1325 lines
242 KiB
Text
1325 lines
242 KiB
Text
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "MhoQ0WE77laV"
|
|
},
|
|
"source": [
|
|
"##### Copyright 2018 The TensorFlow Authors."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"metadata": {
|
|
"cellView": "form",
|
|
"id": "_ckMIh7O7s6D",
|
|
"tags": []
|
|
},
|
|
"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": 5,
|
|
"metadata": {
|
|
"cellView": "form",
|
|
"execution": {
|
|
"iopub.execute_input": "2020-10-15T01:28:48.696906Z",
|
|
"iopub.status.busy": "2020-10-15T01:28:48.695809Z",
|
|
"iopub.status.idle": "2020-10-15T01:28:48.698639Z",
|
|
"shell.execute_reply": "2020-10-15T01:28:48.697957Z"
|
|
},
|
|
"id": "vasWnqRgy1H4"
|
|
},
|
|
"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": "jYysdyb-CaWM"
|
|
},
|
|
"source": [
|
|
"# Basic classification: Classify images of clothing"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "S5Uhzt6vVIB2"
|
|
},
|
|
"source": [
|
|
"<table class=\"tfo-notebook-buttons\" align=\"left\">\n",
|
|
" <td>\n",
|
|
" <a target=\"_blank\" href=\"https://www.tensorflow.org/tutorials/keras/classification\"><img src=\"https://www.tensorflow.org/images/tf_logo_32px.png\" />View on TensorFlow.org</a>\n",
|
|
" </td>\n",
|
|
" <td>\n",
|
|
" <a target=\"_blank\" href=\"https://colab.research.google.com/github/tensorflow/docs/blob/master/site/en/tutorials/keras/classification.ipynb\"><img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab</a>\n",
|
|
" </td>\n",
|
|
" <td>\n",
|
|
" <a target=\"_blank\" href=\"https://github.com/tensorflow/docs/blob/master/site/en/tutorials/keras/classification.ipynb\"><img src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
|
|
" </td>\n",
|
|
" <td>\n",
|
|
" <a href=\"https://storage.googleapis.com/tensorflow_docs/docs/site/en/tutorials/keras/classification.ipynb\"><img src=\"https://www.tensorflow.org/images/download_logo_32px.png\" />Download notebook</a>\n",
|
|
" </td>\n",
|
|
"</table>"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "FbVhjPpzn6BM"
|
|
},
|
|
"source": [
|
|
"This guide trains a neural network model to classify images of clothing, like sneakers and shirts. It's okay if you don't understand all the details; this is a fast-paced overview of a complete TensorFlow program with the details explained as you go.\n",
|
|
"\n",
|
|
"This guide uses [tf.keras](https://www.tensorflow.org/guide/keras), a high-level API to build and train models in TensorFlow."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"metadata": {
|
|
"id": "dzLKpmZICaWN",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"2.6.0\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# TensorFlow and tf.keras\n",
|
|
"import tensorflow as tf\n",
|
|
"\n",
|
|
"# Helper libraries\n",
|
|
"import numpy as np\n",
|
|
"import matplotlib.pyplot as plt\n",
|
|
"\n",
|
|
"print(tf.__version__)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "yR0EdgrLCaWR"
|
|
},
|
|
"source": [
|
|
"## Import the Fashion MNIST dataset"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "DLdCchMdCaWQ"
|
|
},
|
|
"source": [
|
|
"This guide uses the [Fashion MNIST](https://github.com/zalandoresearch/fashion-mnist) dataset which contains 70,000 grayscale images in 10 categories. The images show individual articles of clothing at low resolution (28 by 28 pixels), as seen here:\n",
|
|
"\n",
|
|
"<table>\n",
|
|
" <tr><td>\n",
|
|
" <img src=\"https://tensorflow.org/images/fashion-mnist-sprite.png\"\n",
|
|
" alt=\"Fashion MNIST sprite\" width=\"600\">\n",
|
|
" </td></tr>\n",
|
|
" <tr><td align=\"center\">\n",
|
|
" <b>Figure 1.</b> <a href=\"https://github.com/zalandoresearch/fashion-mnist\">Fashion-MNIST samples</a> (by Zalando, MIT License).<br/> \n",
|
|
" </td></tr>\n",
|
|
"</table>\n",
|
|
"\n",
|
|
"Fashion MNIST is intended as a drop-in replacement for the classic [MNIST](http://yann.lecun.com/exdb/mnist/) dataset—often used as the \"Hello, World\" of machine learning programs for computer vision. The MNIST dataset contains images of handwritten digits (0, 1, 2, etc.) in a format identical to that of the articles of clothing you'll use here.\n",
|
|
"\n",
|
|
"This guide uses Fashion MNIST for variety, and because it's a slightly more challenging problem than regular MNIST. Both datasets are relatively small and are used to verify that an algorithm works as expected. They're good starting points to test and debug code.\n",
|
|
"\n",
|
|
"Here, 60,000 images are used to train the network and 10,000 images to evaluate how accurately the network learned to classify images. You can access the Fashion MNIST directly from TensorFlow. Import and load the Fashion MNIST data directly from TensorFlow:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 4,
|
|
"metadata": {
|
|
"id": "7MqDQO0KCaWS",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stderr",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"2021-12-13 09:20:48.552934: E tensorflow/core/lib/monitoring/collection_registry.cc:77] Cannot register 2 metrics with the same name: /tensorflow/api/keras/optimizers\n"
|
|
]
|
|
},
|
|
{
|
|
"ename": "AlreadyExistsError",
|
|
"evalue": "Another metric with the same name already exists.",
|
|
"output_type": "error",
|
|
"traceback": [
|
|
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
|
|
"\u001b[0;31mAlreadyExistsError\u001b[0m Traceback (most recent call last)",
|
|
"\u001b[0;32m/tmp/ipykernel_195651/2825506439.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mfashion_mnist\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdatasets\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfashion_mnist\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 3\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mtrain_images\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_labels\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mtest_images\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtest_labels\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mfashion_mnist\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload_data\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/tensorflow/python/util/lazy_loader.py\u001b[0m in \u001b[0;36m__getattr__\u001b[0;34m(self, item)\u001b[0m\n\u001b[1;32m 60\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 61\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m__getattr__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 62\u001b[0;31m \u001b[0mmodule\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_load\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 63\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mgetattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodule\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mitem\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 64\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/tensorflow/python/util/lazy_loader.py\u001b[0m in \u001b[0;36m_load\u001b[0;34m(self)\u001b[0m\n\u001b[1;32m 43\u001b[0m \u001b[0;34m\"\"\"Load the module and insert it into the parent's globals.