Example layout and directory flow to set up tensorflow model for training and inference. Example uses image captioning with attention. Purpose of repo to set up a boiler plate for future tensorflow models.
Dev branch contains a implementation which can be deployed to a web server.
pip install -r requirements.txt
python app/server.py serve
Master branch contains the main components for setting up the training process.
conda env create -f ImageCaptioningWithAttention.yml
conda activate ImageCaptioningWithAttention
jupyter notebook
Main notebook for training: ImageCaptioningWithAttention.ipynb