ashrafulparan/arbanking77

Model Description

This is finetuned version of Arabertv02 on arbanking77 dataset for ArafinNLP Task-1

Model Hyperparameters

• —Learning Rate (LR): 4e-2
• —Weight Decay (WD): 0
• —Warmup Steps (WS): 50
• —Epochs (EP): 15