\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 44\u001b[0m \u001b[0;31m# Import the target module and insert it into the parent's namespace\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 45\u001b[0;31m \u001b[0mmodule\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mimportlib\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mimport_module\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 46\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_parent_module_globals\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_local_name\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodule\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 47\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/__init__.py\u001b[0m in \u001b[0;36mimport_module\u001b[0;34m(name, package)\u001b[0m\n\u001b[1;32m 125\u001b[0m \u001b[0;32mbreak\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 126\u001b[0m \u001b[0mlevel\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 127\u001b[0;31m \u001b[0;32mreturn\u001b[0m \u001b[0m_bootstrap\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_gcd_import\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mname\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mlevel\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpackage\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlevel\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 128\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 129\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_gcd_import\u001b[0;34m(name, package, level)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load_unlocked\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_call_with_frames_removed\u001b[0;34m(f, *args, **kwds)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_gcd_import\u001b[0;34m(name, package, level)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load_unlocked\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_call_with_frames_removed\u001b[0;34m(f, *args, **kwds)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_gcd_import\u001b[0;34m(name, package, level)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load_unlocked\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_call_with_frames_removed\u001b[0;34m(f, *args, **kwds)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_gcd_import\u001b[0;34m(name, package, level)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_find_and_load_unlocked\u001b[0;34m(name, import_)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_load_unlocked\u001b[0;34m(spec)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap_external.py\u001b[0m in \u001b[0;36mexec_module\u001b[0;34m(self, module)\u001b[0m\n",
|
|
"\u001b[0;32m/usr/lib/python3.9/importlib/_bootstrap.py\u001b[0m in \u001b[0;36m_call_with_frames_removed\u001b[0;34m(f, *args, **kwds)\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/__init__.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0;31m# See b/110718070#comment18 for more details about this import.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 25\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmodels\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 26\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 27\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minput_layer\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mInput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/models.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 18\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcompat\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mv2\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mtf\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 20\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mmetrics\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mmetrics_module\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 21\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0moptimizer_v1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 22\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mfunctional\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/metrics.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 25\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnumpy\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mactivations\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 27\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 28\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbase_layer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/activations.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 18\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 20\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mlayers\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0madvanced_activations\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 21\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgeneric_utils\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mdeserialize_keras_object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 22\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mgeneric_utils\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mserialize_keras_object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/layers/__init__.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 21\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 22\u001b[0m \u001b[0;31m# Generic layers.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 23\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minput_layer\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mInput\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 24\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minput_layer\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mInputLayer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 25\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0minput_spec\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mInputSpec\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/engine/input_layer.