ibealec / aind2-nlp-capstone

Introduction

In this notebook, you will build a deep neural network that functions as part of an end-to-end machine translation pipeline. Your completed pipeline will accept English text as input and return the French translation.

Launch on EC2 GPU Instance with udacity-aind AMI

• Follow Steps 1 to 10 in instructions_launch_ec2_gpu_instance.pdf
• Follow Step 11 in instructions_launch_ec2_gpu_instance.pdf to Login with:

   ssh aind2@X.X.X.X
  

• Follow Step 12 on EC2 GPU instance. Ensure correct Python kernel version in Jupyter https://stackoverflow.com/questions/30492623/using-both-python-2-x-and-python-3-x-in-ipython-notebook

   git clone https://github.com/ltfschoen/aind2-nlp-capstone
   cd aind2-nlp-capstone
   conda create --name aind-nlp-capstone python=3.5 numpy
   source activate aind-nlp-capstone
   conda install notebook ipykernel
   ipython kernel install --user
   pip install tensorflow-gpu -U
   pip install keras -U
   KERAS_BACKEND=tensorflow python -c "from keras import backend"
   jupyter notebook --ip=0.0.0.0 --no-browser
  

  • Open in browser the URL shown in terminal and replace IP address with that in the AWS EC2 Dashboard http://<EC2_IP_address>:8888/?token=3156e..
  • Click machine_translation.ipynb
  • Within Jupyter Notebook, running !python --version returns: Python 3.5.3 :: Continuum Analytics, Inc.

Setup

Install

• Python 3
• NumPy
• TensorFlow 1.x
• Keras 2.x

Start

This project is within a Jupyter Notebook. To start the notebook, run the command jupyter notebook machine_translation.ipynb in this directory. Follow the instructions within the notebook.

Submission

When you are ready to submit your project, do the following steps:

1. Ensure you pass all points on the rubric.
2. Submit the following in a zip file:

• helper.py
• machine_translation.ipynb
• machine_translation.html - You can export the notebook by navigating to File -> Download as -> HTML (.html).