🧠 I Trained a Massive Word2Vec Model on 13 Billion Russian Fiction Words — Here’s What Happened
TL;DR: I built a lemma‑based Word2Vec model (CBOW, 300d) on a huge corpus of Russian fiction (13B tokens → 7.3B after cleaning). You can load it with Gensim and explore semantic neighborhoods of words like слово , язык , речь . The model captures literary semantics without stop words or grammar tags. Check it out on Hugging Face . Why another word2vec for Russian? Most pre‑trained Russian…
A researcher trained a 300-dimensional lemma-based Word2Vec model (CBOW) on a massive 13 billion token corpus of Russian fiction, resulting in a 7.3 billion token cleaned dataset. This model, available on Hugging Face, captures literary semantics without stop words or grammatical tags, differing from general web-based models. The clean corpus comprised 1.36 billion sentences, with 539 million processed paragraphs.
The model, designed for researchers and developers to explore literary semantics, features a 197 MB main model and two 5.8 GB arrays of negative sampling weights and vectors. Usage examples demonstrate word associations and stylistic analysis potential. The model's limitations include lack of POS tagging and lemma ambiguity, but it is lightweight and easy to integrate.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