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 19\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 20\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdistribute\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mdistributed_training_utils\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbase_layer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 22\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mkeras_tensor\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 23\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnode\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mnode_module\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/engine/base_layer.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 41\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mengine\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnode\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mnode_module\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 42\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmixed_precision\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mautocast_variable\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 43\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmixed_precision\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mloss_scale_optimizer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 44\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmixed_precision\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mpolicy\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 45\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msaving\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msaved_model\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mlayer_serialization\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/mixed_precision/loss_scale_optimizer.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 16\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 17\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 18\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0moptimizers\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 19\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmixed_precision\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mloss_scale\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mkeras_loss_scale_module\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 20\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0moptimizer_v2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/optimizers.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mOptimizer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 25\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v1\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mTFOptimizer\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 26\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0madadelta\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0madadelta_v2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 27\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0madagrad\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0madagrad_v2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 28\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0madam\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0madam_v2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/optimizer_v2/adadelta.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 20\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mnumpy\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 21\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mbackend_config\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 22\u001b[0;31m \u001b[0;32mfrom\u001b[0m \u001b[0mkeras\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moptimizer_v2\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0moptimizer_v2\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 23\u001b[0m \u001b[0;32mfrom\u001b[0m \u001b[0mtensorflow\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpython\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutil\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtf_export\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mkeras_export\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 24\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/keras/optimizer_v2/optimizer_v2.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m 34\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 35\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 36\u001b[0;31m keras_optimizers_gauge = tf.__internal__.monitoring.BoolGauge(\n\u001b[0m\u001b[1;32m 37\u001b[0m \"/tensorflow/api/keras/optimizers\", \"keras optimizer usage\", \"method\")\n\u001b[1;32m 38\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/tensorflow/python/eager/monitoring.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, name, description, *labels)\u001b[0m\n\u001b[1;32m 358\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mlabels\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mThe\u001b[0m \u001b[0mlabel\u001b[0m \u001b[0mlist\u001b[0m \u001b[0mof\u001b[0m \u001b[0mthe\u001b[0m \u001b[0mnew\u001b[0m \u001b[0mmetric\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 359\u001b[0m \"\"\"\n\u001b[0;32m--> 360\u001b[0;31m super(BoolGauge, self).__init__('BoolGauge', _bool_gauge_methods,\n\u001b[0m\u001b[1;32m 361\u001b[0m len(labels), name, description, *labels)\n\u001b[1;32m 362\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;32m~/git/Machine-Learning/.venv/lib/python3.9/site-packages/tensorflow/python/eager/monitoring.py\u001b[0m in \u001b[0;36m__init__\u001b[0;34m(self, metric_name, metric_methods, label_length, *args)\u001b[0m\n\u001b[1;32m 133\u001b[0m self._metric_name, len(self._metric_methods)))\n\u001b[1;32m 134\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 135\u001b[0;31m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_metric\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_metric_methods\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_label_length\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m 136\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m 137\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0m__del__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
|
|
"\u001b[0;31mAlreadyExistsError\u001b[0m: Another metric with the same name already exists."
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"fashion_mnist = tf.keras.datasets.fashion_mnist\n",
|
|
"\n",
|
|
"(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "t9FDsUlxCaWW"
|
|
},
|
|
"source": [
|
|
"Loading the dataset returns four NumPy arrays:\n",
|
|
"\n",
|
|
"* The `train_images` and `train_labels` arrays are the *training set*—the data the model uses to learn.\n",
|
|
