Malik Talha Saeed

Academic

Career

Verbatim text

Malik Talha Saeed i19-0413
CodeVerb
CodeVerb is an AI-based system that simplifies the process of writing high-quality computer programs using natural language input. With its transformer-based model scheme, CodeVerb can generate relevant computer code from massive datasets of python language, scraped from coding tutorial websites, programming forums and GitHub. The system is designed to solve various problems faced by non-experienced programmers, researchers, and algorithm designers. It can also help people with disabilities by providing a voice feature for generating code. CodeVerb will greatly reduce project development times and streamline efficiency in the programming industry, making it accessible and easy to use for from its website and publicly available VS Code Extension.
Features Include:
- Generation of Syntactically Correct Python Language Code from English Language Text.
- Website to Manage User Accounts for Personalized Tracking of Generated Code History etc.
- A VS Code IDE Extension publicly available to be integrated in VS Code for Efficient Development
- More than one code predictions allowing the programmer to use alternate solutions.
- Direct Integration of Code in the Workspace with one tap or shortcut through the Extension.
- Voice Recognition Feature allows seamless generation of code predictions without the need for typing text manually.
- Deployment of Model on cloud for Low Latency in Predictions on Website & VS Code Extension.
Group Members
Malik Talha Saeed i19-0413
Hammad Ahmed i19 0582
Hunaid Sohail Mirza i19-2043
Project Supervisor
Ms. Noor Ul Ain
codeverb
A highly advanced AI Model trained on massive code datasets capable of converting Natural Language Text or Algorithms to Python code
Project Overview
User Text/Speech IDE Extension AI Model Generated Code
e.g. print first 100 terms in fibonacci series.
c, b = 0, 1
while b < 100:
    print(c)
    c, b = b, a+b
Deployment Dataset Generation
Timeline
Data Scraping & Preprocessing Nov/22 - Oct/23
Large Language Model (LLM) Training Jan/23 - Feb/23
Neural Network Based LLM Development Nov/22 - Dec/22
IDL Extension Development & LLM Deployment Apr/23 - May/23
LLM Testing & Improvement Mar/23 - Apr/23
Tools & Technologies
Technology Used:
Python, PyTorch, ReactJS, JavaScript, NodeJS, ExpressJS, Flask, TypeScript, AWS
Supervisor Name:
Ms. Noor Ul Ain
Group Members:
Hammad Ahmed (i19 - 0582)
Malik Talha Saeed (i19 - 0413)
Hunaid Sohail Mirza (i19 - 2043)
National University of Computer and Science (NUCES), Islamabad Campus | Pakistan

AI enrichment