AITeamVN/GRPO-VI-Qwen2-7B-RAG

GRPO-VI-Qwen2-7B-RAG is a 7.6 billion parameter large language model developed by AITeamVN, fine-tuned from Qwen2.5-7B-Instruct. It specializes in Retrieval-Augmented Generation (RAG) tasks, particularly for the Vietnamese language, and also demonstrates strong performance in STEM-related tasks and general question answering. The model utilizes Supervised Fine-Tuning combined with Group Relative Policy Optimization (GRPO) on a Vietnamese dataset to enhance its RAG capabilities and language understanding.