New Tajik Language LLM Developed for Resource-Constrained Environments

2026-05-28

Researchers have introduced Soro, a family of large language models specifically designed for the Tajik language, capable of operating under limited compute and connectivity conditions in Tajikistan. The models are based on open-weight Gemma 3 checkpoints and fine-tuned on a Tajik-centric corpus.

Source: arXiv · cs.AI

Reported by VERA Newswire.