"* The model is tested against the *test set*, the `test_images`, and `test_labels` arrays.\n",
|
|
"\n",
|
|
"The images are 28x28 NumPy arrays, with pixel values ranging from 0 to 255. The *labels* are an array of integers, ranging from 0 to 9. These correspond to the *class* of clothing the image represents:\n",
|
|
"\n",
|
|
"<table>\n",
|
|
" <tr>\n",
|
|
" <th>Label</th>\n",
|
|
" <th>Class</th>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>0</td>\n",
|
|
" <td>T-shirt/top</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>1</td>\n",
|
|
" <td>Trouser</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>2</td>\n",
|
|
" <td>Pullover</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>3</td>\n",
|
|
" <td>Dress</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>4</td>\n",
|
|
" <td>Coat</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>5</td>\n",
|
|
" <td>Sandal</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>6</td>\n",
|
|
" <td>Shirt</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>7</td>\n",
|
|
" <td>Sneaker</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>8</td>\n",
|
|
" <td>Bag</td>\n",
|
|
" </tr>\n",
|
|
" <tr>\n",
|
|
" <td>9</td>\n",
|
|
" <td>Ankle boot</td>\n",
|
|
" </tr>\n",
|
|
"</table>\n",
|
|
"\n",
|
|
"Each image is mapped to a single label. Since the *class names* are not included with the dataset, store them here to use later when plotting the images:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 6,
|
|
"metadata": {
|
|
"id": "IjnLH5S2CaWx",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
|
|
" 'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Brm0b_KACaWX"
|
|
},
|
|
"source": [
|
|
"## Explore the data\n",
|
|
"\n",
|
|
"Let's explore the format of the dataset before training the model. The following shows there are 60,000 images in the training set, with each image represented as 28 x 28 pixels:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 7,
|
|
"metadata": {
|
|
"id": "zW5k_xz1CaWX",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"(60000, 28, 28)"
|
|
]
|
|
},
|
|
"execution_count": 7,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"train_images.shape"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "cIAcvQqMCaWf"
|
|
},
|
|
"source": [
|
|
"Likewise, there are 60,000 labels in the training set:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 10,
|
|
"metadata": {
|
|
"execution": {
|
|
"iopub.execute_input": "2020-10-15T01:28:55.942785Z",
|
|
"iopub.status.busy": "2020-10-15T01:28:55.941935Z",
|
|
"iopub.status.idle": "2020-10-15T01:28:55.945134Z",
|
|
"shell.execute_reply": "2020-10-15T01:28:55.945674Z"
|
|
},
|
|
"id": "TRFYHB2mCaWb"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"60000"
|
|
]
|
|
},
|
|
"execution_count": 10,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"len(train_labels)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "YSlYxFuRCaWk"
|
|
},
|
|
"source": [
|
|
"Each label is an integer between 0 and 9:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 11,
|
|
"metadata": {
|
|
"execution": {
|
|
"iopub.execute_input": "2020-10-15T01:28:55.950046Z",
|
|
"iopub.status.busy": "2020-10-15T01:28:55.949289Z",
|
|
"iopub.status.idle": "2020-10-15T01:28:55.952794Z",
|
|
"shell.execute_reply": "2020-10-15T01:28:55.952207Z"
|
|
},
|
|
"id": "XKnCTHz4CaWg"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"array([9, 0, 0, ..., 3, 0, 5], dtype=uint8)"
|
|
]
|
|
},
|
|
"execution_count": 11,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"train_labels"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "TMPI88iZpO2T"
|
|
},
|
|
"source": [
|
|
"There are 10,000 images in the test set. Again, each image is represented as 28 x 28 pixels:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 12,
|
|
"metadata": {
|
|
"execution": {
|
|
"iopub.execute_input": "2020-10-15T01:28:55.957082Z",
|
|
"iopub.status.busy": "2020-10-15T01:28:55.956303Z",
|
|
"iopub.status.idle": "2020-10-15T01:28:55.959375Z",
|
|
"shell.execute_reply": "2020-10-15T01:28:55.959771Z"
|
|
},
|
|
"id": "2KFnYlcwCaWl"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"(10000, 28, 28)"
|
|
]
|
|
},
|
|
"execution_count": 12,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"test_images.shape"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "rd0A0Iu0CaWq"
|
|
},
|
|
"source": [
|
|
"And the test set contains 10,000 images labels:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 13,
|
|
"metadata": {
|
|
"execution": {
|
|
"iopub.execute_input": "2020-10-15T01:28:55.963873Z",
|
|
"iopub.status.busy": "2020-10-15T01:28:55.963023Z",
|
|
"iopub.status.idle": "2020-10-15T01:28:55.966085Z",
|
|
"shell.execute_reply": "2020-10-15T01:28:55.966508Z"
|
|
},
|
|
"id": "iJmPr5-ACaWn"
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"10000"
|
|
]
|
|
},
|
|
"execution_count": 13,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"len(test_labels)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "ES6uQoLKCaWr"
|
|
},
|
|
"source": [
|
|
"## Preprocess the data\n",
|
|
"\n",
|
|
"The data must be preprocessed before training the network. If you inspect the first image in the training set, you will see that the pixel values fall in the range of 0 to 255:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 8,
|
|
"metadata": {
|
|
"id": "m4VEw8Ud9Quh",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 432x288 with 2 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"plt.figure()\n",
|
|
"plt.imshow(train_images[0])\n",
|
|
"plt.colorbar()\n",
|
|
"plt.grid(False)\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Wz7l27Lz9S1P"
|
|
},
|
|
"source": [
|
|
"Scale these values to a range of 0 to 1 before feeding them to the neural network model. To do so, divide the values by 255. It's important that the *training set* and the *testing set* be preprocessed in the same way:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 9,
|
|
"metadata": {
|
|
"id": "bW5WzIPlCaWv",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"train_images = train_images / 255.0\n",
|
|
"\n",
|
|
"test_images = test_images / 255.0"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Ee638AlnCaWz"
|
|
},
|
|
"source": [
|
|
"To verify that the data is in the correct format and that you're ready to build and train the network, let's display the first 25 images from the *training set* and display the class name below each image."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 10,
|
|
"metadata": {
|
|
"id": "oZTImqg_CaW1",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 720x720 with 25 Axes>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"plt.figure(figsize=(10,10))\n",
