Number Classification with TensorFlow

# @title Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

Number Classification with TensorFlow#

Copyright 2019 The TensorFlow Authors.

View on TensorFlow.org Run in Google Colab View source on GitHub Download notebook

This short introduction uses Keras to:

  1. Load a prebuilt dataset.

  2. Build a neural network machine learning model that classifies images.

  3. Train this neural network.

  4. Evaluate the accuracy of the model.

This tutorial is a Google Colaboratory notebook. Python programs are run directly in the browser—a great way to learn and use TensorFlow. To follow this tutorial, run the notebook in Google Colab by clicking the button at the top of this page.

  1. In Colab, connect to a Python runtime: At the top-right of the menu bar, select CONNECT.

  2. To run all the code in the notebook, select Runtime > Run all. To run the code cells one at a time, hover over each cell and select the Run cell icon.

Set up TensorFlow#

Import TensorFlow into your program to get started:

import tensorflow as tf

print("TensorFlow version:", tf.__version__)

TensorFlow version: 2.17.0

If you are following along in your own development environment, rather than Colab, see the install guide for setting up TensorFlow for development.

Note: Make sure you have upgraded to the latest pip to install the TensorFlow 2 package if you are using your own development environment. See the install guide for details.

Load a dataset#

Load and prepare the MNIST dataset. The pixel values of the images range from 0 through 255. Scale these values to a range of 0 to 1 by dividing the values by 255.0. This also converts the sample data from integers to floating-point numbers:

mnist = tf.keras.datasets.mnist

(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0

import matplotlib.pyplot as plt

%matplotlib inline

num_row = 2
num_col = 5
num = num_row * num_col
images = x_train[:num]
labels = y_train[:num]

# plot images
fig, axes = plt.subplots(num_row, num_col, figsize=(1.5 * num_col, 2 * num_row))
for i in range(num):
    ax = axes[i // num_col, i % num_col]
    ax.imshow(images[i], cmap="gray")
    ax.set_title("Label: {}".format(labels[i]))
plt.tight_layout()
plt.show()
../_images/eab9f54f406b1ab00ba677c65454367af7519b5a2d2ae6d75865b1e9ed0246f0.png

Build a machine learning model#

Build a tf.keras.Sequential model:

model = tf.keras.models.Sequential(
    [
        tf.keras.layers.Flatten(input_shape=(28, 28)),
        tf.keras.layers.Dense(128, activation="relu"),
        tf.keras.layers.Dropout(0.2),
        tf.keras.layers.Dense(10),
    ]
)

/Users/ariefrahmansyah/Library/Caches/pypoetry/virtualenvs/applied-python-training-MLD32oJZ-py3.12/lib/python3.12/site-packages/keras/src/layers/reshaping/flatten.py:37: 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.
  super().__init__(**kwargs)

Sequential is useful for stacking layers where each layer has one input tensor and one output tensor. Layers are functions with a known mathematical structure that can be reused and have trainable variables. Most TensorFlow models are composed of layers. This model uses the Flatten, Dense, and Dropout layers.

For each example, the model returns a vector of logits or log-odds scores, one for each class.

predictions = model(x_train[:1]).numpy()
predictions

array([[-0.3805407 ,  0.80421233, -0.31978986, -0.670506  ,  0.06069487,
        -0.679445  ,  0.40332592,  0.47335294, -0.30201632,  0.285113  ]],
      dtype=float32)

The tf.nn.softmax function converts these logits to probabilities for each class:

tf.nn.softmax(predictions).numpy()

array([[0.06272263, 0.20509543, 0.06665121, 0.04693469, 0.09751029,
        0.04651701, 0.13735777, 0.1473213 , 0.06784642, 0.12204331]],
      dtype=float32)

Note: It is possible to bake the tf.nn.softmax function into the activation function for the last layer of the network. While this can make the model output more directly interpretable, this approach is discouraged as it’s impossible to provide an exact and numerically stable loss calculation for all models when using a softmax output.

Define a loss function for training using losses.SparseCategoricalCrossentropy:

loss_fn = tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True)

The loss function takes a vector of ground truth values and a vector of logits and returns a scalar loss for each example. This loss is equal to the negative log probability of the true class: The loss is zero if the model is sure of the correct class.

