A chatbot with the personality of Alan Turing designed to pass the Turing Test.

The software for this tutorial can be downloaded from: https://github.com/tensorflow/nmt The detailed instructions on how to use this tutorial are contained in the README.md file in that GitHub repository. This appendix lists a summary of the steps to run that tutorial. I had to run this tutorial (or parts of it) several times during the project. I wrote this summary to facilitate doing this so quickly.

Create a Python3 virtualenv. In Oracle Virtualbox, create an Ubuntu Desktop 16.04 VM, 4GB RAM / 80 GB HDD In terminal, sudo apt upgrade sudo apt-get install python3-pip sudo pip3 install virtualenv

Create a virtualenv for this project. virtualenv tensorflow cd tensorflow/ source ./bin/activate

Download code sudo apt install git git clone https://github.com/tensorflow/nmt sudo apt install curl

Download the data: In my project, I replaced this data with the Q&A dataset, keeping the same filenames. cd nmt nmt/scripts/download_iwslt15.sh /tmp/nmt_data -train.en -train.vi -tst2012.en -tst2012.vi -tst2013.en -tst2013.vi -vocab.en -vocab.vi

Install dependencies needed to run the model mkdir /tmp/nmt_model pip install numpy pip install tensorflow

Run the model Enter the following command: python -m nmt.nmt
--src=vi --tgt=en
--vocab_prefix=/tmp/nmt_data/vocab
--train_prefix=/tmp/nmt_data/train
--dev_prefix=/tmp/nmt_data/tst2012
--test_prefix=/tmp/nmt_data/tst2013
--out_dir=/tmp/nmt_model
--num_train_steps=12000
--steps_per_stats=100
--num_layers=2
--num_units=128
--dropout=0.2
—metrics=bleu

Start Tensorboard to monitor the model build Open a new terminal window. tensorboard —logdir=/tmp/nmt_model/ Open a browser: localhost:6000

To delete the model:

21. cd /tmp/nmt_model
22. rm -r *

To ask the model questions:

23. Open new terminal window
24. Navigate to /tensorflow/nmt
25. Copy askAlan.py into this directory
26. python askAlan.py