MODELHugging Face

4/5

masakhane/afroxlmr-large-ner-masakhaner-1.0_2.0

Ghomálá'(bbj)NERAfroXLM-R large (XLM-R large adapted on 25 African langs)XLMRobertaForTokenClassification

Model Info

Language

Ghomálá' (bbj)

Architecture

XLMRobertaForTokenClassification

Institution

Jesujoba Alabi, David Adelani et al. (Masakhane)

Last Updated

22 May 2026

Intended Use Cases

Best multilingual NER across MasakhaNER 1.0 & 2.0 incl. Ghomálá'

License & Access

SPDX identifierAFL-3.0

Licensed under AFL-3.0. Review the source terms before use.

Provenance

Source URL

https://huggingface.co/masakhane/afroxlmr-large-ner-masakhaner-1.0_2.0

Fetch Timestamp

22 May 2026

Submitter

Jesujoba Alabi, David Adelani et al. (Masakhane)

Full Record

19 of 25 fields stated

Every column the registry tracks. Values marked derivedwere filled by a rule or read from the source's text rather than stated in its metadata; the rule or the quoted sentence is shown under the value.

Registry columns (16/18)

iso code: bbj
Model Name: masakhane/afroxlmr-large-ner-masakhaner-1.0_2.0
Where is it hosted: HuggingFace
Link: https://huggingface.co/masakhane/afroxlmr-large-ner-masakhaner-1.0_2.0
Year: 2022
Creator/Author(s): Jesujoba Alabi, David Adelani et al. (Masakhane)
Paper Reference: EMNLP 2022 arXiv:2210.12391
NLP Task: NER
Architecture: XLMRobertaForTokenClassificationderivedHugging Face config: architectures / model_type
Base Model: AfroXLM-R large (XLM-R large adapted on 25 African langs)
Fine-tuned?: Yes
License: AFL-3.0
Open Access: Open
Has Model Card: Yes
SOTA Tasks: Best multilingual NER across MasakhaNER 1.0 & 2.0 incl. Ghomálá'
Verified By: Not stated
Date Indexed: 2026-05-22
Comment: Not stated

Readiness (LRS) (3/7)

model id: masakhane-afroxlmr-large-ner-masakhaner-1-0-2-0-bbj
Model Type: NERderivedHugging Face card: pipeline_tag
Training Data Sources: MasakhaNER 1.0 (10 langs) + MasakhaNER 2.0 (20 langs incl. bbj)
Benchmark Used: Not stated
Benchmark Score: Not stated
Benchmark Date: Not stated
Status: Not stated

Readiness Score Breakdown

4/5

Data Coverage95%

Model Availability100%

Benchmarks0%

Documentation83%

Recency100%

Representativeness

Gender Balance

Regional Distribution

Dialect Coverage

Inclusivity

Relations Explorer link

Also Available In

This is the same multilingual corpus, sliced for other languages the platform indexes it under.

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesAmharic(amh)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesBambara(bam)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesEwe(ewe)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesFon(fon)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesHausa(hau)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesIgbo(ibo)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesKinyarwanda(kin)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesLuganda(lug)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesLuo(luo)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesMooré(mos)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesChewa(nya)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesNigerian Pidgin(pcm)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

masakhane/afroxlmr-large-ner-masakhaner-1.0_2.0

Shona (sna) achieves F1=96.5 on MasakhaNER 2.0 — highest in the benchmark. Multilingual NER model covering 21 African languages.

21 languagesShona(sna)Davlan/afro-xlmr-largeNER

Open17 Jun 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesSwahili(swa)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesTswana(tsn)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesTwi(twi)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesWolof(wol)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesXhosa(xho)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesYoruba(yor)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

afroxlmr-large-ner-masakhaner-1.0_2.0

token-classification model · XLMRobertaForTokenClassification · masakhane

21 languagesZulu(zul)XLMRobertaForTokenClassificationtoken-classification

Restricted27 Oct 2026

Related Models

masakhane/xlm-roberta-base-masakhaner2-bbj (per-language model)

NER model · David Adelani et al. (Masakhane)

Ghomálá'(bbj)XLM-RoBERTa baseNER

Open22 May 2026

facebook/nllb-200-distilled-600M (covers bbj_Latn)

MT model · M2M100ForConditionalGeneration · Meta AI NLLB Team

54 languages13 copiesGhomálá'(bbj)mBART / NLLB architectureMT

Open22 May 2026

afrispeech-gender-id

audio-classification model · AfriSpeech

2,051 languagesGhomálá'(bbj)audio-classification

Open27 Oct 2026