This untrained model gives probabilities close to random (1/10 for each class), so the initial loss should be close to -tf.math.log(1/10) ~= 2.3.

loss_fn(y_train[:1], predictions).numpy()

3.0679374

Before you start training, configure and compile the model using Keras Model.compile. Set the optimizer class to adam, set the loss to the loss_fn function you defined earlier, and specify a metric to be evaluated for the model by setting the metrics parameter to accuracy.

model.compile(optimizer="adam", loss=loss_fn, metrics=["accuracy"])

Train and evaluate your model#

Use the Model.fit method to adjust your model parameters and minimize the loss:

model.fit(x_train, y_train, epochs=5)

Epoch 1/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 995us/step - accuracy: 0.8607 - loss: 0.4769
Epoch 2/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9549 - loss: 0.1528
Epoch 3/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 991us/step - accuracy: 0.9666 - loss: 0.1097
Epoch 4/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 985us/step - accuracy: 0.9727 - loss: 0.0898
Epoch 5/5
1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9771 - loss: 0.0764

<keras.src.callbacks.history.History at 0x17d8168d0>


 314/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.7341 - loss: 0.9244


 356/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7463 - loss: 0.8834


 398/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7564 - loss: 0.8490


 438/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7648 - loss: 0.8204


 480/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7727 - loss: 0.7937


 523/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7797 - loss: 0.7697


 564/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7857 - loss: 0.7490


 604/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7911 - loss: 0.7305


 646/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.7962 - loss: 0.7128


 689/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8009 - loss: 0.6964


 730/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8050 - loss: 0.6820


 772/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8089 - loss: 0.6682


 813/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8125 - loss: 0.6557


 855/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8159 - loss: 0.6437


 897/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8191 - loss: 0.6325


 938/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8221 - loss: 0.6222


 978/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8248 - loss: 0.6127


1019/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8274 - loss: 0.6035


1059/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.8298 - loss: 0.5950


1100/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8321 - loss: 0.5868


1141/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8344 - loss: 0.5789


1181/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8364 - loss: 0.5716


1221/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8384 - loss: 0.5647


1259/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8402 - loss: 0.5583


1298/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8420 - loss: 0.5521


1336/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8437 - loss: 0.5462


1375/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8453 - loss: 0.5404


1414/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8469 - loss: 0.5348


1453/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8484 - loss: 0.5294


1491/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8499 - loss: 0.5243


1532/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8514 - loss: 0.5190


1573/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8528 - loss: 0.5139


1615/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8543 - loss: 0.5089


1655/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8556 - loss: 0.5042


1695/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8569 - loss: 0.4997


1736/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8581 - loss: 0.4953


1777/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8594 - loss: 0.4910


1816/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8605 - loss: 0.4870


1857/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.8616 - loss: 0.4829


1875/1875 ━━━━━━━━━━━━━━━━━━━━ 4s 1ms/step - accuracy: 0.8622 - loss: 0.4811

Epoch 2/5

   1/1875 ━━━━━━━━━━━━━━━━━━━━ 1:30 48ms/step - accuracy: 0.9375 - loss: 0.3879


  42/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9523 - loss: 0.1747   


  82/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9543 - loss: 0.1602


 122/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9551 - loss: 0.1561


 165/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9552 - loss: 0.1541


 207/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9549 - loss: 0.1536


 249/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9544 - loss: 0.1538


 292/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9539 - loss: 0.1547


 334/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9536 - loss: 0.1552


 376/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9534 - loss: 0.1557


 417/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9532 - loss: 0.1562


 458/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1567


 500/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9530 - loss: 0.1570


 542/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9530 - loss: 0.1572


 584/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9530 - loss: 0.1573


 627/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1573


 669/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1573


 711/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1575


 753/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1576


 796/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9531 - loss: 0.1576


 838/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9532 - loss: 0.1576


 881/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9532 - loss: 0.1576


 924/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9533 - loss: 0.1576


 967/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9534 - loss: 0.1576


1010/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9534 - loss: 0.1575


1052/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9535 - loss: 0.1574


1095/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9535 - loss: 0.1573


1137/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9536 - loss: 0.1572


1178/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9537 - loss: 0.1570


