We build a model that can be prompted to generate human like positive and negative medical reviews in German. For that, we fine-tune GPT-2 on an unique data set using Tensorflow in Transformers. For fast results, we use TPUs on Google Colab.

Using a unique German data set containing ratings and comments on doctors, we build a Binary Text Classifier. In this third part, we introduce a state of the art model based on the transformer architecture, namely BERT. Using the transformers library for tensorflow, we push our ability to predict the sentiment of comments towards its limit.

Using a unique German data set containing ratings and comments on doctors, we build a Binary Text Classifier. In part 1 we've introduced a complete machine learning work flow that predicts ratings from comments. In this second part, we improve on our baseline by implementing a LSTM neural network model and using FastText embeddings. Using Keras for feature creation and prediction, we improve on the ability to understand comment sentiments.