A Recurrent Neural Network (RNN) with Long Short Term Memory (LSTM) cells learns how to translate from English to French using only a small subset of a massive 2GB dataset. This machine learning architecture makes connections along a time sequence, so it remembers why certain words come before or after others.
Simply run the Jupyter Notebook dlnd_language_translation.ipynb or you can run the script language_translation.py
python language_translation.py
To translate a sentence, assign the variable translate_sentence a string.
translate_sentence = 'he saw a old yellow truck .'
You can install the required packages through Anaconda's environment manager using the machine-learning.yml file
conda env create -f machine-learning.yml
Then, activate the environment and run language_translation.py
activate machine-learning
Otherwise, check out the machine-learning.yml file for dependencies and their versions
Simply add test cases to problem_unittests.py or run it
python problem_unittests.py
- TensorFlow - The machine learning framework
- Anaconda - The environment manager
- Jupyter Notebook - The code documentation