# African language processing

An Amharic language sentence transformer based on the RoBERTa architecture, designed for sentence similarity calculation and feature extraction.

This model is a fine-tuned multilingual translation model based on NLLB-200-distilled-600M, specifically optimized for bidirectional translation between Wolof, French, and English.

Whosper-large-v2 is a cutting-edge speech recognition model specifically designed for Wolof, the primary language of Senegal. Built upon OpenAI's Whisper-large-v2, it significantly improves Word Error Rate (WER) and Character Error Rate (CER).

A Wolof text-to-speech (TTS) model fine-tuned based on the SpeechT5 architecture, using a custom tokenizer adapted to Wolof language characteristics

This model is a fine-tuned version of facebook/mms-1b-all on the Isma/alffa_wolof dataset, specifically designed for Wolof automatic speech recognition (ASR) tasks.

A fine-tuned version of OpenAI's Whisper model, specifically designed for automatic speech recognition tasks in the low-resource Ghanaian language Akan

A pre-trained Amharic language model based on the bert-medium architecture, with 40.5 million parameters trained on 290 million tokens, achieving performance comparable to larger multilingual models.

A wav2vec2-based speech recognition model, fine-tuned at the phoneme level for Amharic

A speech recognition model fine-tuned based on facebook/wav2vec2-large-xlsr-53, supporting Bemba language

This is a small ELECTRA model pretrained on the TLMD dataset containing over 40 million tokens

This is a Wolof automatic speech recognition model fine-tuned based on facebook/wav2vec2-xls-r-300m, aimed at addressing the scarcity of Wolof language resources.

A RoBERTa model for classifying Swahili news texts, trained on Google Cloud TPUv3-8

This is a RoBERTa model pretrained on Kinyarwanda datasets using Masked Language Modeling (MLM) objective, with case-insensitive tokenization during pretraining.

This is a BERT base model pretrained specifically for Tigrinya, trained for 40 epochs on a dataset of 40 million tokens.

A token classification model for Wolof named entity recognition (NER) tasks, fine-tuned on the Wolof portion of the MasakhaNER dataset based on xlm-roberta-base

This model is fine-tuned on the Swahili portion of the MasakhaNER dataset for named entity recognition tasks in Swahili text.

A Transformer-based machine translation model for translating Tsonga (ts) to English (en).

A model fine-tuned on Amharic text based on XLM-RoBERTa, outperforming the original XLM-RoBERTa in Named Entity Recognition tasks

This is a BERT model optimized for the Igbo language, obtained by fine-tuning on top of the multilingual BERT, and it performs better than the base multilingual model in Igbo text processing tasks.

Hausa automatic speech recognition model fine-tuned from facebook/wav2vec2-xls-r-300m

Luganda (lg) to Swedish (sv) machine translation model trained on OPUS dataset, using transformer-align architecture

This is an automatic speech recognition (ASR) model fine-tuned on the Kabyle language dataset based on the facebook/wav2vec2-xls-r-300m model

A Transformer-based machine translation model supporting translation tasks from Tumbuka (tum) to French (fr)

Hausa automatic speech recognition model based on wav2vec2-xls-r-300m architecture, fine-tuned on Common Voice 8.0 Hausa dataset

This is a multilingual translation model that supports translation from English to various Bantu languages, including Shona, Zulu, Kinyarwanda, and more.

A Transformer-based machine translation model for translating from the Tumbuka language to Spanish.

This is a Hausa automatic speech recognition model fine-tuned from facebook/wav2vec2-xls-r-300m, trained on the Common Voice 8.0 dataset.

A model fine-tuned on Luganda text based on XLM-RoBERTa, outperforming the base model in Named Entity Recognition tasks

Swahili automatic speech recognition model fine-tuned from facebook/wav2vec2-xls-r-300m, trained on Common Voice 8 dataset