Know a great AI tool or AI news source? Submit it to AiByts
2 min read

Adapting Qwen3 to Tajik from 0.6B to 8B: Lower Text Loss Did Not Mean Better Understanding

Small LoRA adapters trained on Tajik news and books lowered Qwen3's Tajik text loss at every size we tried, from 0.6B to 8B, by 0.60 to 1.45 nats per token, but none of them improved Tajik reading comprehension where it can be measured.

By Tohir Saidzoda, Madina Abdusalomzoda, Nasimjon Rajabov, Akhrorkhon Ikromov, Komron Giyosiev, Abubakr Abdullozoda, Amin Jalilov, Jonibek Kishvarov, Rahmonali Sharipov, Mariyamkhon ShodyevaarXiv - cs.CL (NLP)

Read the full story

Originally published by arXiv - cs.CL (NLP) on Nov 5, 2026.

Read on arXiv - cs.CL (NLP)(opens in a new tab)

Comments

Sign in to join the discussion.

No comments yet — be the first to share your thoughts.