keras.Sequential([\n", " normalizer,\n", " layers.Dense(units=1)\n", "])" ] }, { "cell_type": "markdown", "metadata": { "id": "IHlx6WeIWyAr" }, "source": [ "When you call `Model.predict` on a batch of inputs, it produces `units=1` outputs for each example:\n" ] }, { "cell_type": "code", "execution_count": 34, "metadata": { "id": "DynfJV18WiuT" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 190ms/step\n" ] }, { "data": { "text/plain": [ "array([[ 1.912],\n", " [-0.136],\n", " [-0.748],\n", " [ 0.861],\n", " [ 0.664],\n", " [-0.449],\n", " [ 0.667],\n", " [-2.866],\n", " [ 0.24 ],\n", " [ 1.051]], dtype=float32)" ] }, "execution_count": 34, "metadata": {}, "output_type": "execute_result" } ], "source": [ "linear_model.predict(train_features[:10])" ] }, { "cell_type": "markdown", "metadata": { "id": "hvHKH3rPXHmq" }, "source": [ "When you call the model, its weight matrices will be built—check that the `kernel` weights (the $m$ in $y=mx+b$) have a shape of `(9, 1)`:\n" ] }, { "cell_type": "code", "execution_count": 35, "metadata": { "id": "DwJ4Fq0RXBQf" }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 35, "metadata": {}, "output_type": "execute_result" } ], "source": [ "linear_model.layers[1].kernel" ] }, { "cell_type": "markdown", "metadata": { "id": "eINAc6rZXzOt" }, "source": [ "Configure the model with Keras `Model.compile` and train with `Model.fit` for 100 epochs:\n" ] }, { "cell_type": "code", "execution_count": 36, "metadata": { "id": "A0Sv_Ybr0szp" }, "outputs": [], "source": [ "linear_model.compile(\n", " optimizer=tf.keras.optimizers.Adam(learning_rate=0.1),\n", " loss='mean_absolute_error')" ] }, { "cell_type": "code", "execution_count": 37, "metadata": { "id": "EZoOYORvoTSe" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "CPU times: user 4.32 s, sys: 413 ms, total: 4.73 s\n", "Wall time: 4.32 s\n" ] } ], "source": [ "%%time\n", "history = linear_model.fit(\n", " train_features,\n", " train_labels,\n", " epochs=100,\n", " # Suppress logging.\n", " verbose=0,\n", " # Calculate validation results on 20% of the training data.\n", " validation_split = 0.2)" ] }, { "cell_type": "markdown", "metadata": { "id": "EdxiCbiNYK2F" }, "source": [ "Using all the inputs in this regression model achieves a much lower training and validation error than the `horsepower_model`, which had one input:\n" ] }, { "cell_type": "code", "execution_count": 38, "metadata": { "id": "4sWO3W0koYgu" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_loss(history)" ] }, { "cell_type": "markdown", "metadata": { "id": "NyN49hIWe_NH" }, "source": [ "Collect the results on the test set for later:\n" ] }, { "cell_type": "code", "execution_count": 39, "metadata": { "id": "jNC3D1DGsGgK" }, "outputs": [], "source": [ "test_results['linear_model'] = linear_model.evaluate(\n", " test_features, test_labels, verbose=0)" ] }, { "cell_type": "markdown", "metadata": { "id": "SmjdzxKzEu1-" }, "source": [ "## Regression with a deep neural network (DNN)\n" ] }, { "cell_type": "markdown", "metadata": { "id": "DT_aHPsrzO1t" }, "source": [ "In the previous section, you implemented two linear models for single and multiple inputs.\n", "\n", "Here, you will implement single-input and multiple-input DNN models.\n", "\n", "The code is basically the same except the model is expanded to include some \"hidden\" non-linear layers. The name \"hidden\" here just means not directly connected to the inputs or outputs.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "6SWtkIjhrZwa" }, "source": [ "These models will contain a few more layers than the linear model:\n", "\n", "- The normalization layer, as before (with `horsepower_normalizer` for a single-input model and `normalizer` for a multiple-input model).\n", "- Two hidden, non-linear, `Dense` layers with the ReLU (`relu`) activation function nonlinearity.\n", "- A linear `Dense` single-output layer.\n", "\n", "Both models will use the same training procedure, so the `compile` method is included in the `build_and_compile_model` function below.\n" ] }, { "cell_type": "code", "execution_count": 40, "metadata": { "id": "c26juK7ZG8j-" }, "outputs": [], "source": [ "def build_and_compile_model(norm):\n", " model = keras.Sequential([\n", " norm,\n", " layers.Dense(64, activation='relu'),\n", " layers.Dense(64, activation='relu'),\n", " layers.Dense(1)\n", " ])\n", "\n", " model.compile(loss='mean_absolute_error',\n", " optimizer=tf.keras.optimizers.Adam(0.001))\n", " return model" ] }, { "cell_type": "markdown", "metadata": { "id": "6c51caebbc0d" }, "source": [ "### Regression using a DNN and a single input\n" ] }, { "cell_type": "markdown", "metadata": { "id": "xvu9gtxTZR5V" }, "source": [ "Create a DNN model with only `'Horsepower'` as input and `horsepower_normalizer` (defined earlier) as the normalization layer:\n" ] }, { "cell_type": "code", "execution_count": 41, "metadata": { "id": "cGbPb-PHGbhs" }, "outputs": [], "source": [ "dnn_horsepower_model = build_and_compile_model(horsepower_normalizer)" ] }, { "cell_type": "markdown", "metadata": { "id": "Sj49Og4YGULr" }, "source": [ "This model has quite a few more trainable parameters than the linear models:\n" ] }, { "cell_type": "code", "execution_count": 42, "metadata": { "id": "ReAD0n6MsFK-" }, "outputs": [ { "data": { "text/html": [ "
Model: \"sequential_2\"\n",
       "
\n" ], "text/plain": [ "\u001b[1mModel: \"sequential_2\"\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃ Layer (type)                    ┃ Output Shape           ┃       Param # ┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ normalization_1 (Normalization) │ (None, 1)              │             3 │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_2 (Dense)                 │ (None, 64)             │           128 │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_3 (Dense)                 │ (None, 64)             │         4,160 │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_4 (Dense)                 │ (None, 1)              │            65 │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "
\n" ], "text/plain": [ "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n", "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n", "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n", "│ normalization_1 (\u001b[38;5;33mNormalization\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m3\u001b[0m │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_2 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m128\u001b[0m │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_3 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m64\u001b[0m) │ \u001b[38;5;34m4,160\u001b[0m │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_4 (\u001b[38;5;33mDense\u001b[0m) │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m) │ \u001b[38;5;34m65\u001b[0m │\n", "└─────────────────────────────────┴────────────────────────┴───────────────┘\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Total params: 4,356 (17.02 KB)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m4,356\u001b[0m (17.02 KB)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Trainable params: 4,353 (17.00 KB)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m4,353\u001b[0m (17.00 KB)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Non-trainable params: 3 (16.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m3\u001b[0m (16.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dnn_horsepower_model.summary()" ] }, { "cell_type": "markdown", "metadata": { "id": "0-qWCsh6DlyH" }, "source": [ "Train the model with Keras `Model.fit`:\n" ] }, { "cell_type": "code", "execution_count": 43, "metadata": { "id": "sD7qHCmNIOY0" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Epoch 1/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 16ms/step - loss: 23.3406 - val_loss: 23.5376\n", "Epoch 2/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 22.6237 - val_loss: 23.2827\n", "Epoch 3/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 22.3482 - val_loss: 23.0177\n", "Epoch 4/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 22.6963 - val_loss: 22.6979\n", "Epoch 5/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 22.4835 - val_loss: 22.2934\n", "Epoch 6/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 20.8064 - val_loss: 21.7769\n", "Epoch 7/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 21.3272 - val_loss: 21.1064\n", "Epoch 8/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 20.5179 - val_loss: 20.2638\n", "Epoch 9/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 20.1822 - val_loss: 19.2042\n", "Epoch 10/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 18.8464 - val_loss: 17.8975\n", "Epoch 11/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 16.7458 - val_loss: 16.3702\n", "Epoch 12/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 15.6274 - val_loss: 14.6095\n", "Epoch 13/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 14.1242 - val_loss: 12.7258\n", "Epoch 14/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 12.4111 - val_loss: 10.7173\n", "Epoch 