Giter VIP home page Giter VIP logo

nti_ml_20-21's Introduction

Hi there, I'm George Kokush

Computer science student, Intern ML-engineer from Russia ๐Ÿ‡ท๐Ÿ‡บ

Kaggle Badge

I'm George Kokush, 18 y. o. HSE and novice ML/DL-engineer from Russia.

Contact me on: Telegram VK
CV

PyTorch NumPy Pandas Sklearn

Python C++

My last projects

  • "Semantically-Informed Regressive Encoder Score" submission for WMT23 Shared task workshop [ Paper ] [ Repo ]
    • Our task was to develop NN-based metric for text evaluation(machine translation)
    • We improved our developments from AIRI research project trying different approaches(including use of additional vector representations and contrastive learning)
    • Our approach was on 5th place in chinese-english and hebrew-english language pairs and 11th on english-german language pair
    • Our paper was reviewed and we were invited to EMNLP 23 conference
  • Team submission for Eval4NLP Shared task workshop [ Paper ] [ Repo ]
    • Our task was to develop metric for text evaluation(MT&Summarization) only using prompt-engineering techniques and approaches
    • We tried the new approach based on AutoMQM work
    • Our paper was reviewed and we were invited to IJCNLP-AACL 23 conference
  • "Efficient LLM-based metrics for NLG" research project for AIRI Summer School [ Presentation ] [ Repo ]
    • Our task was to develop NN-based metric for text evaluation(machine translation)
    • We tried to beat GPT4-based GEMBA metric by fine-tuning LLMs for translation evaluation
    • I implemented LLM Encoder+MLP decoder architecture which got the best quality
  • "Multimodality in image2text tasks" research project for 1st year of HSE [ Poster ] [ Repo ]
    • Our task was to develop image2text model for russian language
    • We implemented the BLIP-2 architecture and tested it on various configurations
    • We adapted architecture for russian language and achieved tolerable quality
  • NTI ML contest, 2021 [ Repo ]
    • I used lots of classic ML algorithms(linear and logistic regression, trees, boostings, etc), web-scrapping for data extraction and grid-search for hyperparams search
    • We achieved one of the best scores in final rating
  • Toxic detector bot, pet project [ Repo ]
    • I trained CatBoostClassifiers for toxicity prediction using word2vec embeddings
  • Other pet-projects

nti_ml_20-21's People

Contributors

egoluback avatar gorg1t avatar izitckiyt avatar

Stargazers

 avatar  avatar  avatar

Watchers

 avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.