In this section, we will show you how to implement a relatively simple CNN architecture. We will also look at how to train it to classify the CIFAR-10 dataset.

Start by importing all the necessary libraries:

import fire 
import numpy as np 
import os 
import tensorflow as tf 
from tf.keras.datasets import cifar10 

We will define a Python class that will implement the training process. The class name is Train, and it implements two methods: build_graph and train. The train function is fired when the main program is executed:

class Train:  

   __x_ = []
     __y_ = []
     __logits = []
     __loss = []
     __train_step = []
     __merged_summary_op = []
     __saver = []
     __session = []
     __writer = []
     __is_training = []
     __loss_val = []
     __train_summary = []
     __val_summary = []

 

   def __init__(self):
         pass 

   def build_graph(self): 

   [...] 

   def train(self, save_dir='./save', batch_size=500): 

[...] 

if __name__ == '__main__': 

   cnn= Train() 

   cnn.train