shamilmohammedi/Azhari_Model_v0.4_Academic
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Mar 28, 2026Architecture:Transformer Cold

The Azhari Model v0.4 Academic by shamilmohammedi is a 7.6 billion parameter Llama-3-8B variant, fine-tuned on academic Sharia datasets from Al-Maktaba Al-Shamila. This research prototype, with a 32768 token context length, focuses on evaluating the effectiveness of combining Fine-Tuning (FT) and Retrieval-Augmented Generation (RAG) to reduce hallucinations in Islamic Jurisprudence contexts. It is specifically designed for academic benchmarking and AI research in specialized religious domains.

Loading preview...