/TensorFlow Guide

Get Started with Eager Execution

Run in Google Colab View source on GitHub

This tutorial describes how to use machine learning to categorize Iris flowers by species. It uses TensorFlow's eager execution to (1) build a model, (2) train the model on example data, and (3) use the model to make predictions on unknown data. Machine learning experience isn't required to follow this guide, but you'll need to read some Python code.

TensorFlow programming

There many TensorFlow APIs available, but we recommend starting with these high-level TensorFlow concepts:

This tutorial shows these APIs and is structured like many other TensorFlow programs:

  1. Import and parse the data sets.
  2. Select the type of model.
  3. Train the model.
  4. Evaluate the model's effectiveness.
  5. Use the trained model to make predictions.

To learn more about using TensorFlow, see the Getting Started guide and the example tutorials. If you'd like to learn about the basics of machine learning, consider taking the Machine Learning Crash Course.

Run the notebook

This tutorial is available as an interactive Colab notebook for you to run and change the Python code directly in the browser. The notebook handles setup and dependencies while you "play" cells to execute the code blocks. This is a fun way to explore the program and test ideas. If you are unfamiliar with Python notebook environments, there are a couple of things to keep in mind:

  1. Executing code requires connecting to a runtime environment. In the Colab notebook menu, select Runtime > Connect to runtime...
  2. Notebook cells are arranged sequentially to gradually build the program. Typically, later code cells depend on prior code cells, though you can always rerun a code block. To execute the entire notebook in order, select Runtime > Run all. To rerun a code cell, select the cell and click the play icon on the left.

Setup program

Install the latest version of TensorFlow

This tutorial uses eager execution, which is available in TensorFlow 1.8. (You may need to restart the runtime after upgrading.)

!pip install -q --upgrade tensorflow

Configure imports and eager execution

Import the required Python modules, including TensorFlow, and enable eager execution for this program. Eager execution makes TensorFlow evaluate operations immediately, returning concrete values instead of creating a computational graph that is executed later. If you are used to a REPL or the python interactive console, you'll feel at home.

Once eager execution is enabled, it cannot be disabled within the same program. See the eager execution guide for more details.

from __future__ import absolute_import, division, print_function

import os
import matplotlib.pyplot as plt

import tensorflow as tf
import tensorflow.contrib.eager as tfe


print("TensorFlow version: {}".format(tf.VERSION))
print("Eager execution: {}".format(tf.executing_eagerly()))
/usr/local/lib/python3.5/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
TensorFlow version: 1.8.0
Eager execution: True

The Iris classification problem

Imagine you are a botanist seeking an automated way to categorize each Iris flower you find. Machine learning provides many algorithms to statistically classify flowers. For instance, a sophisticated machine learning program could classify flowers based on photographs. Our ambitions are more modest—we're going to classify Iris flowers based on the length and width measurements of their sepals and petals.

The Iris genus entails about 300 species, but our program will classify only the following three:

  • Iris setosa
  • Iris virginica
  • Iris versicolor
Petal geometry compared for three iris species: Iris setosa, Iris virginica, and Iris
Figure 1. Iris setosa (by Radomil, CC BY-SA 3.0), Iris versicolor, (by Dlanglois, CC BY-SA 3.0), and Iris virginica (by Frank Mayfield, CC BY-SA 2.0).

Fortunately, someone has already created a data set of 120 Iris flowers with the sepal and petal measurements. This is a classic dataset that is popular for beginner machine learning classification problems.

Import and parse the training dataset

We need to download the dataset file and convert it to a structure that can be used by this Python program.

Download the dataset

Download the training dataset file using the tf.keras.utils.get_file function. This returns the file path of the downloaded file.

train_dataset_url = "http://download.tensorflow.org/data/iris_training.csv"

train_dataset_fp = tf.keras.utils.get_file(fname=os.path.basename(train_dataset_url),

print("Local copy of the dataset file: {}".format(train_dataset_fp))
Downloading data from http://download.tensorflow.org/data/iris_training.csv
8192/2194 [================================================================================================================] - 0s 0us/step
Local copy of the dataset file: /root/.keras/datasets/iris_training.csv

Inspect the data

This dataset, iris_training.csv, is a plain text file that stores tabular data formatted as comma- separated values (CSV). Use the head -n5 command to take a peak at the first five entries:

!head -n5 {train_dataset_fp}

From this view of the dataset, we see the following:

  1. The first line is a header containing information about the dataset:
  2. There are 120 total examples. Each example has four features and one of three possible label names.
  3. Subsequent rows are data records, one example per line, where:
  4. The first four fields are features: these are characteristics of an example. Here, the fields hold float numbers representing flower measurements.
  5. The last column is the label: this is the value we want to predict. For this dataset, it's an integer value of 0, 1, or 2 that corresponds to a flower name.

