# Low-resource Language Optimization

Sentiment analysis model optimized for Khmer financial texts, capable of classifying positive/negative sentiments

A speech recognition model adapted from whisper-large-v3, specifically optimized for multiple languages in Uganda

Jina Embeddings V3 is a multilingual sentence embedding model supporting over 100 languages, specializing in sentence similarity and feature extraction tasks.

A multilingual text processing model supporting over 100 languages and writing systems, covering various Arabic dialects and minority languages

Aya 101 is a large-scale multilingual generative language model supporting instructions in 101 languages, outperforming similar models in various evaluations.

MMS-1B-FL102 is part of Facebook's Massively Multilingual Speech project, an automatic speech recognition model supporting 102 languages, based on the 1-billion-parameter Wav2Vec2 architecture, achieving multilingual transcription through adapter technology.

Historical multilingual and monolingual ByT5 base model, current version focuses on English text processing.

Danish BERT is a language model optimized for Danish, excelling in the Danish ScandEval benchmark.

A multilingual model for Maltese pre-trained on the Maltese corpus v4.0 based on multilingual BERT initial checkpoints

BERTu is a monolingual Maltese model based on the BERT architecture, specifically designed for low-resource languages, supporting various natural language processing tasks.

XGLM-2.9B is a multilingual autoregressive language model with 2.9 billion parameters, trained on a diverse and balanced corpus of 500 billion subword tokens across multiple languages.

A named entity recognition model fine-tuned based on xlm-roberta-base, specifically optimized for Nigerian Pidgin

XGLM-1.7B is a multilingual autoregressive language model with 1.7 billion parameters, trained on a diverse and balanced corpus of 500 billion subword tokens.

A Bulgarian speech recognition model fine-tuned on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - BG dataset based on facebook/wav2vec2-xls-r-300m

A speech recognition model based on Microsoft's UniSpeech architecture, specifically fine-tuned for the Kyrgyz language

This is an automatic speech recognition model fine-tuned on Galician speech datasets based on facebook/wav2vec2-xls-r-300m.

AfriBERTa is a multilingual pretrained model supporting 11 African languages with 111 million parameters, suitable for tasks like text classification and named entity recognition.

Portuguese clinical named entity recognition model based on BioBERTpt, supporting 13 UMLS-compatible clinical entity types

Portuguese clinical named entity recognition model based on BioBERTpt, supporting 13 UMLS-compatible clinical entity types

A Filipino RoBERTa model trained on the TLUnified corpus, improved from previous versions with case-sensitive processing support.

A Transformer-based machine translation model from English to Malagasy, developed by the Helsinki-NLP team.

Transformer-based English to Lozi machine translation model developed by the Helsinki-NLP team

Marathi to English machine translation model based on OPUS dataset, using transformer-align architecture

A Transformer-based English to Mizo (Lushai) machine translation model developed by the Helsinki-NLP team.

opus-mt-en-cy is a machine translation model based on the transformer-align architecture, specifically designed for translating English to Welsh. This model was developed by the Helsinki-NLP team and trained on the OPUS dataset.

opus-mt-en-st is a machine translation model based on the Transformer architecture, specifically designed for translating English to Southern Sotho.

OPUS-MT machine translation model from French to Kinyarwanda, based on Transformer architecture, trained using OPUS datasets.