Make sure you have the Python 3.7.6 installed IDE : Anaconda(Jupyter and Spyder),Pycharm.

• Create a virtual enviornment: This is a good practice but not necessary

  • Goto Anaconda Prompt and type in the below command. This command will create a virtual enviornment. conda create -y --name tensorflow python=3.7
  • Now activate the virtual enviornment, by executing the below command. This command will change the prompt from "base" to "tensorflow" activate tensorflow

• In the Terminal type the following commands to install the libraries,this will check the exsisiting version and will install the latest one.

  • conda install -y scipy
  • pip install --exists-action i --upgrade tensorflow==2.0.0
  • pip install --exists-action i --upgrade keras
  • pip install --exists-action i --upgrade nltk
  • pip install --exists-action i --upgrade numpy
  • pip install --exists-action i --upgrade pandas
  • pip install --exists-action i --upgrade pickle
  • pip install --exists-action i --upgrade

• Check the version installed using below commands:

  • tf.version (where tf is the alias of tensorflow)
  • k.version (where k is the alias of keras)
  • sys.version
  • pd.version (where pd is the alias of pandas)

• Select the virtual Environment:

  • In Anaconda Navigator --> Home, Click on the "Applications on" dropdown and Select "tensorflow".
  • This will take a few minutes, after that install the prefered IDE (Jupyter, Spyder),installation is required for first time only
  • Launch the IDE and start coding.

• In IDE Pycharm : Pycharm uses a virtual enviornment to run tensorflow, hence install the tensorflow to the specific virtual enviornment.

  • Project Setting --> Project Interpreter --> Click on the "+" icon at the bottom left of the package window --> type in tensorflow and specify the version
  • Click on Install Package

tensorboard 2.0.0 has requirement setuptools>=41.0.0, but you'll have setuptools 40.8.0 which is incompatible.

This is a possible error which can be solved using the command

pip install --upgrade --user tensorflow

Open Source Software

Pycharm, Anaconda Jupyter

• for creating model. from keras.models import Sequential
• for creating layers for the model. from keras.layers import Dense, Activation, Dropout
• for selecting optimiser for minising loss and hence increasing accuracy. from keras.optimizers import Adam,SGD
• for chatbot GUI. from tkinter import *
• for displaying date and time. import time import datetime
• for processing natural language. from nltk import word_tokenize
• wordNet is a collection of nouns and verbs and adjectives and their synonyms from nltk.stem import WordNetLemmatizer
• for create light weight binary files. import pickle
• for importing the model created. from keras.models import load_model
• for formating the Intents.json file. import json
• for selecting a random "response" from the slected "tag" import random
• for converting the bag-of-words (bag) from customer input and feeding it to the model. import numpy as np
• for making database related operations(insert, select) import sqlite3
• for matching words from customer for unlimited data. from fuzzywuzzy import fuzz
• for processing the images used in the aplication. from PIL import Image from PIL import ImageTk

##General Instructions

• To download the corpus i.e nltk_data either :

1. Run the below command in the terminal window of PyCharm or on the command prompt. python -m nltk.downloader all
2. In your .py file add 2 line: import nltk nltk.download() A pop-up window will apper and click on download --> it will take some time, Once all the collections turn green close the window.

• Make sure proper libraries are imported while executing and please follow best coding practices.

Happy Coding!