A novel approach to the classification of antimicrobial peptides (AMPs) using pre-trained language models to create contextual vectorized embeddings of each peptide sequence before a convolutional neural network is used as the classifier.