results

This model is a fine-tuned version of SI2M-Lab/DarijaBERT on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5291
• Macro F1: 0.7697
• Accuracy: 0.8007
• Recall: 0.7687

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Macro F1	Accuracy	Recall
0.6848	0.9877	40	0.6040	0.6869	0.7504	0.6821
0.5937	2.0	81	0.5376	0.7396	0.7799	0.7286
0.4946	2.9877	121	0.5313	0.7474	0.7816	0.7434
0.386	4.0	162	0.5291	0.7697	0.8007	0.7687
0.3114	4.9877	202	0.5690	0.7391	0.7782	0.7329
0.2477	6.0	243	0.5891	0.7480	0.7834	0.7441
0.1804	6.9877	283	0.6194	0.7422	0.7764	0.7366

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.2
• Tokenizers 0.19.1