mbart50

Fine-tuned and compared 3 🤗 pre-trained Multilingual LLMs

Fine-tune mBart 50 for Bengali Sentence Error Correction

Polyglot Power: mBART & mT5 Translation Toolkit mBart50 multilanguage translation plus docker image and bulk pdf translation, also older English to german translator local/server using T5 model and german to polish english to spanish MarianMT transformer for file translations.

This repository contains a simple and beginner-level noteboook which employs the mBART LLM/model for the translation of English text into Indian Languages.

NLP course project: Comparison of mBART-50 and Helsinki-NLP MarianMT models for Polish–Croatian machine translation with BLEU evaluation

Using Hugging Face MBart50 model for language translation from English to Hindi.

A multilingual text summarization app which uses the mBART-50 language model fine-tuned on the CNN DAILY NEWS Dataset. Enables users to upload, summarize and translate text data.

A project for fine-tuning a pre-trained mBart 50 model for English-Persian subtitle translation using NLP techniques. The fine-tuned model has been made available on Hugging Face 🤗 for open-source access and further development by the community.

Benchmarking Study of Bloomz-560m, mBART-large, IndicBART on the Indic Languages

In the context of my bachelor's thesis, I conducted an evaluation of german text simplification with mono-, bi- and cross-lingual fine-tuning of the mBART50 model, using the OneStopEnglish and the DE-Plain-APA corpora.

Experimental Eng->Sin Subtitle Translation Model

Comparing mBART-50, NLLB-200, Helsinki-NLP & mT5 for English→Malayalam NMT using Back-Translation data augmentation in low-resource settings.

🌍 AI-powered multilingual text translator built with mBART50. Fine-tuned on OPUS Books & IIT Bombay datasets to support English ↔ Spanish, French, Italian, Portuguese, and Hindi. Clean Streamlit UI for fast local inference.

🚀 Model API with Flask, BART Base & mBART

A Python-based machine translation system for English-to-Hinglish, leveraging models like mBART and T5-small to address the complexities of code-mixed languages, achieving state-of-the-art performance with a BLEU score of 43.23.

multi-language translate using UKPLab/EasyNMT