cefr

NokaMan: multi-language learning ability assessment (CEFR writing samples)

The CEFR Level Classifier Project is a AI Streamlit-based application that utilizes a two-phase Camembert model to accurately classify texts into six CEFR language proficiency levels. This innovative tool offers user-friendly navigation for both training and text classification.

Free open-source Ukrainian course. A1 to C2. Based on Ukrainian State Standard 2024. «Мова – душа народу»

Elaisa Search - CEFR Categerorizer

Claude Code plugin for CEFR-aligned English exam prep: IELTS, TOEFL iBT (2026 format), CEFR B1-C2. Original content only; append-only progress tracking.

Code and data for TSAR 2025 Shared Task

Open multilingual frequency dictionary with POS, gender, inflection and CEFR.

Bulgarian NLP content-enrichment engine — morphology, CEFR-style difficulty, vocabulary, contextual glosses. Source-available work sample.

英语教练：Hermes 英语学习 skill by Rac 🦝

Multilingual AI language tutor for listening, speaking, reading, and writing using local Qwen, Whisper, FastAPI, PostgreSQL, and Docker.

📝 The Swedish Kelly list as a Hugging Face dataset

Listen to Bulgarian audio and understand it word by word — local Whisper transcription in a synced karaoke reader. BYO-audio, private by design.

An interactive Telegram bot that helps users improve their English grammar, expand vocabulary, and prepare for IELTS, TOEFL, and CEFR exams through real-time feedback and practice.

Web-based PDF vocabulary analyzer with CEFR level highlighting and automatic glossary generation

Measuring semantic robustness in LLM-based CEFR essay scoring through systematic prompt paraphrasing. University of Exeter Year 3 Computer Science research project.

End-to-end grammar scoring engine that evaluates spoken English using speech-to-text, NLP-based grammar analysis, CEFR mapping, and a cloud-deployed Streamlit app.

LangGraph multi-agent system that generates CEFR-aligned language curricula as strictly-typed JSON, grounded by RAG over Qdrant.

A high-performance CEFR level predictor API for English text, powered by DeBERTa-V3 and optimized with ONNX Runtime.

Pilot application within the PedagoReLearn framework, implementing Reflexive Reciprocity Theory (RRT) through a reinforcement learning–enhanced micro-tutoring system for German vocabulary acquisition (CEFR A1). Learner interaction data feeds upstream to the parent PedagoReLearn repository for instructional policy analysis.

🇷🇸 Telegram bot for learning Serbian vocabulary via spaced repetition. Levels A1–C2, two quiz modes, smart scheduling, per-user progress. Built with Python & aiogram.