lapa-llm/lapa-v0.1.2-instruct

Lapa LLM v0.1.2 is a 12 billion parameter open-source language model developed by a consortium of Ukrainian researchers, based on Gemma-3. It is specifically optimized for Ukrainian language processing, featuring a custom tokenizer that significantly reduces token count for Ukrainian text. This model excels in tasks such as English-to-Ukrainian translation, image processing in Ukrainian, summarization, and Q&A, making it highly efficient for RAG systems and culturally relevant text generation.