Deep Learning

How much DEEP learning do we need? Word2Vec vs. CNN vs. LSTM vs. BERT for Sentiment Analysis [experiment]

Photo by Jesse Martini on Unsplash

The growing trend of large pre-trained networks to push for better accuracy in Natural Language Processing (NLP) concerns me. Surely they are great, but do they worth the resources we are using on training them? Looking at their number of parameters, from the…