Used Recurrent Neural Networks (RNN) to build RNNs that can generate sequences based on input data - with a focus on two applications: With the first I used real market data in order to predict future Apple stock prices using an RNN model. The second one is trained on Sir Arthur Conan Doyle's classic novel Sherlock Holmes and will generate wacky sentences based on it that may - or may not - become the next great Sherlock Holmes novel!
RNN_project.ipynb
- Interactive Python Notebook.report.pdf
- Report of Python Notebook.report.html
- Report of Python Notebook.
To get this code on your machine you can fork the repo or open a terminal and run this command.
$ git clone https://github.com/JonathanKSullivan/text_generation.git
$ cd text_generation
$ jupyter notebook RNN_project.ipynb
This project requires Python 3:
- It is highly recommended that you install the Anaconda distribution of Python and load the environment included below.
- Run
git clone https://github.com/udacity/aind2-dl.git; cd aind2-dl
- Run
conda env create -f requirements/aind-dl-mac-linux.yml
- Run
source activate aind-dl
- Run
KERAS_BACKEND=tensorflow python -c "from keras import backend"
- Run
git clone https://github.com/udacity/aind2-dl.git; cd aind2-dl
- Run
conda env create -f requirements/aind-dl-windows.yml
- Run
activate aind-dl
- Run
set KERAS_BACKEND=tensorflow
- Run
python -c "from keras import backend"
- Anaconda - The data science platform used
- Udacity - Initial work - AIND-Isolation
- Jonathan Sulivan
- Hackbright Academy
- Udacity