SILMA Kashif 2B Instruct v1.0 is an advanced open-weight model designed for Retrieval-Augmented Generation (RAG) tasks, excelling in answering questions based on context in Arabic and English. It also has minor capabilities in Entity Extraction. Built on Google Gemma’s foundational models, it delivers top-tier performance within the 3-9 billion parameter range, as validated by the
SILMA RAGQA Benchmark. The model supports a 12k context length