language_model: true
license: apache-2.0
tags:
- text-generation
- language-modeling
- multilingual
- transformer
- wikipedia
- darija
- arabic
datasets:
- wikimedia/wikipedia
metrics:
- cross_entropy_loss
model_name: Darija-GPT
dataset: Wikipedia (Subset)
dataset_args: 'config: ary'
language: ary
finetuned_from: scratch
tasks: text-generation