15/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 9.9853 - val_loss: 8.6982\n", "Epoch 16/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 7.2959 - val_loss: 6.7763\n", "Epoch 17/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 5.5447 - val_loss: 5.4299\n", "Epoch 18/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 4.3193 - val_loss: 5.4205\n", "Epoch 19/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.9461 - val_loss: 5.3927\n", "Epoch 20/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 4.4559 - val_loss: 5.1505\n", "Epoch 21/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 4.0093 - val_loss: 4.9382\n", "Epoch 22/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.9185 - val_loss: 4.8527\n", "Epoch 23/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.9001 - val_loss: 4.8145\n", "Epoch 24/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.1636 - val_loss: 4.7823\n", "Epoch 25/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.9694 - val_loss: 4.6969\n", "Epoch 26/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 4.0743 - val_loss: 4.6290\n", "Epoch 27/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8491 - val_loss: 4.5746\n", "Epoch 28/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.6691 - val_loss: 4.5141\n", "Epoch 29/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.4493 - val_loss: 4.4631\n", "Epoch 30/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.7181 - val_loss: 4.4172\n", "Epoch 31/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.6709 - val_loss: 4.3712\n", "Epoch 32/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.7494 - val_loss: 4.3313\n", "Epoch 33/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 27ms/step - loss: 3.4456 - val_loss: 4.2959\n", "Epoch 34/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 9ms/step - loss: 3.4759 - val_loss: 4.2737\n", "Epoch 35/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.5438 - val_loss: 4.2423\n", "Epoch 36/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.4742 - val_loss: 4.2203\n", "Epoch 37/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.3549 - val_loss: 4.2060\n", "Epoch 38/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.8108 - val_loss: 4.1825\n", "Epoch 39/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4052 - val_loss: 4.1514\n", "Epoch 40/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.5751 - val_loss: 4.1276\n", "Epoch 41/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3500 - val_loss: 4.1026\n", "Epoch 42/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4536 - val_loss: 4.0852\n", "Epoch 43/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4466 - val_loss: 4.0595\n", "Epoch 44/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.5157 - val_loss: 4.0430\n", "Epoch 45/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4757 - val_loss: 4.0223\n", "Epoch 46/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.5406 - val_loss: 4.0039\n", "Epoch 47/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2980 - val_loss: 3.9903\n", "Epoch 48/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4312 - val_loss: 3.9700\n", "Epoch 49/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4820 - val_loss: 3.9547\n", "Epoch 50/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4456 - val_loss: 3.9428\n", "Epoch 51/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3002 - val_loss: 3.9257\n", "Epoch 52/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.7905 - val_loss: 3.9159\n", "Epoch 53/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4443 - val_loss: 3.9034\n", "Epoch 54/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.5159 - val_loss: 3.8950\n", "Epoch 55/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3604 - val_loss: 3.8823\n", "Epoch 56/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2979 - val_loss: 3.8736\n", "Epoch 57/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3905 - val_loss: 3.8652\n", "Epoch 58/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1933 - val_loss: 3.8555\n", "Epoch 59/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.2922 - val_loss: 3.8466\n", "Epoch 60/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4644 - val_loss: 3.8358\n", "Epoch 61/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.6650 - val_loss: 3.8254\n", "Epoch 62/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3851 - val_loss: 3.8231\n", "Epoch 63/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3929 - val_loss: 3.8137\n", "Epoch 64/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3725 - val_loss: 3.8117\n", "Epoch 65/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3882 - val_loss: 3.8005\n", "Epoch 66/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2898 - val_loss: 3.7951\n", "Epoch 67/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3415 - val_loss: 3.7916\n", "Epoch 68/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.6329 - val_loss: 3.7866\n", "Epoch 69/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 2.9822 - val_loss: 3.7831\n", "Epoch 70/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2633 - val_loss: 3.7888\n", "Epoch 71/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1712 - val_loss: 3.7781\n", "Epoch 72/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3555 - val_loss: 3.7804\n", "Epoch 73/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.3267 - val_loss: 3.7767\n", "Epoch 74/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.0617 - val_loss: 3.8089\n", "Epoch 75/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3320 - val_loss: 3.7788\n", "Epoch 76/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3989 - val_loss: 3.7714\n", "Epoch 77/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2895 - val_loss: 3.7742\n", "Epoch 78/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.6588 - val_loss: 3.8086\n", "Epoch 79/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.1576 - val_loss: 3.7774\n", "Epoch 80/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1521 - val_loss: 3.7812\n", "Epoch 81/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3732 - val_loss: 3.7715\n", "Epoch 82/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4121 - val_loss: 3.7625\n", "Epoch 83/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.3590 - val_loss: 3.7798\n", "Epoch 84/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2066 - val_loss: 3.7660\n", "Epoch 85/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2276 - val_loss: 3.7804\n", "Epoch 86/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.0587 - val_loss: 3.7714\n", "Epoch 87/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1696 - val_loss: 3.7900\n", "Epoch 88/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 6ms/step - loss: 3.2637 - val_loss: 3.7748\n", "Epoch 89/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2026 - val_loss: 3.7755\n", "Epoch 90/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step - loss: 3.2265 - val_loss: 3.7859\n", "Epoch 91/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1260 - val_loss: 3.7714\n", "Epoch 92/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4126 - val_loss: 3.7771\n", "Epoch 93/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2984 - val_loss: 3.7717\n", "Epoch 94/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4742 - val_loss: 3.7874\n", "Epoch 95/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1427 - val_loss: 3.7836\n", "Epoch 96/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2167 - val_loss: 3.7601\n", "Epoch 97/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.2843 - val_loss: 3.7644\n", "Epoch 98/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1456 - val_loss: 3.7879\n", "Epoch 99/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.4607 - val_loss: 3.7821\n", "Epoch 100/100\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 4ms/step - loss: 3.1820 - val_loss: 3.7745\n", "CPU times: user 5.96 s, sys: 709 ms, total: 6.67 s\n", "Wall time: 6.5 s\n" ] } ], "source": [ "%%time\n", "history = dnn_horsepower_model.fit(\n", " train_features['Horsepower'],\n", " train_labels,\n", " validation_split=0.2,\n", " verbose=1, epochs=100)" ] }, { "cell_type": "markdown", "metadata": { "id": "dArGGxHxcKjN" }, "source": [ "This model does slightly better than the linear single-input `horsepower_model`:\n" ] }, { "cell_type": "code", "execution_count": 44, "metadata": { "id": "NcF6UWjdCU8T" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_loss(history)" ] }, { "cell_type": "markdown", "metadata": { "id": "TG1snlpR2QCK" }, "source": [ "If you plot the predictions as a function of `'Horsepower'`, you should notice how this model takes advantage of the nonlinearity provided by the hidden layers:\n" ] }, { "cell_type": "code", "execution_count": 45, "metadata": { "id": "hPF53Rem14NS" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\u001b[1m1/8\u001b[0m \u001b[32m━━\u001b[0m\u001b[37m━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[1m0s\u001b[0m 46ms/stepWARNING:tensorflow:5 out of the last 17 calls to .one_step_on_data_distributed