This repository elucidates the video and the blog where I discussed how we can create our own custom data generator for different kind of inputs like single image input, a sequence of input or multiple input to be fed to multiple model (ensemble of model).

The links of the videos are given below:

Part-1: Introduction - https://youtu.be/oy5EeamF_M8

Part-2: What is a generator function - https://youtu.be/2tRR45vcn3o

The links of the blog are given below:

Chapter -1 : What is a generator function in python and the difference between yield and return

Chapter-2: Writing a generator function to read your data that can be fed for training an image classifier in Keras.

I have used the code and concept of this blog as reference - http://www.jessicayung.com/using-generators-in-python-to-train-machine-learning-models/

To create a data genrator you can follow the note book flowers_recognition\data_generator.ipynb

I used conda environment and I have installed few libraries explicitly like keras, opencv,scikit-laern, pandas.

All the libraries may not be present in the requirements.txt file

if you are getting error in installing some libraries from requirements.txt the you can just use pip install library name without any specific version.

  pip install -r requirements.txt
  pip install library_name