Creating a custom TensorFlow Lite model for image recognition

Once you have had a fair go at Colaboratory, we're all set up to build the custom TensorFlow Lite model for the task of recognizing plant species. To do so, we will begin with a new Colaboratory notebook and perform the following steps:

1. Import the necessary modules for the project. Firstly, we import TensorFlow and NumPy. NumPy will be useful for handling the image arrays, and TensorFlow will be used to build the CNN. The code to import the modules can be seen in the following snippet:

!pip install tf-nightly-gpu-2.0-preview
import tensorflow as tf
import numpy as np
import os

Notice the !pip install <package-name> command used on the first line. This is used to install packages in a running Colaboratory notebook, which, in this case, installs the latest TensorFlow release that internally implements the...