Describe your perfect cup of coffee. We'll classify your preferences into one of six roast classes to help you select the best bag of beans from your favorite coffee shop.
CoffeeMate is an NLP classification project that aligns free-text coffee preferences with six categories of roast levels. The inference model was built and tested using Python 3, Tensorflow, Keras, and Google Colab.
Users provide a description of their favorite cup of coffee - this may include preferred flavors, strength, or origin. The neural network encodes this text into a set of high dimensional vectors that are passed into the model. The output layer is a 1D vector of categorical probabilities for 6 classes of coffee, light to dark!