1220/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9537 - loss: 0.1568


1261/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9538 - loss: 0.1566


1304/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9539 - loss: 0.1564


1347/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9540 - loss: 0.1562


1389/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9540 - loss: 0.1561


1431/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9541 - loss: 0.1559


1473/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9542 - loss: 0.1557


1514/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9542 - loss: 0.1555


1556/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9543 - loss: 0.1553


1597/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9544 - loss: 0.1551


1638/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9544 - loss: 0.1550


1680/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9545 - loss: 0.1548


1723/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9545 - loss: 0.1546


1764/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9546 - loss: 0.1544


1806/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9546 - loss: 0.1542


1847/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9547 - loss: 0.1541


1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9547 - loss: 0.1539

Epoch 3/5

   1/1875 ━━━━━━━━━━━━━━━━━━━━ 1:26 46ms/step - accuracy: 0.9688 - loss: 0.2087


  41/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9718 - loss: 0.1163   


  80/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9684 - loss: 0.1186


 122/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9669 - loss: 0.1197


 166/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9667 - loss: 0.1187


 210/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9669 - loss: 0.1177


 254/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9671 - loss: 0.1166


 299/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9674 - loss: 0.1157


 344/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9675 - loss: 0.1152


 388/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9676 - loss: 0.1145


 431/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9676 - loss: 0.1141


 471/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1138


 514/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9678 - loss: 0.1135


 558/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9678 - loss: 0.1131


 601/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9678 - loss: 0.1129


 644/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9678 - loss: 0.1127


 686/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9678 - loss: 0.1126


 727/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1125


 768/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1125


 811/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1125


 853/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1124


 896/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1122


 938/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1121


 980/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1120


1023/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9677 - loss: 0.1119


1066/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1117


1110/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1116


1154/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1115


1198/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1115


1241/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1114


1285/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1113


1327/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1112


1370/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1112


1414/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9677 - loss: 0.1111


1458/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1111


1501/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1110


1543/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1110


1585/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1109


1628/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1109


1670/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1109


1713/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1108


1756/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1108


1799/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1107


1843/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9676 - loss: 0.1107


1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9676 - loss: 0.1107

Epoch 4/5

   1/1875 ━━━━━━━━━━━━━━━━━━━━ 1:26 46ms/step - accuracy: 0.9375 - loss: 0.1125


  42/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9758 - loss: 0.0701   


  85/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9752 - loss: 0.0739


 128/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9754 - loss: 0.0742


 173/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9758 - loss: 0.0745


 217/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9758 - loss: 0.0755


 261/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9756 - loss: 0.0766


 305/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9755 - loss: 0.0774


 349/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9753 - loss: 0.0781


 393/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9751 - loss: 0.0789


 438/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9749 - loss: 0.0794


 483/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9749 - loss: 0.0798


 528/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9748 - loss: 0.0802


 572/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9747 - loss: 0.0806


 616/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9746 - loss: 0.0810


 660/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9746 - loss: 0.0813


 705/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9745 - loss: 0.0815


 749/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9745 - loss: 0.0816


 792/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9745 - loss: 0.0818


 836/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9744 - loss: 0.0820


 880/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9744 - loss: 0.0821


 924/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9744 - loss: 0.0822


 967/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9744 - loss: 0.0823


1011/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9743 - loss: 0.0824


1055/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9743 - loss: 0.0826


1097/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9742 - loss: 0.0827


