uaritm/gemmamed_cardio

uaritm/gemmamed_cardio is a specialized instruction-following Gemma-4B-Instruct model, fine-tuned by Uaritm for cardiology-related information and medical queries in Ukrainian. This model underwent a two-stage LoRA fine-tuning process, including linguistic adaptation on a large Ukrainian corpus and domain specialization on cardiovascular health data. It is optimized for efficient inference on consumer hardware, provided as a GGUF Q4KM quantized file (~2.4 GB) with a context length of 4096 tokens, excelling as a high-quality cardiology assistant in Ukrainian.