数据集:

1.5B Arabic Corpus 1.5B-Arabic-Corpus 1.5B_Arabic_Corpus 3A1.5B+Arabic+Corpus 3A1.5B-Arabic-Corpus 3A1.5B_Arabic_Corpus 3AAssafir(private) 3AOSCAR+Arabic+Unshuffled 3Aoscar-arabic-unshuffled 3AOsian 3ATwitter(private) 3Awikipedia Assafir(private) OSCAR Arabic Unshuffled oscar-arabic-unshuffled OSIAN Twitter(private) wikipedia

语言:

ar

预印本库:

arxiv:1406.2661 arxiv:2003.00104 arxiv:2012.15516 arxiv:2012.15520

其他:

AutoTrain Compatible bert electra gpt2 pretraining text-generation-inference

任务:

填充掩码 文本分类 文本生成

类库:

JAX PyTorch Safetensors TensorBoard TensorFlow Transformers