How to create amazing Tensorflow model in few lines of code
--
Although TensorFlow looks very hare but it is very easy to create a machine learning model using this library in very easy and few steps. In this tutorial I will share how to create a machine learning model in minimum lines of code
- What is Tensoflow: It is open source library mainly designed to handle large scale machine learning. It has state of the arts function and capability to handle end to end machine learning tasks.
- Why are we using TensorFlow: It is one of the most popular library used for deep learning task at enterprise level. It is very easy to learn and implement
- Steps to create model : You need to follow below simple steps to create your first deep learning model in TensorFlow
- Importing the libraries: We need to import TensorFlow library in python. If TensorFlow is not installed you can use pip install tensorflow
import tensorflow as tf
- Downloading the data: In this tutorial we are going to use inbuilt data set available in tensorflow. The MNIST database of handwritten digits, has a training set of 60,000 examples, and a test set of 10,000 examples. It is a subset of a larger set available from NIST. The digits have been size-normalized and centered in a fixed-size image.
mnist = tf.keras.datasets.mnist
- Transforming the data : To check the accuracy of our model we need to divide the dataset in two parts training set and test set.
(x_train, y_train), (x_test, y_test) = mnist.load_data()
x_train, x_test = x_train / 255.0, x_test / 255.0
- Building the model : In this step we are creating the architecture of deep learning model. We are building a Sequential model with Flat and Dense layer. (You can play with the layers and see the impact on accuracy)
First layer is Flatten layer that will transform the input matrix [28*28] into flat array
Second layer will be Dense layer with ‘relu’ activation function
Third is also a dense layer. This layer also works as output layer
model = tf.keras.models.Sequential([
tf.keras.layers.Flatten(input_shape=(28, 28)),
tf.keras.layers.Dense(128, activation='relu'),
tf.keras.layers.Dense(10)
])
- Training the model : Lets train the model on the training dataset. You can select less epochs if you want to train your model fast. The accuracy may be impacted with less epochs.
model.fit(x_train, y_train, epochs=5)
- Checking the accuracy of model : Now it is time to evaluate the accuracy of your model.
model.evaluate(x_test, y_test, verbose=2)
That’s it. Kudos to you. You have created your first deep learning model