CUAIStudents/Qwen-Ar-GEC
TEXT GENERATIONConcurrency Cost:1Model Size:7.6BQuant:FP8Ctx Length:32kPublished:Sep 11, 2025Architecture:Transformer0.0K Cold

CUAIStudents/Qwen-Ar-GEC is a 7.6 billion parameter Qwen 2.5-7B-Instruct based model fine-tuned by CUAIStudents for Arabic Grammatical Error Correction (GEC). Utilizing QLoRA on 50,000 samples, this model is specifically designed to detect and correct grammatical, spelling, and stylistic errors in Arabic text. It excels at automated proofreading, language learning assistance, and academic writing support for Arabic content, ensuring full diacritization.

Loading preview...