Monglish_Arabic_FAQ-V3

This model is a fine-tuned version of CAMeL-Lab/bert-base-arabic-camelbert-msa on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.3059
• Accuracy: 0.9558
• F1: 0.9520
• Precision: 0.9674
• Recall: 0.9558

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: 4
• eval_batch_size: 4
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Accuracy	F1	Validation Loss	Precision	Recall
1.1065	1.0	226	0.8850	0.8681	1.0925	0.9284	0.8850
0.12	2.0	452	0.3676	0.9381	0.9392	0.9639	0.9381
0.0403	3.0	678	0.2427	0.9558	0.9562	0.9705	0.9558
0.0179	4.0	904	0.2722	0.9469	0.9443	0.9645	0.9469
0.0129	5.0	1130	0.2694	0.9558	0.9520	0.9674	0.9558
0.0151	6.0	1356	0.2493	0.9558	0.9520	0.9674	0.9558
0.0046	7.0	1582	0.2811	0.9558	0.9520	0.9674	0.9558
0.0036	8.0	1808	0.2630	0.9558	0.9520	0.9674	0.9558
0.0051	9.0	2034	0.2787	0.9558	0.9520	0.9674	0.9558
0.0038	10.0	2260	0.2924	0.9558	0.9520	0.9674	0.9558
0.0027	11.0	2486	0.3098	0.9558	0.9520	0.9674	0.9558
0.0034	12.0	2712	0.3010	0.9558	0.9520	0.9674	0.9558
0.0046	13.0	2938	0.3051	0.9558	0.9520	0.9674	0.9558
0.0033	14.0	3164	0.3070	0.9558	0.9520	0.9674	0.9558
0.0016	15.0	3390	0.3059	0.9558	0.9520	0.9674	0.9558

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.0.0
• Tokenizers 0.19.1