How to Train a Word2Vec Model

Question: How do we train a Word2Vec model? Key points: Word2Vec learns dense word embeddings by predicting context from a target word (Skip-gram) or predicting a target word from its context (CBOW), using a shallow neural network trained via negative sampling or hierarchical softmax for efficiency over large vocabularies. The resulting embedding vectors capture semantic/syntactic similarity such that similar words end up close together in the embedding space.