Wrappers for FastText Library used for fast text representation and classification.

Simple command-line tool to identify the language of a given text using 2 different identification models.

LinearCosine: Adding beats multiplying for lower-precision efficient cosine similarity

AI From Scratch in C++ - for educational purposes. Available now: Fully Connected Network and Convolution Neural Network.

This code models cross-linked syntactic and semantic trees of words in sentences from a lexicon of graph rewriting rules (created in GrGen.NET). Graphs are visualized in realtime using YComp and a front-end is created for the user in C.

Knowledge Embedding Project for Tsinghua Deep Learning Summer School 2018

This program automatically identifies the subject of posts from the EECS 280 Piazza using natural language processing and machine learning techniques. It figures out what project the post is about by using previous training data and analyzing all the words in a post.

🛠️ Create and execute code with live compilation in an interactive environment tailored for your device using the Words programming language.

Deep Learning Chinese Word Segment

Code velds encapsulating UDPipe.

Client for LLM Chat/RAG/Fine-Tuning

Лабораторные работы по курсу «Компьютерная лингвистика-1» (ФИТ НГУ, 2022).

Bachelor thesis - language identification of short texts (2014)

🍫 ~ small and customizable BPE tokenizer

Collection of reading notes and solutions to the problems I encountered in my research.

💌 Love Letters Project – Process WhatsApp exported chats, apply NLP techniques like lemmatization, and optimize text for AI analysis. Run on Google Colab, deploy APIs with Ngrok, and manage tokens efficiently using GPT-4o’s tokenizer (Tiktoken). 🚀

A Parallel Term Frequency–Inverse Document Frequency Implementation in C++. Final Project for COP 4520 UCF

just a french soundex