Each label is associated with string name (for example, "setosa"), but machine learning typically relies on numeric values. The label numbers are mapped to a named representation, such as:

  • 0: Iris setosa
  • 1: Iris versicolor
  • 2: Iris virginica

For more information about features and labels, see the ML Terminology section of the Machine Learning Crash Course.

Parse the dataset

Since our dataset is a CSV-formatted text file, we'll parse the feature and label values into a format our Python model can use. Each line—or row—in the file is passed to the parse_csv function which grabs the first four feature fields and combines them into a single tensor. Then, the last field is parsed as the label. The function returns both the features and label tensors:

def parse_csv(line):
  example_defaults = [[0.], [0.], [0.], [0.], [0]]  # sets field types
  parsed_line = tf.decode_csv(line, example_defaults)
  # First 4 fields are features, combine into single tensor
  features = tf.reshape(parsed_line[:-1], shape=(4,))
  # Last field is the label
  label = tf.reshape(parsed_line[-1], shape=())
  return features, label

Create the training tf.data.Dataset

TensorFlow's Dataset API handles many common cases for feeding data into a model. This is a high-level API for reading data and transforming it into a form used for training. See the Datasets Quick Start guide for more information.

This program uses tf.data.TextLineDataset to load a CSV-formatted text file and is parsed with our parse_csv function. A tf.data.Dataset represents an input pipeline as a collection of elements and a series of transformations that act on those elements. Transformation methods are chained together or called sequentially—just make sure to keep a reference to the returned Dataset object.

Training works best if the examples are in random order. Use tf.data.Dataset.shuffle to randomize entries, setting buffer_size to a value larger than the number of examples (120 in this case). To train the model faster, the dataset's batch size is set to 32 examples to train at once.

train_dataset = tf.data.TextLineDataset(train_dataset_fp)
train_dataset = train_dataset.skip(1)             # skip the first header row
train_dataset = train_dataset.map(parse_csv)      # parse each row
train_dataset = train_dataset.shuffle(buffer_size=1000)  # randomize
train_dataset = train_dataset.batch(32)

# View a single example entry from a batch
features, label = iter(train_dataset).next()
print("example features:", features[0])
print("example label:", label[0])
example features: tf.Tensor([6.  2.7 5.1 1.6], shape=(4,), dtype=float32)
example label: tf.Tensor(1, shape=(), dtype=int32)

Select the type of model

Why model?

A model is the relationship between features and the label. For the Iris classification problem, the model defines the relationship between the sepal and petal measurements and the predicted Iris species. Some simple models can be described with a few lines of algebra, but complex machine learning models have a large number of parameters that are difficult to summarize.

Could you determine the relationship between the four features and the Iris species without using machine learning? That is, could you use traditional programming techniques (for example, a lot of conditional statements) to create a model? Perhaps—if you analyzed the dataset long enough to determine the relationships between petal and sepal measurements to a particular species. And this becomes difficult—maybe impossible—on more complicated datasets. A good machine learning approach determines the model for you. If you feed enough representative examples into the right machine learning model type, the program will figure out the relationships for you.

Select the model

We need to select the kind of model to train. There are many types of models and picking a good one takes experience. This tutorial uses a neural network to solve the Iris classification problem. Neural networks can find complex relationships between features and the label. It is a highly-structured graph, organized into one or more hidden layers. Each hidden layer consists of one or more neurons. There are several categories of neural networks and this program uses a dense, or fully-connected neural network: the neurons in one layer receive input connections from every neuron in the previous layer. For example, Figure 2 illustrates a dense neural network consisting of an input layer, two hidden layers, and an output layer:

A diagram of the network architecture: Inputs, 2 hidden layers, and outputs
Figure 2. A neural network with features, hidden layers, and predictions.

When the model from Figure 2 is trained and fed an unlabeled example, it yields three predictions: the likelihood that this flower is the given Iris species. This prediction is called inference. For this example, the sum of the output predictions are 1.0. In Figure 2, this prediction breaks down as: 0.03 for Iris setosa, 0.95 for Iris versicolor, and 0.02 for Iris virginica. This means that the model predicts—with 95% probability—that an unlabeled example flower is an Iris versicolor.

Create a model using Keras

The TensorFlow tf.keras API is the preferred way to create models and layers. This makes it easy to build models and experiment while Keras handles the complexity of connecting everything together. See the Keras documentation for details.

