The Automatic Code Evaluation Using LLM Model is a web-based system designed to automate the evaluation of programming tasks for students. The system integrates two primary roles: Trainer and Student, each with specific functionalities. Trainers post programming questions, while students view, write, and submit solutions to these questions. Upon submission, the system evaluates the code using two distinct processes. First, the Trainer evaluates the submission based on a pre-defined scoring rubric. Then, the AI Model, using the Qwen2.5-coder:1.5b model running on a local machine, evaluates the same code for correctness, logic, and performance. The final score is calculated by combining 70% Trainer's score and 30% AI score, ensuring a comprehensive assessment of the student's abilities.

The system features a Trainer Dashboard for managing questions, tracking submissions, and generating scores, and a Student Dashboard for applying to tasks, submitting solutions, and receiving feedback. Additionally, the system facilitates smooth communication between the Trainer and Student and provides real-time notifications about submissions and evaluations.By incorporating both human and AI-based evaluations, the system offers a scalable, efficient, and transparent solution for programming task evaluation. This system ensures timely and constructive feedback for students while minimizing the administrative burden on trainers.

Keywords: Automatic Code Evaluation, Programming Tasks, Trainer Dashboard, AI Evaluation, Student Feedback, Large Language Model, Qwen2.5-coder, Grading System, Real-time Notifications.