GitHub-based Hugging Face Portfolio

Sangyeon Ryu
YEonleo

NLP and data-centric AI portfolio built from my GitHub projects. This page focuses on practical dataset engineering, evaluation tooling, and reproducible workflows.

GitHub Profile Hugging Face

FiscalLeaderboard

Benchmark leaderboard for Korean tax/accounting LLM evaluation with Streamlit app and structured result pipelines.

Open on GitHub

CPA_datasets

CPA exam QA dataset curation workflow with review status tracking, correction tools, and JSONL-first dataset management.

Open on GitHub

LACD

Research project repository for NLP methodology and implementation focused on language data quality and modeling.

Open on GitHub

FractalLLM

LLM research codebase for experimental modeling, evaluation, and paper-linked reproducible components.

Open on GitHub

Core Stack

Python PyTorch NLP Dataset Curation Evaluation Pipelines Streamlit