- are based on counting of words and
- are precursor to modern word embeddings such as Word2Vec,
- Before attempting the best-in-class Language models (pre-trained or otherwise) that involve seq2seq networks and transformers, I wanted to take a step back and understand the fundamentals better
- That is the reason behind developing codes and notes for Bigram and Trigram Markov Language Models.
- The below link has my notes on
Statistical Markov Language Models
:
https://senthilkumarm1901.github.io/StatisticalLanguageModels/