The tf.keras.Sequential model is a linear stack of layers. Its constructor takes a list of layer instances, in this case, two Dense layers with 10 nodes each, and an output layer with 3 nodes representing our label predictions. The first layer's input_shape parameter corresponds to the amount of features from the dataset, and is required.

model = tf.keras.Sequential([
  tf.keras.layers.Dense(10, activation="relu", input_shape=(4,)),  # input shape required
  tf.keras.layers.Dense(10, activation="relu"),

The activation function determines the output of a single neuron to the next layer. This is loosely based on how brain neurons are connected. There are many available activations, but ReLU is common for hidden layers.

The ideal number of hidden layers and neurons depends on the problem and the dataset. Like many aspects of machine learning, picking the best shape of the neural network requires a mixture of knowledge and experimentation. As a rule of thumb, increasing the number of hidden layers and neurons typically creates a more powerful model, which requires more data to train effectively.

Train the model

Training is the stage of machine learning when the model is gradually optimized, or the model learns the dataset. The goal is to learn enough about the structure of the training dataset to make predictions about unseen data. If you learn too much about the training dataset, then the predictions only work for the data it has seen and will not be generalizable. This problem is called overfitting—it's like memorizing the answers instead of understanding how to solve a problem.

The Iris classification problem is an example of supervised machine learning: the model is trained from examples that contain labels. In unsupervised machine learning, the examples don't contain labels. Instead, the model typically finds patterns among the features.

Define the loss and gradient function

Both training and evaluation stages need to calculate the model's loss. This measures how off a model's predictions are from the desired label, in other words, how bad the model is performing. We want to minimize, or optimize, this value.

Our model will calculate its loss using the tf.losses.sparse_softmax_cross_entropy function which takes the model's prediction and the desired label. The returned loss value is progressively larger as the prediction gets worse.

def loss(model, x, y):
  y_ = model(x)
  return tf.losses.sparse_softmax_cross_entropy(labels=y, logits=y_)

def grad(model, inputs, targets):
  with tf.GradientTape() as tape:
    loss_value = loss(model, inputs, targets)
  return tape.gradient(loss_value, model.variables)

The grad function uses the loss function and the tf.GradientTape to record operations that compute the gradients used to optimize our model. For more examples of this, see the eager execution guide.

Create an optimizer

An optimizer applies the computed gradients to the model's variables to minimize the loss function. You can think of a curved surface (see Figure 3) and we want to find its lowest point by walking around. The gradients point in the direction of steepest ascent—so we'll travel the opposite way and move down the hill. By iteratively calculating the loss and gradient for each batch, we'll adjust the model during training. Gradually, the model will find the best combination of weights and bias to minimize loss. And the lower the loss, the better the model's predictions.

Optimization algorthims visualized over time in 3D space.
Figure 3. Optimization algorthims visualized over time in 3D space. (Source: Stanford class CS231n, MIT License)

TensorFlow has many optimization algorithms available for training. This model uses the tf.train.GradientDescentOptimizer that implements the stochastic gradient descent (SGD) algorithm. The learning_rate sets the step size to take for each iteration down the hill. This is a hyperparameter that you'll commonly adjust to achieve better results.

optimizer = tf.train.GradientDescentOptimizer(learning_rate=0.01)

Training loop

With all the pieces in place, the model is ready for training! A training loop feeds the dataset examples into the model to help it make better predictions. The following code block sets up these training steps:

  1. Iterate each epoch. An epoch is one pass through the dataset.
  2. Within an epoch, iterate over each example in the training Dataset grabbing its features (x) and label (y).
  3. Using the example's features, make a prediction and compare it with the label. Measure the inaccuracy of the prediction and use that to calculate the model's loss and gradients.
  4. Use an optimizer to update the model's variables.
  5. Keep track of some stats for visualization.
  6. Repeat for each epoch.

The num_epochs variable is the amount of times to loop over the dataset collection. Counter- intuitively, training a model longer does not guarantee a better model. num_epochs is a hyperparameter that you can tune. Choosing the right number usually requires both experience and experimentation.

## Note: Rerunning this cell uses the same model variables

# keep results for plotting
train_loss_results = []
train_accuracy_results = []

num_epochs = 201

for epoch in range(num_epochs):
  epoch_loss_avg = tfe.metrics.Mean()
  epoch_accuracy = tfe.metrics.Accuracy()

  # Training loop - using batches of 32
  for x, y in train_dataset:
    # Optimize the model
    grads = grad(model, x, y)
    optimizer.apply_gradients(zip(grads, model.variables),

    # Track progress
    epoch_loss_avg(loss(model, x, y))  # add current batch loss
    # compare predicted label to actual label
    epoch_accuracy(tf.argmax(model(x), axis=1, output_type=tf.int32), y)

  # end epoch

  if epoch % 50 == 0:
    print("Epoch {:03d}: Loss: {:.3f}, Accuracy: {:.3%}".format(epoch,
Epoch 000: Loss: 1.271, Accuracy: 35.000%
Epoch 050: Loss: 0.507, Accuracy: 79.167%
Epoch 100: Loss: 0.297, Accuracy: 94.167%
Epoch 150: Loss: 0.197, Accuracy: 96.667%
Epoch 200: Loss: 0.136, Accuracy: 99.167%

Visualize the loss function over time

While it's helpful to print out the model's training progress, it's often more helpful to see this progress. TensorBoard is a nice visualization tool that is packaged with TensorFlow, but we can create basic charts using the mathplotlib module.