at 0x30c0f84a0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "\u001b[1m8/8\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 5ms/step \n" ] } ], "source": [ "x = tf.linspace(0.0, 250, 251)\n", "y = dnn_horsepower_model.predict(x)" ] }, { "cell_type": "code", "execution_count": 46, "metadata": { "id": "rsf9rD8I17Wq" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_horsepower(x, y)" ] }, { "cell_type": "markdown", "metadata": { "id": "WxCJKIUpe4io" }, "source": [ "Collect the results on the test set for later:\n" ] }, { "cell_type": "code", "execution_count": 47, "metadata": { "id": "bJjM0dU52XtN" }, "outputs": [], "source": [ "test_results['dnn_horsepower_model'] = dnn_horsepower_model.evaluate(\n", " test_features['Horsepower'], test_labels,\n", " verbose=0)" ] }, { "cell_type": "markdown", "metadata": { "id": "S_2Btebp2e64" }, "source": [ "### Regression using a DNN and multiple inputs\n" ] }, { "cell_type": "markdown", "metadata": { "id": "aKFtezDldLSf" }, "source": [ "Repeat the previous process using all the inputs. The model's performance slightly improves on the validation dataset.\n" ] }, { "cell_type": "code", "execution_count": 48, "metadata": { "id": "c0mhscXh2k36" }, "outputs": [ { "data": { "text/html": [ "
Model: \"sequential_3\"\n",
       "
\n" ], "text/plain": [ "\u001b[1mModel: \"sequential_3\"\u001b[0m\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃ Layer (type)                    ┃ Output Shape           ┃       Param # ┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ normalization (Normalization)   │ (10, 9)                │            19 │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_5 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_6 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_7 (Dense)                 │ ?                      │   0 (unbuilt) │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "
\n" ], "text/plain": [ "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n", "┃\u001b[1m \u001b[0m\u001b[1mLayer (type) \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m Param #\u001b[0m\u001b[1m \u001b[0m┃\n", "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n", "│ normalization (\u001b[38;5;33mNormalization\u001b[0m) │ (\u001b[38;5;34m10\u001b[0m, \u001b[38;5;34m9\u001b[0m) │ \u001b[38;5;34m19\u001b[0m │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_5 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_6 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "├─────────────────────────────────┼────────────────────────┼───────────────┤\n", "│ dense_7 (\u001b[38;5;33mDense\u001b[0m) │ ? │ \u001b[38;5;34m0\u001b[0m (unbuilt) │\n", "└─────────────────────────────────┴────────────────────────┴───────────────┘\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Total params: 19 (80.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m19\u001b[0m (80.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Trainable params: 0 (0.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "text/html": [ "
 Non-trainable params: 19 (80.00 B)\n",
       "
\n" ], "text/plain": [ "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m19\u001b[0m (80.00 B)\n" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dnn_model = build_and_compile_model(normalizer)\n", "dnn_model.summary()" ] }, { "cell_type": "code", "execution_count": 49, "metadata": { "id": "CXDENACl2tuW" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "CPU times: user 5.18 s, sys: 515 ms, total: 5.69 s\n", "Wall time: 5.6 s\n" ] } ], "source": [ "%%time\n", "history = dnn_model.fit(\n", " train_features,\n", " train_labels,\n", " validation_split=0.2,\n", " verbose=0, epochs=100)" ] }, { "cell_type": "code", "execution_count": 50, "metadata": { "id": "-9Dbj0fX23RQ" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plot_loss(history)" ] }, { "cell_type": "markdown", "metadata": { "id": "hWoVYS34fJPZ" }, "source": [ "Collect the results on the test set:\n" ] }, { "cell_type": "code", "execution_count": 51, "metadata": { "id": "-bZIa96W3c7K" }, "outputs": [], "source": [ "test_results['dnn_model'] = dnn_model.evaluate(\n", " test_features, test_labels, verbose=0)" ] }, { "cell_type": "markdown", "metadata": { "id": "uiCucdPLfMkZ" }, "source": [ "## Performance\n" ] }, { "cell_type": "markdown", "metadata": { "id": "rDf1xebEfWBw" }, "source": [ "Since all models have been trained, you can review their test set performance:\n" ] }, { "cell_type": "code", "execution_count": 52, "metadata": { "id": "e5_ooufM5iH2" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Mean absolute error [MPG]