|
|
"for i in range(25):\n",
|
|
" plt.subplot(5,5,i+1)\n",
|
|
" plt.xticks([])\n",
|
|
" plt.yticks([])\n",
|
|
" plt.grid(False)\n",
|
|
" plt.imshow(train_images[i], cmap=plt.cm.binary)\n",
|
|
" plt.xlabel(class_names[train_labels[i]])\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "59veuiEZCaW4"
|
|
},
|
|
"source": [
|
|
"## Build the model\n",
|
|
"\n",
|
|
"Building the neural network requires configuring the layers of the model, then compiling the model."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Gxg1XGm0eOBy"
|
|
},
|
|
"source": [
|
|
"### Set up the layers\n",
|
|
"\n",
|
|
"The basic building block of a neural network is the *layer*. Layers extract representations from the data fed into them. Hopefully, these representations are meaningful for the problem at hand.\n",
|
|
"\n",
|
|
"Most of deep learning consists of chaining together simple layers. Most layers, such as `tf.keras.layers.Dense`, have parameters that are learned during training."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 41,
|
|
"metadata": {
|
|
"id": "9ODch-OFCaW4",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"model = tf.keras.Sequential([\n",
|
|
" tf.keras.layers.Flatten(input_shape=(28, 28)),\n",
|
|
" tf.keras.layers.Dense(512, kernel_regularizer=tf.keras.regularizers.l2(0.0001),\n",
|
|
" activation='elu'),\n",
|
|
" tf.keras.layers.Dropout(0.5),\n",
|
|
" tf.keras.layers.Dense(512, kernel_regularizer=tf.keras.regularizers.l2(0.0001),\n",
|
|
" activation='elu'),\n",
|
|
" tf.keras.layers.Dropout(0.5),\n",
|
|
" tf.keras.layers.Dense(512, kernel_regularizer=tf.keras.regularizers.l2(0.0001),\n",
|
|
" activation='elu'),\n",
|
|
" tf.keras.layers.Dropout(0.5),\n",
|
|
" tf.keras.layers.Dense(512, kernel_regularizer=tf.keras.regularizers.l2(0.0001),\n",
|
|
" activation='elu'),\n",
|
|
" tf.keras.layers.Dropout(0.5),\n",
|
|
" tf.keras.layers.Dense(10)\n",
|
|
"])\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "gut8A_7rCaW6"
|
|
},
|
|
"source": [
|
|
"The first layer in this network, `tf.keras.layers.Flatten`, transforms the format of the images from a two-dimensional array (of 28 by 28 pixels) to a one-dimensional array (of 28 * 28 = 784 pixels). Think of this layer as unstacking rows of pixels in the image and lining them up. This layer has no parameters to learn; it only reformats the data.\n",
|
|
"\n",
|
|
"After the pixels are flattened, the network consists of a sequence of two `tf.keras.layers.Dense` layers. These are densely connected, or fully connected, neural layers. The first `Dense` layer has 128 nodes (or neurons). The second (and last) layer returns a logits array with length of 10. Each node contains a score that indicates the current image belongs to one of the 10 classes.\n",
|
|
"\n",
|
|
"### Compile the model\n",
|
|
"\n",
|
|
"Before the model is ready for training, it needs a few more settings. These are added during the model's *compile* step:\n",
|
|
"\n",
|
|
"* *Loss function* —This measures how accurate the model is during training. You want to minimize this function to \"steer\" the model in the right direction.\n",
|
|
"* *Optimizer* —This is how the model is updated based on the data it sees and its loss function.\n",
|
|
"* *Metrics* —Used to monitor the training and testing steps. The following example uses *accuracy*, the fraction of the images that are correctly classified."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 42,
|
|
"metadata": {
|
|
"id": "Lhan11blCaW7",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"model.compile(optimizer='adam',\n",
|
|
" loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
|
|
" metrics=['accuracy'])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "qKF6uW-BCaW-"
|
|
},
|
|
"source": [
|
|
"## Train the model\n",
|
|
"\n",
|
|
"Training the neural network model requires the following steps:\n",
|
|
"\n",
|
|
"1. Feed the training data to the model. In this example, the training data is in the `train_images` and `train_labels` arrays.\n",
|
|
"2. The model learns to associate images and labels.\n",
|
|
"3. You ask the model to make predictions about a test set—in this example, the `test_images` array.\n",
|
|
"4. Verify that the predictions match the labels from the `test_labels` array.\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Z4P4zIV7E28Z"
|
|
},
|
|
"source": [
|
|
"### Feed the model\n",
|
|
"\n",
|
|
"To start training, call the `model.fit` method—so called because it \"fits\" the model to the training data:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 43,
|
|
"metadata": {
|
|
"id": "xvwvpA64CaW_",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"Epoch 1/10\n",
|
|
"1875/1875 [==============================] - 22s 12ms/step - loss: 1.3382 - accuracy: 0.6500\n",
|
|
"Epoch 2/10\n",
|
|
"1875/1875 [==============================] - 25s 13ms/step - loss: 0.8738 - accuracy: 0.7803\n",
|
|
"Epoch 3/10\n",
|
|
"1875/1875 [==============================] - 20s 11ms/step - loss: 0.8724 - accuracy: 0.7894\n",
|
|
"Epoch 4/10\n",
|
|
"1875/1875 [==============================] - 17s 9ms/step - loss: 0.8851 - accuracy: 0.7940\n",
|
|
"Epoch 5/10\n",
|
|
"1875/1875 [==============================] - 21s 11ms/step - loss: 0.8805 - accuracy: 0.7995\n",
|
|
"Epoch 6/10\n",
|
|
"1875/1875 [==============================] - 17s 9ms/step - loss: 0.8647 - accuracy: 0.8057\n",
|
|
"Epoch 7/10\n",
|
|
"1875/1875 [==============================] - 22s 12ms/step - loss: 0.8748 - accuracy: 0.8038\n",
|
|
"Epoch 8/10\n",
|
|
"1875/1875 [==============================] - 17s 9ms/step - loss: 0.8639 - accuracy: 0.8056\n",
|
|
"Epoch 9/10\n",
|
|
"1875/1875 [==============================] - 20s 11ms/step - loss: 0.8648 - accuracy: 0.8060\n",
|
|
"Epoch 10/10\n",
|
|
"1875/1875 [==============================] - 21s 11ms/step - loss: 0.8676 - accuracy: 0.8057\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"<tensorflow.python.keras.callbacks.History at 0x7f5e984740d0>"
|
|
]
|
|
},
|
|
"execution_count": 43,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"model.fit(train_images, train_labels, epochs=10)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "W3ZVOhugCaXA"
|
|
},
|
|
"source": [
|
|
"As the model trains, the loss and accuracy metrics are displayed. This model reaches an accuracy of about 0.91 (or 91%) on the training data."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "wCpr6DGyE28h"
|
|
},
|
|
"source": [
|
|
"### Evaluate accuracy\n",
|
|
"\n",
|
|
"Next, compare how the model performs on the test dataset:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 46,
|
|
"metadata": {
|
|
