MODELHugging Face

5/5

Glot500-m

Moroccan Arabic (Darija)(ary)LM / EncoderXLM-RoBERTa BaseXLMRobertaForMaskedLM

Model Info

Language

Moroccan Arabic (Darija) (ary)

Architecture

XLMRobertaForMaskedLM

Institution

Imani-Googhari et al. — LMU Munich

Last Updated

14 May 2026

Intended Use Cases

Multilingual LM

License & Access

SPDX identifierCC BY-NC-SA

Open with restrictions — attribution required; some commercial or derivative use limited.

Provenance

Source URL

https://huggingface.co/cis-lmu/glot500-base

Fetch Timestamp

14 May 2026

Submitter

Imani-Googhari et al. — LMU Munich

Full Record

23 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 (17/18)

iso code: ary
Model Name: Glot500-m
Where is it hosted: HuggingFace
Link: https://huggingface.co/cis-lmu/glot500-base
Year: 2023
Creator/Author(s): Imani-Googhari et al. — LMU Munich
Paper Reference: ACL 2023
NLP Task: LM / Encoder
Architecture: XLMRobertaForMaskedLMderivedHugging Face config: architectures / model_type
Base Model: XLM-RoBERTa Base
Fine-tuned?: Yes
License: CC BY-NC-SA
Open Access: Open
Has Model Card: Yes
SOTA Tasks: Multilingual LM
Verified By: Not stated
Date Indexed: 2026-05-14
Comment: ary_Arab covered

Readiness (LRS) (6/7)

model id: glot500-m-ary
Model Type: LLMderivedHugging Face card: pipeline_tag
Training Data Sources: Glot500-c (511 langs)
Benchmark Used: Glot500-m Evaluation Suite (Bukusu)derivedregistry benchmark records for this model
Benchmark Score: LM evaluation for Bukusu
Benchmark Date: 2023
Status: Not stated

Readiness Score Breakdown

5/5

Data Coverage100%

Model Availability100%

Benchmarks100%

Documentation69%

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.

Glot500-m

Afrikaans included

13 languagesAfrikaans(afr)XLM-RoBERTa BaseLM / Encoder (XLM-R-base extension)

Open12 May 2026

Glot500-m

Libyan Arabic covered via Tatoeba data

13 languagesLibyan Arabic(ayl)XLM-RoBERTa BaseLM / Encoder

Open14 May 2026

Glot500-m

Tuki covered via SIL Bloom Library

13 languagesTuki (Ki)(bag)XLM-RoBERTa BaseLM / Encoder

Open14 May 2026

Glot500-m

Bamum covered via SIL Bloom Library

13 languagesBamum(bax)XLM-RoBERTa BaseLM / Encoder

Open14 May 2026

Glot500-m

Babanki covered via SIL Bloom Library

13 languagesBabanki(bbk)XLM-RoBERTa BaseLM / Encoder

Open14 May 2026

Glot500-m

Kpwe/Mokpwe covered via SIL Bloom Library

13 languagesKpwe (Mokpwe)(bri)XLM-RoBERTa BaseLM / Encoder

Open12 May 2026

Glot500-m

Bukusu covered in Glot500

13 languagesBukusu(bxk)XLM-RoBERTa BaseLM / Encoder

Open21 May 2026

Glot500-m

Medumba covered in Glot500

13 languagesMedumba(byv)XLM-RoBERTa BaseLM / Encoder

Open21 May 2026

Glot500-m

Chokwe covered in Glot500

13 languagesChokwe(cjk)XLM-RoBERTa BaseLM / Encoder

Open22 May 2026

Glot500-m

Chenoua covered in Glot500

13 languagesChenoua(cnu)XLM-RoBERTa BaseLM / Encoder

Open22 May 2026

Glot500-m

Guinea-Bissau Creole covered in Glot500

13 languagesGuinea-Bissau Creole(cpp)XLM-RoBERTa BaseLM / Encoder

Open22 May 2026

Glot500-m

Chuka covered in Glot500

13 languagesChuka(cuh)XLM-RoBERTa BaseLM / Encoder

Open22 May 2026

Related Models

DarijaBERT

First BERT model dedicated to Moroccan Darija

Moroccan Arabic (Darija)(ary)BERT-baseLM

Open14 May 2026

Atlas-Chat (Darija LLM)

First Darija-specific instruction-tuned LLM; 9B and 27B variants

2 copiesMoroccan Arabic (Darija)(ary)Gemma-2-9B / 27BMT

Open14 May 2026

NLLB-200 (ary_Arab)

Moroccan Arabic supported as ary_Arab

54 languages13 copiesMoroccan Arabic (Darija)(ary)Transformer (NLLB)MT / Encoder-Decoder

Open14 May 2026