Stepan Tytarenko's repositories
crypto-news
Leveraging the power of OpenAI GPT to process cryptocurrency news and delivering concise summaries directly to your Telegram and Twitter!
go-EdgeGPT
GoBingChat: A GoLang-powered CLI tool to seamlessly interact with Bing Chat, enabling easy integration into any application.
nano-BERT
Nano-BERT is a straightforward, lightweight and comprehensible custom implementation of BERT, inspired by the foundational "Attention is All You Need" paper. The primary objective of this project is to distill the essence of transformers by simplifying the complexities and unnecessary details.
llm-friend-bot
This project seeks to fine-tune the GPT-2.5 model with a personal touch. By training it on personal Telegram chats, we aim to capture the essence of individualized writing styles. To spice things up, we've also fed the model a sprinkle of anecdotes and a dash of math tasks.
space-model
Space Model framework that allows for maintaining generalizability, and enhances the performance on the downstream task by utilizing task-specific context attribution. It is an external LLM layer, that improves accuracy in classification task for multiple datasets, such as HateXplain, IMDB movies reviews and more.
warehouse-management-system
An integrated Warehouse Management System leveraging Django, Vue.js, Kotlin, and Raspberry Pi for intelligent logistics automation, role-based access, and real-time package tracking.
bstree-tango
Tango trees are an advanced binary search tree that aims to optimize the search operation by minimizing rotations. By maintaining a set of preferred paths and adjusting them dynamically, Tango trees ensure efficient access to frequently searche
go-BingDALLE
GoBingChat: A GoLang-powered CLI tool to seamlessly interact with Bing Image Generator, enabling easy integration into any application.
news-contest
This repository contains Jupyter notebooks detailing the experiments conducted in our research paper on Ukrainian news classification. We introduce a framework for simple classification dataset creation with minimal labeling effort, and further compare several pretrained models for the Ukrainian language.
q-learning
a Python-based platformer infused with Q-Learning and dynamic level creation from simple JSON files.
rdrone-front
A comprehensive software system leveraging convolutionary neural networks, drones, and IoT for real-time road quality monitoring and alerts. Built using Python, TensorFlow, Django, and Vue.js.
semi-supervised-regression
COREG with Transformers Exploration: This research project delves into enhancing the COREG algorithm by incorporating transformer models, particularly DistilBERT and DeBERTa. Through rigorous experimentation on various GPU-equipped platforms, the modified COREG seeks to identify confident labels based on the predictions' similarity between models.
interview-prep
This repository contains C++ solutions for various LeetCode problems, categorized into multiple sections by topics
pwa-timetable
By offering features like an interactive calendar, real-time notifications, and offline access, this PWA ensures that students can seamlessly manage classes, assignments, and exams, keeping education on track despite the challenges of remote learning
rdrone-back
A comprehensive software system leveraging convolutionary neural networks, drones, and IoT for real-time road quality monitoring and alerts. Built using Python, TensorFlow, Django, and Vue.js.
students_nure
A predictive model for postgraduate program enrollment based on historical data, student rankings, and various influencing factors.
llm-stable-diffusion-movie-creator
ipynb for Google Colab to create a movie using LLM to generate the scenes description and stable diffusion to create visualizations. State of the art stable-diffusion with the most powerful open-source 7B Mistral LLM
cam-bert
We propose a novel method of fine-tuning the model for a particular downstream task, which proves to be more efficient and generalizable. We show that in an example of a fake news detection task, utilizing three distinct datasets and outperforming the baseline model in both the same dataset and cross-dataset zero-shot test.
crypto-text-classification
Application of NLP techniques, leveraging transformer models, to predict the impact of news articles on Bitcoin prices. A pre-trained transformer model, on a small manually labeled dataset is expanded by the dataset labeled automatically using a custom Bing API and Bing GPT with internet access.
stable-diffusion-intro
Educational project on Variational Autoencoders (VAEs) and Stable Diffusion Models. Examples of "dreaming" diffusion models and walking over the latent space. Configuring parameters and understanding the architecture
stepantita.github.io
Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes