Hello! I’m Mohammadreza Ghofrani, a PhD student at Aalto University, focusing on sustainable AI innovation, under supervision of Dr. Gopika Premsankar*. I obtained my Master’s degree in Artificial Intelligence, specializing in Natural Language Processing, from Amirkabir University of Technology (Tehran Polytechnic). During master’s studies, I had the privilege of working under the guidance of Dr. MohammadMehdi Homayounpour* and Dr. Saeedeh Momtazi**, focusing on my thesis titled “Response Generation in Task-based Dialogue Systems with Limited Training Data.”

Prior to my master’s, I completed my bachelor’s degree under the supervision of Dr. Saeedeh Momtazi**, where I worked on “Implementing a Relation Extraction System for Persian Texts using Distant Supervision.”

Education

  • PhD, Artificial Intelligence at Aalto University
    Supervisor: Dr. Gopika Premsankar*
  • M.Sc., Artificial Intelligence at Amirkabir University of Technology (Tehran Polytechnic)
    Thesis: Response Generation in Task-based Dialogue Systems with Limited Training Data
    Supervisor: Dr. MohammadMehdi Homayounpour* & Dr. Saeedeh Momtazi**
  • B.Sc., Computer Engineering at Amirkabir University of Technology (Tehran Polytechnic)
    Overal Grade: GPA 3.98/4 (19.18/20)
  • High School, Mathematics, Sampad Shahid Beheshti High School
    Overal Grade: GPA 4/4 (19.87/20)

Publication

2- LVLM-COUNT: Enhancing the Counting Ability of Large Vision-Language Models 2024 arXiv, GitHub
Mohammad Fetrat Qarabagh, Mohammadreza Ghofrani, and Kimon Fountoulakis

1- Persian Address Geocoding: an LALR Parsing and Dynamic Programming Approach JADM 2023 [pdf]
Alireza Mazochi, Sara Bourbour, Mohammadreza Ghofrani, and Saeedeh Momtazi

Awards

  • Ranked 5th/∼40, in M.Sc. Artificial Intelligence, ArmirKabir University of Technology (AUT), Iran, 2023
  • Merit-Based admission, M.Sc. program, AmirKabir University, Iran, 2021
  • Ranked 4th/∼140, in B.Sc. Computer Science field, AmirKabir University of Technology (AUT), Iran, 2021
  • Ranked 458/137000+, the nationwide university entrance exam, Iran, 2017

Selected Projects


CommonCrawl Language Specific Downloader github
This script downloads only the segments of the CommonCrawl dataset for the specified language. CommonCrawl archives and publishes internet content every few months, making it a valuable resource for extracting data to train language models (or large language models, as they are commonly known today).

Computer Skills

Programming Languages Python, MATLAB, SQL, C++, C, Java, HTML, CSS, JS

Frameworks:

  • Machine Learning Numpy, Scikit-learn, Scipy, Pandas
  • Deep Learning Pytorch, Tensorflow, Huggingface Transformers, Huggingface Optimum, Hugging face’s Datasets, Hugging face’s Evaluate, Hugging face’s Tokenizers, Pytorch Lighting, Pytorch ignite, ONNX, OpenRE, Gensim
  • Large Language Model LangChain, OpenAI
  • Plotting Matplotlib, Plotly, Seaborn
  • Web Programming Flask, BentoML, Django, Django REST Framework
  • Web Scraping Selenium, Scrapy, Requests, Beautiful Soup, Urllib
  • Image Opencv, Pillow/PIL

Databases Redis, ElasticSearch, MongoDB, PostgreSQL, MySQL, SQLite
Tools Git, LATEX, Docker, Weka, Apache Hadoop, Arduino
OS Ubuntu, Windows, Pop! OS, Manjaro