translate your prompt text to enlish using local translation models
Few NLP augmentation techniques: synonym/antonym and back translation.
Hybrid RAG system for translating Ancient Egyptian transliterations using the TLA dataset.
Links to my repositories, where I implement a wide variety of Natural Language Processing models using TensorFlow and Hugging Face.
This AI-powered MVP (Minimum Viable Product) is designed to help users enhance their English-speaking skills through real-time, personalized feedback in Marathi.
🎮 Real-time game subtitle translator with AI-powered OCR. Context-aware translation for 20+ languages. Free offline models + dirt cheap APIs. Perfect for gaming in foreign languages!
A real-time Speech-to-Speech translation pipeline (ASR ➡️ NMT ➡️ TTS) using OpenAI Whisper, MarianMT, and gTTS. Features a Flask backend and a responsive web UI for low-latency multilingual communication and audio synthesis.
Python and C++ implementations of MarianMT that run on the Rockchip NPU
This project is a AI tool translator using NLP techniques
Bidirectional Urdu↔English translation service on Helsinki-NLP MarianMT — FastAPI + vanilla-JS web UI with RTL handling, mocked API tests, CPU-runnable.
Persona Styled Recipes are generated given a list of ingredients
NLP course project: Comparison of mBART-50 and Helsinki-NLP MarianMT models for Polish–Croatian machine translation with BLEU evaluation
Fine-tuned transformer that translates English into Quenya, Tolkien's Elvish language, with a Flask web interface.
A multi-language translator built using pre-trained MarianMT models from HuggingFace. It supports English-to-French, Hindi, Russian, and Spanish translation with BLEU score evaluation to assess translation quality. Ideal for real-time translation and NLP experimentation.
Encoder-Decoder NLP model with Beam Search and MarianMT for sequence-to-sequence tasks. Includes a web UI frontend and Flask/Streamlit backend.
Multilingual AI translation and text-to-speech application built with Python, Streamlit, Transformers, and MarianMT.
This repository uses BERT for medical diagnosis, incorporating data augmentation techniques like backtranslation and paraphrasing to improve model performance in predicting diseases from symptom descriptions.
Encoder-Decoder NLP model with Beam Search and MarianMT for sequence-to-sequence tasks. Includes a web UI frontend and Flask/Streamlit backend.
To associate your repository with the marianmt topic, visit your repo's landing page and select "manage topics."