AraBERT_token_classification_AraEval24_rand_trun_multi_fixed

This model is a fine-tuned version of aubmindlab/bert-base-arabertv2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 4.0911
• Precision: 0.0588
• Recall: 0.1184
• F1: 0.0786
• Accuracy: 0.1794

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: 8
• eval_batch_size: 8
• seed: 42
• 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	Precision	Recall	F1	Accuracy
1.5667	1.0	1534	2.2108	0.0351	0.0556	0.0430	0.2275
1.2795	2.0	3068	2.5072	0.0444	0.0745	0.0556	0.1989
1.0642	3.0	4602	2.5903	0.0598	0.0986	0.0745	0.2405
0.8661	4.0	6136	3.3476	0.0635	0.1145	0.0817	0.1784
0.6954	5.0	7670	3.6462	0.0611	0.1064	0.0776	0.1658
0.5748	6.0	9204	3.7515	0.0621	0.1181	0.0814	0.1714
0.4858	7.0	10738	3.7003	0.0621	0.1167	0.0811	0.1884
0.4276	8.0	12272	3.9753	0.0548	0.1103	0.0732	0.1768
0.3903	9.0	13806	4.1462	0.0577	0.1150	0.0768	0.1699
0.3339	10.0	15340	4.0911	0.0588	0.1184	0.0786	0.1794

Framework versions

• Transformers 4.30.2
• Pytorch 2.3.1+cu121
• Datasets 2.20.0
• Tokenizers 0.13.3