"id": "VflXLEeECaXC",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"313/313 - 1s - loss: 0.7762 - accuracy: 0.8343\n",
|
|
"\n",
|
|
"Test accuracy: 0.8342999815940857\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"test_loss, test_acc = model.evaluate(test_images, test_labels, verbose=2)\n",
|
|
"\n",
|
|
"print('\\nTest accuracy:', test_acc)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "yWfgsmVXCaXG"
|
|
},
|
|
"source": [
|
|
"It turns out that the accuracy on the test dataset is a little less than the accuracy on the training dataset. This gap between training accuracy and test accuracy represents *overfitting*. Overfitting happens when a machine learning model performs worse on new, previously unseen inputs than it does on the training data. An overfitted model \"memorizes\" the noise and details in the training dataset to a point where it negatively impacts the performance of the model on the new data. For more information, see the following:\n",
|
|
"* [Demonstrate overfitting](https://www.tensorflow.org/tutorials/keras/overfit_and_underfit#demonstrate_overfitting)\n",
|
|
"* [Strategies to prevent overfitting](https://www.tensorflow.org/tutorials/keras/overfit_and_underfit#strategies_to_prevent_overfitting)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "v-PyD1SYE28q"
|
|
},
|
|
"source": [
|
|
"### Make predictions\n",
|
|
"\n",
|
|
"With the model trained, you can use it to make predictions about some images.\n",
|
|
"The model's linear outputs, [logits](https://developers.google.com/machine-learning/glossary#logits). Attach a softmax layer to convert the logits to probabilities, which are easier to interpret. "
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 16,
|
|
"metadata": {
|
|
"id": "DnfNA0CrQLSD",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"probability_model = tf.keras.Sequential([model, \n",
|
|
" tf.keras.layers.Softmax()])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 17,
|
|
"metadata": {
|
|
"id": "Gl91RPhdCaXI",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"predictions = probability_model.predict(test_images)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "x9Kk1voUCaXJ"
|
|
},
|
|
"source": [
|
|
"Here, the model has predicted the label for each image in the testing set. Let's take a look at the first prediction:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 19,
|
|
"metadata": {
|
|
"id": "3DmJEUinCaXK",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"array([6.7990785e-10, 7.6563593e-14, 1.7743880e-12, 1.7664440e-14,\n",
|
|
" 3.1395782e-12, 1.1033747e-04, 3.6911777e-13, 2.4555137e-03,\n",
|
|
" 1.5690568e-10, 9.9743420e-01], dtype=float32)"
|
|
]
|
|
},
|
|
"execution_count": 19,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"predictions[0]"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "-hw1hgeSCaXN"
|
|
},
|
|
"source": [
|
|
"A prediction is an array of 10 numbers. They represent the model's \"confidence\" that the image corresponds to each of the 10 different articles of clothing. You can see which label has the highest confidence value:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 20,
|
|
"metadata": {
|
|
"id": "qsqenuPnCaXO",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"9"
|
|
]
|
|
},
|
|
"execution_count": 20,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"np.argmax(predictions[0])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "E51yS7iCCaXO"
|
|
},
|
|
"source": [
|
|
"So, the model is most confident that this image is an ankle boot, or `class_names[9]`. Examining the test label shows that this classification is correct:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 21,
|
|
"metadata": {
|
|
"id": "Sd7Pgsu6CaXP",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"9"
|
|
]
|
|
},
|
|
"execution_count": 21,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"test_labels[0]"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "ygh2yYC972ne"
|
|
},
|
|
"source": [
|
|
"Graph this to look at the full set of 10 class predictions."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 23,
|
|
"metadata": {
|
|
"id": "DvYmmrpIy6Y1",
|
|
"tags": []
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"def plot_image(i, predictions_array, true_label, img):\n",
|
|
" true_label, img = true_label[i], img[i]\n",
|
|
" plt.grid(False)\n",
|
|
" plt.xticks([])\n",
|
|
" plt.yticks([])\n",
|
|
"\n",
|
|
" plt.imshow(img, cmap=plt.cm.binary)\n",
|
|
"\n",
|
|
" predicted_label = np.argmax(predictions_array)\n",
|
|
" if predicted_label == true_label:\n",
|
|
" color = 'blue'\n",
|
|
" else:\n",
|
|
" color = 'red'\n",
|
|
"\n",
|
|
" plt.xlabel(\"{} {:2.0f}% ({})\".format(class_names[predicted_label],\n",
|
|
" 100*np.max(predictions_array),\n",
|
|
" class_names[true_label]),\n",
|
|
" color=color)\n",
|
|
"\n",
|
|
"def plot_value_array(i, predictions_array, true_label):\n",
|
|
" true_label = true_label[i]\n",
|
|
" plt.grid(False)\n",
|
|
" plt.xticks(range(10))\n",
|
|
" plt.yticks([])\n",
|
|
" thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
|
|
" plt.ylim([0, 1])\n",
|
|
" predicted_label = np.argmax(predictions_array)\n",
|
|
"\n",
|
|
" thisplot[predicted_label].set_color('red')\n",
|
|
" thisplot[true_label].set_color('blue')"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "Zh9yABaME29S"
|
|
},
|
|
"source": [
|
|
"### Verify predictions\n",
|
|
"\n",
|
|
"With the model trained, you can use it to make predictions about some images."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "d4Ov9OFDMmOD"
|
|
},
|
|
"source": [
|
|
"Let's look at the 0th image, predictions, and prediction array. Correct prediction labels are blue and incorrect prediction labels are red. The number gives the percentage (out of 100) for the predicted label."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 24,
|
|
"metadata": {
|
|
"id": "HV5jw-5HwSmO",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 432x216 with 2 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"i = 0\n",
|
|
"plt.figure(figsize=(6,3))\n",
|
|
"plt.subplot(1,2,1)\n",
|
|
"plot_image(i, predictions[i], test_labels, test_images)\n",
|
|
"plt.subplot(1,2,2)\n",
|
|
"plot_value_array(i, predictions[i], test_labels)\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 25,
|
|
"metadata": {
|
|
"id": "Ko-uzOufSCSe",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 432x216 with 2 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"i = 12\n",
|
|
"plt.figure(figsize=(6,3))\n",
|
|
"plt.subplot(1,2,1)\n",
|
|
"plot_image(i, predictions[i], test_labels, test_images)\n",
|
|
"plt.subplot(1,2,2)\n",
|
|
"plot_value_array(i, predictions[i], test_labels)\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "kgdvGD52CaXR"