Interpreting these charts takes some experience, but you really want to see the loss go down and the accuracy go up.

fig, axes = plt.subplots(2, sharex=True, figsize=(12, 8))
fig.suptitle('Training Metrics')

axes[0].set_ylabel("Loss", fontsize=14)

axes[1].set_ylabel("Accuracy", fontsize=14)
axes[1].set_xlabel("Epoch", fontsize=14)



Evaluate the model's effectiveness

Now that the model is trained, we can get some statistics on its performance.

Evaluating means determining how effectively the model makes predictions. To determine the model's effectiveness at Iris classification, pass some sepal and petal measurements to the model and ask the model to predict what Iris species they represent. Then compare the model's prediction against the actual label. For example, a model that picked the correct species on half the input examples has an accuracy of 0.5. Figure 4 shows a slightly more effective model, getting 4 out of 5 predictions correct at 80% accuracy:

Example features Label Model prediction
5.9 3.0 4.3 1.5 1 1
6.9 3.1 5.4 2.1 2 2
5.1 3.3 1.7 0.5 0 0
6.0 3.4 4.5 1.6 1 2
5.5 2.5 4.0 1.3 1 1
Figure 4. An Iris classifier that is 80% accurate.

Setup the test dataset

Evaluating the model is similar to training the model. The biggest difference is the examples come from a separate test set rather than the training set. To fairly assess a model's effectiveness, the examples used to evaluate a model must be different from the examples used to train the model.

The setup for the test Dataset is similar to the setup for training Dataset. Download the CSV text file and parse that values, then give it a little shuffle:

test_url = "http://download.tensorflow.org/data/iris_test.csv"

test_fp = tf.keras.utils.get_file(fname=os.path.basename(test_url),

test_dataset = tf.data.TextLineDataset(test_fp)
test_dataset = test_dataset.skip(1)             # skip header row
test_dataset = test_dataset.map(parse_csv)      # parse each row with the funcition created earlier
test_dataset = test_dataset.shuffle(1000)       # randomize
test_dataset = test_dataset.batch(32)           # use the same batch size as the training set
Downloading data from http://download.tensorflow.org/data/iris_test.csv
8192/573 [============================================================================================================================================================================================================================================================================================================================================================================================================================================] - 0s 0us/step

Evaluate the model on the test dataset

Unlike the training stage, the model only evaluates a single epoch of the test data. In the following code cell, we iterate over each example in the test set and compare the model's prediction against the actual label. This is used to measure the model's accuracy across the entire test set.

test_accuracy = tfe.metrics.Accuracy()

for (x, y) in test_dataset:
  prediction = tf.argmax(model(x), axis=1, output_type=tf.int32)
  test_accuracy(prediction, y)

print("Test set accuracy: {:.3%}".format(test_accuracy.result()))
Test set accuracy: 100.000%

Use the trained model to make predictions

We've trained a model and "proven" that it's good—but not perfect—at classifying Iris species. Now let's use the trained model to make some predictions on unlabeled examples; that is, on examples that contain features but not a label.

In real-life, the unlabeled examples could come from lots of different sources including apps, CSV files, and data feeds. For now, we're going to manually provide three unlabeled examples to predict their labels. Recall, the label numbers are mapped to a named representation as:

  • 0: Iris setosa
  • 1: Iris versicolor
  • 2: Iris virginica
class_ids = ["Iris setosa", "Iris versicolor", "Iris virginica"]

predict_dataset = tf.convert_to_tensor([
    [5.1, 3.3, 1.7, 0.5,],
    [5.9, 3.0, 4.2, 1.5,],
    [6.9, 3.1, 5.4, 2.1]

predictions = model(predict_dataset)

for i, logits in enumerate(predictions):
  class_idx = tf.argmax(logits).numpy()
  name = class_ids[class_idx]
  print("Example {} prediction: {}".format(i, name))
Example 0 prediction: Iris setosa
Example 1 prediction: Iris versicolor
Example 2 prediction: Iris virginica

These predictions look good!

To dig deeper into machine learning models, take a look at the TensorFlow Programmer's Guide and check out the community.

© 2018 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.