horsepower_model3.649566
linear_model2.476359
dnn_horsepower_model2.927792
dnn_model1.714351
\n", "
" ], "text/plain": [ " Mean absolute error [MPG]\n", "horsepower_model 3.649566\n", "linear_model 2.476359\n", "dnn_horsepower_model 2.927792\n", "dnn_model 1.714351" ] }, "execution_count": 52, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.DataFrame(test_results, index=['Mean absolute error [MPG]']).T" ] }, { "cell_type": "markdown", "metadata": { "id": "DABIVzsCf-QI" }, "source": [ "These results match the validation error observed during training.\n" ] }, { "cell_type": "markdown", "metadata": { "id": "ft603OzXuEZC" }, "source": [ "### Make predictions\n", "\n", "You can now make predictions with the `dnn_model` on the test set using Keras `Model.predict` and review the loss:\n" ] }, { "cell_type": "code", "execution_count": 53, "metadata": { "id": "Xe7RXH3N3CWU" }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "WARNING:tensorflow:5 out of the last 11 calls to .one_step_on_data_distributed at 0x30c8fc860> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "\u001b[1m3/3\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 30ms/step\n" ] }, { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "test_predictions = dnn_model.predict(test_features).flatten()\n", "\n", "a = plt.axes(aspect='equal')\n", "plt.scatter(test_labels, test_predictions)\n", "plt.xlabel('True Values [MPG]')\n", "plt.ylabel('Predictions [MPG]')\n", "lims = [0, 50]\n", "plt.xlim(lims)\n", "plt.ylim(lims)\n", "_ = plt.plot(lims, lims)" ] }, { "cell_type": "markdown", "metadata": { "id": "19wyogbOSU5t" }, "source": [ "It appears that the model predicts reasonably well.\n", "\n", "Now, check the error distribution:\n" ] }, { "cell_type": "code", "execution_count": 54, "metadata": { "id": "f-OHX4DiXd8x" }, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "error = test_predictions - test_labels\n", "plt.hist(error, bins=25)\n", "plt.xlabel('Prediction Error [MPG]')\n", "_ = plt.ylabel('Count')" ] }, { "cell_type": "markdown", "metadata": { "id": "KSyaHUfDT-mZ" }, "source": [ "If you're happy with the model, save it for later use with `Model.save`:\n" ] }, { "cell_type": "code", "execution_count": 55, "metadata": { "id": "4-WwLlmfT-mb" }, "outputs": [], "source": [ "dnn_model.save('dnn_model.keras')" ] }, { "cell_type": "markdown", "metadata": { "id": "Benlnl8UT-me" }, "source": [ "If you reload the model, it gives identical output:\n" ] }, { "cell_type": "code", "execution_count": 56, "metadata": { "id": "dyyyj2zVT-mf" }, "outputs": [], "source": [ "reloaded = tf.keras.models.load_model('dnn_model.keras')\n", "\n", "test_results['reloaded'] = reloaded.evaluate(\n", " test_features, test_labels, verbose=0)" ] }, { "cell_type": "code", "execution_count": 57, "metadata": { "id": "f_GchJ2tg-2o" }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Mean absolute error [MPG]
horsepower_model3.649566
linear_model2.476359
dnn_horsepower_model2.927792
dnn_model1.714351
reloaded1.714351
\n", "
" ], "text/plain": [ " Mean absolute error [MPG]\n", "horsepower_model 3.649566\n", "linear_model 2.476359\n", "dnn_horsepower_model 2.927792\n", "dnn_model 1.714351\n", "reloaded 1.714351" ] }, "execution_count": 57, "metadata": {}, "output_type": "execute_result" } ], "source": [ "pd.DataFrame(test_results, index=['Mean absolute error [MPG]']).T" ] }, { "cell_type": "markdown", "metadata": { "id": "vgGQuV-yqYZH" }, "source": [ "## Conclusion\n", "\n", "This notebook introduced a few techniques to handle a regression problem. Here are a few more tips that may help:\n", "\n", "- Mean squared error (MSE) (`tf.keras.losses.MeanSquaredError`) and mean absolute error (MAE) (`tf.keras.losses.MeanAbsoluteError`) are common loss functions used for regression problems. MAE is less sensitive to outliers. Different loss functions are used for classification problems.\n", "- Similarly, evaluation metrics used for regression differ from classification.\n", "- When numeric input data features have values with different ranges, each feature should be scaled independently to the same range.\n", "- Overfitting is a common problem for DNN models, though it wasn't a problem for this tutorial. Visit the [Overfit and underfit](overfit_and_underfit.ipynb) tutorial for more help with this.\n" ] } ], "metadata": { "colab": { "collapsed_sections": [], "name": "regression.ipynb", "toc_visible": true }, "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.11.0" } }, "nbformat": 4, "nbformat_minor": 4 }