|
|
},
|
|
"source": [
|
|
"Let's plot several images with their predictions. Note that the model can be wrong even when very confident."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 47,
|
|
"metadata": {
|
|
"id": "hQlnbqaw2Qu_",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "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",
|
|
"text/plain": [
|
|
"<Figure size 1440x720 with 50 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"# Plot the first X test images, their predicted labels, and the true labels.\n",
|
|
"# Color correct predictions in blue and incorrect predictions in red.\n",
|
|
"num_rows = 5\n",
|
|
"num_cols = 5\n",
|
|
"num_images = num_rows*num_cols\n",
|
|
"plt.figure(figsize=(2*2*num_cols, 2*num_rows))\n",
|
|
"for i in range(num_images):\n",
|
|
" plt.subplot(num_rows, 2*num_cols, 2*i+1)\n",
|
|
" plot_image(i, predictions[i], test_labels, test_images)\n",
|
|
" plt.subplot(num_rows, 2*num_cols, 2*i+2)\n",
|
|
" plot_value_array(i, predictions[i], test_labels)\n",
|
|
"plt.tight_layout()\n",
|
|
"plt.show()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "R32zteKHCaXT"
|
|
},
|
|
"source": [
|
|
"## Use the trained model\n",
|
|
"\n",
|
|
"Finally, use the trained model to make a prediction about a single image."
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 27,
|
|
"metadata": {
|
|
"id": "yRJ7JU7JCaXT",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"(28, 28)\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Grab an image from the test dataset.\n",
|
|
"img = test_images[1]\n",
|
|
"\n",
|
|
"print(img.shape)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "vz3bVp21CaXV"
|
|
},
|
|
"source": [
|
|
"`tf.keras` models are optimized to make predictions on a *batch*, or collection, of examples at once. Accordingly, even though you're using a single image, you need to add it to a list:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 28,
|
|
"metadata": {
|
|
"id": "lDFh5yF_CaXW",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"(1, 28, 28)\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Add the image to a batch where it's the only member.\n",
|
|
"img = (np.expand_dims(img,0))\n",
|
|
"\n",
|
|
"print(img.shape)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "EQ5wLTkcCaXY"
|
|
},
|
|
"source": [
|
|
"Now predict the correct label for this image:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 29,
|
|
"metadata": {
|
|
"id": "o_rzNSdrCaXY",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"[[3.8127650e-06 1.1930970e-19 9.9880171e-01 2.7449465e-13 9.5398573e-04\n",
|
|
" 4.1459834e-17 2.4049767e-04 2.3141040e-15 3.1803546e-15 1.4722940e-13]]\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"predictions_single = probability_model.predict(img)\n",
|
|
"\n",
|
|
"print(predictions_single)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 30,
|
|
"metadata": {
|
|
"id": "6Ai-cpLjO-3A",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAWAAAAEbCAYAAADkhF5OAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8+yak3AAAACXBIWXMAAAsTAAALEwEAmpwYAAAdjElEQVR4nO3debhdVXnH8e97E0ICGIYkEAYhQAYmAQmIyGQTwgwRTMAIBatBhlCGMMSiQLAgQ6WtiGhEEIgDkTAEUaBipYBiFdAoCqI4FIcKaBULCsS8/eN9T+7O6U3IvWefsy6X3+d58tyzz7m5a4+/vfZaa+9j7o6IiHReV+kZEBF5rVIAi4gUogAWESlEASwiUogCWESkEAWwiEghg3vzyyNHjvQxY8a0aVYGpsWLYcmS9pYxeDDssEN7yxCRvnv44YefdfdRze/3KoDHjBnDQw89VN9cvQaYtb+MJUtAm0Wk/zKzX/T0vpogREQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkUIUwCIihSiARUQKUQCLiBSiABYRKUQBLCJSiAJYRKQQBbCISCEKYBGRQhTAIiKFKIBFRApRAIuIFKIAFhEpRAEsIlKIAlhEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gUogAWESlEASwiUogCWESkEAWwiEghCmARkULM3Vf9l82eAX7RvtlZzkjg2Q6VpbL7R/kqW2UP1LI3c/dRzW/2KoA7ycwecvedVfZrp3yVrbJfC2VXqQlCRKQQBbCISCH9OYA/qbJfc+WrbJX9Wih7mX7bBiwiMtD15xqwiMiApgAWESlEAdwHZmal50E6w8zW0vYWADPryp+17Q8K4F4yM/NsODezKWa2cYl5GIhlZXnrVV5P6GTZPczLOGA+sGOHyhvciXJ6o7r9zWxIJ8rpj8xsXWB4Tu5Y198dMAHc2IBmNszM1mhXOZXw3R2YA/ypXWX1pOkEcIiZrdOhsiaZ2TbtKivL6AImmdkVZnYCMMfMhr/S/2sXd/8x8DPgfWa2fTvLMrO1gV3y9ZR2r+tV0bT9jwHe2agF1lzOeOCYdgZ8DfYk9scLgPm1XRm5+4D5B0wFvgJ8A/hbYM02lTONuCX7kJweUmBZpwD3AiM6UNZpwLeBzTu0bN8Bfg+8PqdX6/C6NaCrMn0JcDOwfRvLHE+c0BcBj7dr3+3jvO0MLATWqnMd5889829/GzgCWL308q5knr8K/AHYq66/OZBqwFsBs4D3AxcRAXxUftbSmaqH/78IeI4IJtz9JTMb1EoZvZyf3YFPA//s7r8zs9XbWNYkYj3u4e4/M7M3mtnkmsuoXuauTuzoDwAXm9lgd3+5zvJeaV48LDWzEQDu/j7gh8AFddeEG8vu7k8AGwNvAr4AvFhnOX1lZhOBTxFXen+u6++6u5vZnsDHgWuA7wJ7AUf2l5pwD8f9FcS2OTrzpvUyMtlf1cxsLHAxsTzT8r3dgAXADHf/egt/u3oZth/wV+BR4Blip3nE3Y/Nzwe5+19bWZZXmoecXpNYto3cfac6y+6hrHHA+4gz/1JgV+AF4Dp3v7HO8sxsa+A5d/9VTn8eGOzu083srUTt6O5Wy1zF+TqZWNangGvc/UkzmwtsC1zs7o/UUEZ12Y8F9gH+HRhHXAF8zt1/bWYjgd95Bw7W5u2f7x0HHAucDjxU13yY2dlErfq8PPG+BziMCPxb3f2lOsrp47xVt83bgcHA8+5+h5n9E7AhsT4OB/7i7tf3pZxXbQ246ez0c+ARYN1sP1vD3R8EbiSeetRnlY1wJhFE+xMdM+OAnYDtzOy2/N22hq+Z7WNmBwPrAkcD3zWzRY3wbbUW3lTWiWZ2OBG2XwU2Am4CDgXuI3bIllXKOx2YB9xgZvOyrfG9wCAzewT4F+AndZT5SjJwjgD+IX9ebGZ7uvtcYl87vdWrDjMbXln2NxP71Ynu/mngP4FNgGlmdg5wHtC2q5yqyjzNNLNzc7+fT2z784Cda2wHfhzY1cy2cfcX3f2qfP/NQNEO2Mp6OAU4G9gBeLeZXefuZwFPE/vkWUSTWZ8LetX9o7vmvgcwAzg0p88kbjE8B5hEtNPuUUN5Y4kzMkTzxiKyrQpYDbifCChr4zKfAfwHcCXR9rsbsBZwdU531VjWyUSb3NgePjsqd7itaizvaOD+fP0h4Hng2srn04AxbVy3uxD9B6vnOv0nYDRwCnAP0QZ8N7Bn/v7IFsvbkgj3ocTJ9FbgIeAtld85CDgXeJA2tj2vYP5OzeWeBHwPmJ3vz8l9fac+/M3GMTsRmAxsCgzJ7f1+oullPHAn0d5+bieXeQXzPDS3zdicHgJ8FpiT01sB67dURumFbGHl7Ad8PwPxq8At+f4pwLeAjwBT8r1ehVNzkAKbATfk37yjEr6H1xl8K5mf8ZXlm5M7aVf+Gw58jOywqqGskXnwbZthNAOYDexNXJJ/FXhDi2U0r9/t84CcRZzc1gaeBG4hmiDavX7fmfvMQTk9JEPy7srvPEk0c61RQ3mbA+sQw5nGVvav2cCWld/rogOdcSzf4dhF1OyMqNB8iTgxDcrPT+7tvlYJ3wOAJ4APAL8lOvcm5j79TaLmvxUwHbisUWan/vWwX65BVHoOqrx3ANH3Uk+ZnVzAOneY3GGPrLx3J/CxfH0+0WC+O73sQa9uBOBIYOs8IK8kmjk2yc9mAg/T4hlwFeZnJPC6XJ5rgS83lok4AQxv3nFa3OkG57IuIGoi1xCXn6cQtf11Wlye6vpdGxjetE0PyOkP5kE5qp37UeX1PxIdf9NzelPiErnRPHAzsHGNy/46ovJwDTCGuOSeT7QrjmvnPrWS+dsvt8lC4iS8kBzhA5xAVmj6+Le3Ja6ctsxyfpvrd3J+Poq4GtiP6PDcrsPLXt02uxGVnuG57Z8kr06A44hKwtBWjrtlZZXY0H1YOUOACfl6TG6sK4AjKr+zGXB9vl4NuJw4i/apxkLUxr4PbJHTU4hL0a8Rl02PAtu2ebl3zDLXy+W9n6whAX9HXB6Ormmnm0rUdidkudPIYWd58C2kxtooUbu6jejIPIpo8zyLqM2fT9R+N+nQ/nUS8Hni8vIp4LDKcn+dONG2tK17OliB1xOX35/I/Xp8rpNZda7rlczTTnQ3q7wO+Ga+npyh0zgZHQs81jgWVvFvb0lUEKZW3htPVIoeyuk5wEvApJweTvQDtPW4eoX5/nui+e16omlvX+BA4FfAVcTJYZvayiu1oL1YISOI4SnHAZfmAbsmMczsN42NBfwNUYMZldOD6UXtqSmM3pAH3cY5vTfRHtZoK5xG5VKxjcu+C1Hr3oroBJhHtEl9DPhBXTsqcCIxdvp44GVg98Y6IYL++63udMSl5puIS+9JxCX/cKKTax7wLqKJ41Tgi504CHP5tiTaXxtjjmfkATgtp9en1Xa+SnARQxc/TXRobUBc4ZyX23QLokliow4s++Dc7v9e2d6NbTI09/GfAtfRyxMQEbSPEpWgbwAnVD6bCVyVr/fO8t9c+byjY76b5ntcLutoovlhnzz+xhEVk22oqalvWZmlFnYVV8gQos3p7cRl8YvA+ZXPTyXOzFcQZ6ZGG15L7bJEwH+UqJl8gqh53kalxt3m5R7dWIY8YG/MA2aL3Cmm08JNESx/slk/l21t4N1EG++gymeX03r4HgQsJmpSmxBNO/Mrn+9LpYbVzoOQ/9/kMoSo/e5J1jqJ5o8/AfvXUN4I4EdEu+ebcl86mqhM3E4MZxpBXOn8Mx2o+TbN23uJE95kYngdle0/juhcXuUTUIbUd+i+SeloIuh3zOk9ibG0HyFOdG/qabt0aPm7qmXn8fWlpt+ZC8xs2zx0eqH7sJJmEb3BE4hL/4uAd9DdDro7UWN9Y6sbkmh/+o98fQhxKbx9Tl8AXNrunYWoBV5NjOYYnoF1Kb24/HuFv18N3xlEDXBOlnlXZb2ekQHcUiAQtZyfALtW3tuRuMSrvnctOWKlXeu3adnH0H2Fc1nuY2Ny+sAMiVbbfA/OdTuJuHK7Gzg4PxuVB/dtxA0Y69Hi6IreroOcHk40wXyLGOd9A3ES/ixR+ehVRxgxMmlpZfp7RP/Md3MbDyIqEXOp4QRXx7ogR2Tk69vIpsycvgi4sG3zUHIFrOpKyg14Yk6fCvxrHiR7EO10farx9rAzdhEN7Dc1vf9Oah561dM8EG3XQ4HtiJrnv+XBsRi4suZypxIdLVvmgfYo3R1i07PMMTWUMxs4NV83aphrEyfTS4m24HcRbY5ta/NtWs+ziSacL2b5w4hL7flE5+NDtHjCy/BdTHdTxsb5d6+u/M4IYsjbjb0NuhrWwWSi8rJBTp9ItHm+mzjpj6GPzWzESIGf5v51Xr43hBgBccaK5qlT/3LZ98nXf0+cHD6Tx9w6RIfr3cRJeTHZ/9SWeen0wq/CyhkKvC5fb5o/x+bBuj5xxj6duIXxaSpDRFooczw53Ic4Qy8AvpzTWxO1oZaGXq3CPJyYO8HVwN753iTgGODHRBPLujWVtTNRy2mc1NbPg+W6DIOHaLEXmu7axUfJGgSVZywQPd4fIO56uoYOdbwQVxifIYaCbUW0+TXGdb6R6FtoqX2faEL6GrBLTjf2rT2Jh/ucXPndjtR8m+bvZKLGe27uW43j7ASiiWT3GsqYDCxh+ZEm72kO4BL/iCu/pcTInqtyP9gu98MbiIrY8UT/R+2VruXmpfTK6GHlHEDUco8h2gVHER0VH6K7XWkNogOjMTKiT2fRXNGb5UH4d+SICeIk8CBwR07X9hCSFczHYURH17bEWNPLgaMrn0+gtTbfccRwqklE7XNDoonjXmCH/J21iSuKKY0DsqZlm5ThPrGyzhs14dOJk1/bH8CS4b8DcUl8Ld3tnJtlGH2kxrLWJa5e3pD70txc1wuIuwifAi5o9zKvYN72IUZ2rEnc4fVfxG3mjfb346hvTPmBwE/y9Vhi2Nm+JZa7uh/kz8ZdnvNzejBRuVtInjg7Mj8lV8YKVlAXUXv4E3n5lu+/lWi0n1jHBmh6bw+iRngM3bWVc4hB2C21A65gHvYix7vm9NnA+/P1IOKS/BbqGfR/ENGTe2sG4c+Js/0mRLv2R2njmMs80OcS7awTK++/I+er1l7lVdjWx2UIvoXu9u7Nc5/boKf/05dyiTb0u4FfElcWM4lL/kuIZo9lI3ba+a95eYga94ZEh+g9+d58YjhY7duCGEf7AtHkU7TNt7JtGiH8dqImvHfl888Bh3dsfkqvkB5W0La5c3yJaADfrHKgTCM6xlq+O4io8V4NXEjcifWGDOEzMjBubdcBQrSxPtPYITMkb6dyyylZg2qxnP2JmxmqO9j5GcJbE00P52dAtK+dK9o/zyNOaB/Odf54O4O/qfxpRG//Djl9fK7fPSr7Vq2jD4i7CHej6RGLxCXulDqCfhXmodrmOwHYujL9AWBWvn4PcYfn/7v1vKb5mEyOre4P/5pC+Chi6OUHiOecLKaDN8IUXxlNK2YLohNsNFETXkAMMVuT6E3/B6LG1qtxkpWV3fh5HNHOeSRxGfz9PFi2IR648/lWw28F8zGR7vbdw4mOiv2JQfAXEs0PhxAdZI+0cgIgajpL6e51H1r57AKi7W8YcfI5ixZu6FjF+RmWgTc3A3B8G8tao/L6NKJd83yilnt8vj+TuPV1tw7u39OJ5q62BN1Kyp1NVC7uIvoz1iPaPz+Zx9d97d7+OR8lOtx6LDPzpZEH0/NYuYoO3fyzbD46vUKaVsJo4maDLqKN6KfAhyufr5U7zKeAXxPDxPbuQzljmv7mqcCBlffelsHfaH5oy/Md8kB4gO67j6YTvf97EZfBJxAjPm4ia2stlndQnlxG5HS1JnYv3UP3+u1DsPu4zP9K1Lp3BRbk+2fmSe2T5I0BRJNTbe3dK5mnDfNE8AM6f4vtFODOfH0h+XyLDOEZROd2R+epg8tevQI4gLgamQAMa3xeCeEDaeNV4ArnsfAKmtm0Qi4kOt42rfzOMKK9bsfqSl3VsykxHOgJojPkvURn3jzgtsrvbEC0/azTgR1hFvGtHXvl9BEZwo174odRQ9tv0473JDmCgu5L7kV0+ClbHdifGkO/3lZZlxtnKN9LdLTMzSA8voPzNSznoaM13yx7J2Jo2YVEm3Rj+3es5l9gmZvbvWcTtfxLiKuefaq/u6pZ0pZ57Qcra32iV/pvcvrSXEk9NjP0ZmXR/WCPCRmyt1Q+u5e4JXQQ0SH0ddr/YJ3G7a4nNIXw24me6Lb0EPcQwsfkOm7r8nZ4P2oe+jWMeIrXJkTt95zKsn98IC17ZR00B48RTUzfIh7i1Ki8zCSaZdYtGT5tXA8j82fjyvrGnD6NaIbpIsbcF1/2jn8La+UrWDx/Pm1mvwRmmNnL7j7HzD4EfMXM9vX8doSGxv9bhXL2JTo8HiA68i4GRpnZtu7+A6K2dGv+zljgPe7+dD1L2eP8bA+cbWYL3f0TuR7ONbMPuvvNZvYy0QRTO3e/M7/h4T4zu4oY69rW5S3gRaIz5S9mNpS4A20PYizqesSDxMcSV1OHDrBlB5Z7iPipRPD+nGjjvYI46Z9kZhsQTW4z3P1/ysxpe+QxNQr4mZnNcPfbzez3wDNm9jnihHOwx9dNHUGcsH9bcJY7G8Bmtrq7v5iv30KsrAc9vpLkTOKJ8+7u55jZasTdOL9a8V9cYTmTiWdHnE7UjHYi2kLHALuY2R/d/ZfAlPx6n0Hu/lzrS7is/OVOMvn6e2b2HWA/M1vq7h83MwcuN7PT3P32usrvSYbwIGJ42xvzJDSQ/IG4xP4wMZLmHuKmkseI9r3riOFQH3L3jnyzRqfkN8C8kK/3IEZ9zCNC+Dqi1v80MfxwEDG884kyc9teWaF7N/BpM3uXu3/RzJ4nbrJ5r7svya9/OpO4CiiqY98JZ/H16beRtw4Tw65+QgwEvyPPVrOJTrmPu/t9LZS1C9HW9Y38nrF3EF8ouA5x++cDwL3u/os+L9DKyx/s7kvy9X7E5e78nD6FeBbCF9z9rtxZ7nH3/2rHvPQwb8sO1oHGzNYihhO+HlhUOdlfD9zu7jeXnL92MLODiI62y4jO3JOAyzy+u2wUcavtdsBpndrH+gMz25/owJ9KnITnEv0BzxCds0f0h0pIR7+U08zm0P2Yu/Pd/XEzO544aP4tQ/hs4C53/14N5XXl5cYE4nkOzxN3fG1JnAAWeM3f42ZmU4hOj8VE5x/EsLdr3f2m/J1ridrJue5+Z53ly/LMbDoxtPAId3+y9PzUKb8f8CLieQuLzOz1xPj5b7j7Cfk7I4jmmM2JY+Cv7r601Dy3Q+M4z9fvBDZ098vNbBrRv3Sguz+QV90jgMX95WTUkSaIxpc9uvulZvYsMZb3emIw/k2AA1PzyyUvq6vcxkZx9x9lG9CRRDvhY8DX2hC++xOPMpxPdC7uT9zaeC3wt7kevkDcBDCEGBMqbWBmGxLb+zjim1MGWviOJm4amunu3zazNd39KTM7CZhvZie7+5Xu/jszu4QYWvly2bmun5ntAFxiZtPc/Xm6a7m4+0IzWwosMrOZ7n5ryXntSdsDuBG+2Tb1R3e/xuJrti81s9+6+8NmtpBom2pbu1SG8ELibper3f13df59M1uP6Gmemu1OmxKXhasTIQzwQTObSrRRThuIHUH9yB+Im02mDrQ239Tc6XiWmb2V6FR6CphjZqPc/Xx3/3252Wwvd19sZkuABWZ2GNHM+Ezl81uyT+YKM/sK8EJ/ugLoSBOEmR1APPH/WHe/P987ibhUn+Xu/1ltN23zvKzWrppAtsddRoyxfM7MPgvc5+7z8vNtiSdifWWg1cikszJUZhMPs290Oj5AXN0dAjxLjHY4zN2fWcGfedXK5e9qXMWa2c3E8yyezJ+PEScpiCGmL7j7n0vM68q0PYDNbGOiZnicu38rLxmGE7WTw4lOg7cA/9ufzkx9lSebK4ge+Y2Ao9z9z40rgbJzJwPJSjodbyCawe4ZiPtc9Vgys40bQ1XNbB7R5DSP7v6eYcQY8H7R5tus9gA2s22I5ygsyOm1ieeOvkAMOxtHDBD/vLt/ysw2d/ef1ToThZnZPkQ77+gcFjPU3f9Ser5k4Kt0Oh45EJtemsL3ZOJhOg8T3zP3QzP7GHHD06H5O0Pc/aVyc7xyXXX+MTMbT3yVyZqN99z9j8TdaEOI23/3IdpEJ+bnAyp8Adz9HuLW06+Z2foKX2k3M9vQzE4jhlsdOxDDF5a72eRtxLOmTyYqdMeb2W7uPgvoMrNG22/bmzVbUVsnXA71ugNY6O7X5nvDst1lPvE9UX81s12Jp2HNqavs/ihvfBgC3GVmO8dbA+9yUPqNPzCwOx2XMbOtiNFGn8tO/J8SD9h6R46kOtjMNsrjrV8fc7XUgLPZ4TPErY9/NLPdAbLtcwviubujLW7HPZ0Yt3h3nqEGLHdfRDzvYanCV9rJ3f/s7l8aiOFrcft01XPEHZ1HZa33f4hvkXkJODSb/H7d6fnsi5bbgM1sGNHJ9imiBnwG2dxAPHN3ETEw/KL8/Q3d/TfqlBKRV5K13R8Sjxh9zN2vzveHEnfVTiZuL38wOyWHuvuzpea3t2rphDOz0e7+3/l6AtEwPpgI4Cfc/dHq3SoiIqsi7+67kbhzdRIxzvkLxI1Uz5vZLOKGm7Pd/Zvl5rRvammCqIRvl7v/iGjzXUJ8CeI6+TsKXxHpFXd/inic5k5Ex/adxFCzL5vZROKW/yvpw0O7+oNaR0FUbv39MRHCQ4k2mXXrLEdEBr5KH9EcojNtJPAbYuzz48D7iQdt3Z1B/arT1hsxzGwcLAtkEZFeyRAeQtxLsAVRE36fu9+Ww16f8Vfxc407+jQ0EZG+yL6l+4GPuvs/lp6futTaBCEi0g7ZtzQHGGRma5Sen7oogEXk1eJB8g7agUJNECLyqmED7BtdFMAiIoWoCUJEpBAFsIhIIQpgEZFCFMAiIoUogEVEClEAi4gU8n+zV4cFpYRKHQAAAABJRU5ErkJggg==",
|
|
"text/plain": [
|
|
"<Figure size 432x288 with 1 Axes>"
|
|
]
|
|
},
|
|
"metadata": {
|
|
"needs_background": "light"
|
|
},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"plot_value_array(1, predictions_single[0], test_labels)\n",
|
|
"_ = plt.xticks(range(10), class_names, rotation=45)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "cU1Y2OAMCaXb"
|
|
},
|
|
"source": [
|
|
"`tf.keras.Model.predict` returns a list of lists—one list for each image in the batch of data. Grab the predictions for our (only) image in the batch:"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 31,
|
|
"metadata": {
|
|
"id": "2tRmdq_8CaXb",
|
|
"tags": []
|
|
},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"2"
|
|
]
|
|
},
|
|
"execution_count": 31,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
}
|
|
],
|
|
"source": [
|
|
"np.argmax(predictions_single[0])"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "markdown",
|
|
"metadata": {
|
|
"id": "YFc2HbEVCaXd"
|
|
},
|
|
"source": [
|
|
"And the model predicts a label as expected."
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"colab": {
|
|
"collapsed_sections": [],
|
|
"name": "classification.ipynb",
|
|
"toc_visible": true
|
|
},
|
|
"kernelspec": {
|
|
"display_name": "Python 3",
|
|
"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": "3.9.9"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 4
|
|
}
|