4 ms·
FastText isn't a LLM, it's a token embedding model with a simple classifier on top.
by tensor 3y ago
FastText isn't a LLM, it's a token embedding model with a simple classifier on top.
- marcinzm 3y agoSure but it's existence means the statement is really "gzip approach not better than dnn models, and doesn't compete or be cheaper to run than previous models like FastText." That's not a very meaningful value statement for the approach (although why gzip is even half-decent might be a very interesting research question).