Word2vec -> for word embeddings
The RNN architecture -> Model: "sequential", 1 masking layer, 1 lstm layer, 2 dense layers, last one being binary classification layer (sigmoid).
The dataset -> TensorFlow dataset "imdb_reviews"
Goal -> I am comparing the accuracy of sentiment analysis (a movie review being positive or negative) on
reviews embedded with word2vec, and later with pre-trained embeddings from glove-wiki-gigaword-50 (transfer learning)