DarijaBERT

Moroccan Arabic (Darija)(ary)LMBERT-baseClassification, NER / Encoder (BERT)

View Source

Model Info

Language

Moroccan Arabic (Darija) (ary)

Architecture

Classification, NER / Encoder (BERT)

Institution

Gaanoun et al. — AIOX Lab + SI2M Lab INSEA

Last Updated

14 May 2026

Intended Use Cases

Classification SOTA: best F1 on Moroccan dialect NLU tasks

License & Access

SPDX identifierMIT

Open — attribution required; commercial use allowed.

Provenance

Source URL

https://huggingface.co/SI2M-Lab/DarijaBERT

Fetch Timestamp

14 May 2026

Submitter

Gaanoun et al. — AIOX Lab + SI2M Lab INSEA

Full Record

21 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: DarijaBERT
Where is it hosted: HuggingFace
Link: https://huggingface.co/SI2M-Lab/DarijaBERT
Year: 2024
Creator/Author(s): Gaanoun et al. — AIOX Lab + SI2M Lab INSEA
Paper Reference: Gaanoun et al. 2024
NLP Task: LM
Architecture: Classification, NER / Encoder (BERT)
Base Model: BERT-base
Fine-tuned?: No
License: MIT
Open Access: Open
Has Model Card: Yes
SOTA Tasks: Classification SOTA: best F1 on Moroccan dialect NLU tasks
Verified By: Not stated
Date Indexed: 2026-05-14
Comment: First BERT model dedicated to Moroccan Darija

Readiness (LRS) (4/7)

model id: darijabert-ary
Model Type: LLM
Training Data Sources: 100M tokens: Darija stories, YouTube comments, tweets
Benchmark Used: Not stated
Benchmark Score: Not stated
Benchmark Date: Not stated
Status: Activederivedupdated within 3 years