1140/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9742 - loss: 0.0828


1184/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9742 - loss: 0.0830


1228/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9741 - loss: 0.0831


1272/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9741 - loss: 0.0832


1316/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9741 - loss: 0.0832


1360/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9741 - loss: 0.0833


1403/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9740 - loss: 0.0834


1447/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9740 - loss: 0.0835


1491/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9740 - loss: 0.0836


1535/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9739 - loss: 0.0837


1578/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9739 - loss: 0.0838


1621/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9739 - loss: 0.0839


1664/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9739 - loss: 0.0840


1708/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9739 - loss: 0.0840


1752/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9738 - loss: 0.0841


1796/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9738 - loss: 0.0842


1839/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9738 - loss: 0.0843


1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9738 - loss: 0.0843

Epoch 5/5

   1/1875 ━━━━━━━━━━━━━━━━━━━━ 1:25 46ms/step - accuracy: 1.0000 - loss: 0.0105


  42/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9725 - loss: 0.0800   


  84/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9727 - loss: 0.0810


 128/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9725 - loss: 0.0817


 172/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9731 - loss: 0.0814


 216/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9738 - loss: 0.0805


 260/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9745 - loss: 0.0792


 304/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9749 - loss: 0.0784


 349/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9751 - loss: 0.0778


 392/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9754 - loss: 0.0775


 435/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9755 - loss: 0.0773


 479/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9756 - loss: 0.0773


 522/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9758 - loss: 0.0772


 565/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9759 - loss: 0.0772


 609/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9760 - loss: 0.0772


 651/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9760 - loss: 0.0772


 693/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9761 - loss: 0.0772


 736/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9762 - loss: 0.0772


 778/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9762 - loss: 0.0771


 821/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9762 - loss: 0.0771


 864/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9763 - loss: 0.0771


 906/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9763 - loss: 0.0771


 950/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9763 - loss: 0.0771


 994/1875 ━━━━━━━━━━━━━━━━━━━━ 1s 1ms/step - accuracy: 0.9764 - loss: 0.0770


1037/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0770


1081/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1125/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1169/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1214/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1258/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1302/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1346/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1389/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1432/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1476/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1518/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1558/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1600/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1642/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1684/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1727/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1770/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1814/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1859/1875 ━━━━━━━━━━━━━━━━━━━━ 0s 1ms/step - accuracy: 0.9764 - loss: 0.0769


1875/1875 ━━━━━━━━━━━━━━━━━━━━ 2s 1ms/step - accuracy: 0.9764 - loss: 0.0769

<keras.src.callbacks.history.History at 0x7f0184ec7490>

The Model.evaluate method checks the model’s performance, usually on a validation set or test set.

model.evaluate(x_test, y_test, verbose=2)

313/313 - 0s - 468us/step - accuracy: 0.9753 - loss: 0.0798

[0.07982742041349411, 0.9753000140190125]

The image classifier is now trained to ~98% accuracy on this dataset. To learn more, read the TensorFlow tutorials.

If you want your model to return a probability, you can wrap the trained model, and attach the softmax to it:

probability_model = tf.keras.Sequential([model, tf.keras.layers.Softmax()])

probability_model(x_test[:5])

<tf.Tensor: shape=(5, 10), dtype=float32, numpy=
array([[1.09569520e-07, 2.81650453e-10, 2.76106198e-06, 1.09500957e-04,
        5.27111133e-11, 2.14683382e-06, 2.39292028e-12, 9.99881864e-01,
        4.06290326e-07, 3.12417910e-06],
       [9.36238642e-10, 9.94658796e-04, 9.99004304e-01, 9.69198368e-07,
        1.19458539e-14, 1.32637403e-08, 3.25292042e-08, 3.50566516e-12,
        6.19934113e-08, 1.05918810e-14],
       [2.11339966e-06, 9.99089956e-01, 2.31639220e-04, 9.04002718e-06,
        1.27908625e-05, 6.04543357e-06, 1.57848790e-05, 5.41534158e-04,
        8.97237696e-05, 1.44716455e-06],
       [9.99949574e-01, 6.36652331e-09, 2.91350098e-05, 6.50657412e-08,
        3.27179244e-08, 2.34230447e-06, 7.51126618e-06, 9.89659202e-06,
        7.71864705e-08, 1.30633589e-06],
       [4.48162268e-07, 7.70599016e-08, 1.51247777e-05, 3.49523965e-08,
        9.97868180e-01, 1.17272137e-07, 2.39243946e-06, 6.84397091e-05,
        7.22275331e-07, 2.04430916e-03]], dtype=float32)>

Conclusion#

Congratulations! You have trained a machine learning model using a prebuilt dataset using the Keras API.

For more examples of using Keras, check out the tutorials. To learn more about building models with Keras, read the guides. If you want learn more about loading and preparing data, see the tutorials on image data loading or CSV data loading.

Contents