Bu yazıda 8 Mart için küçük bir “biz dili” manifestosuna başlıyoruz: Türkiye’de kadınların kadınlara yazdığı güzellik, stil ve yaşam yazılarında dayanışma nasıl kuruluyor? SüslüTrendler derleminden yola çıkarak; kombin, makyaj, iyi...
Read More
We proudly present the first part of our Turkish subword manifesto: how Turkish language modeling benefits from morphology. In this article, we compare character-, word-, and morphology-aware subword tokenization, where...
Read More
For a long time, Turkish models climbed leaderboards written somewhere else—translated datasets, English‑first assumptions, noisy web scrapes. Useful, yes. Representative of real Turkish? Not quite. TrGLUE is our answer: a...
Read More
We proudly introduce BellaTurca, the ultimate Turkish large corpus, bringing diversity and high quality to fight the dullness and blandness in Turkish language modeling. We’re talking about 250GB of text...
Read More
Gender bias in language models is nothing new, but here’s the catch: most of the research out there focuses on English. So what about other languages — like Turkish? That’s...
Read More
Bir varmış; bir yokmuş / Evvel zaman içinde / Kalbur saman içinde / Develer tellal iken / Pireler berber iken / Ben dedemin beşiğini tıngır mıngır sallarken…
Read More
Turkish morphology is fascinating and even more fascinating when you process it with spaCy Turkish models. In this article, quantify and characterize Turkish morphological info in news using spaCy Turkish...
Read More
Natural Language Processing (NLP) has grown leaps and bounds in the past decade, but if you’re working with Turkish, you’ve probably faced some challenges. Turkish is a morphologically rich, agglutinative...
Read More