ArabicNewSplits4_withSameOriginalTrainFileOfSplit3_FineTuningAraBERT_noAug_task1_organization

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8934
• Qwk: 0.6871
• Mse: 0.8934
• Rmse: 0.9452

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	Qwk	Mse	Rmse
No log	0.5	2	3.7117	0.1429	3.7117	1.9266
No log	1.0	4	2.1519	0.2390	2.1519	1.4669
No log	1.5	6	1.0964	0.4800	1.0964	1.0471
No log	2.0	8	0.9221	0.3734	0.9221	0.9602
No log	2.5	10	0.8148	0.5234	0.8148	0.9026
No log	3.0	12	0.7699	0.5859	0.7699	0.8774
No log	3.5	14	0.8276	0.6326	0.8276	0.9097
No log	4.0	16	0.7629	0.6927	0.7629	0.8734
No log	4.5	18	0.7543	0.6973	0.7543	0.8685
No log	5.0	20	0.7117	0.7219	0.7117	0.8436
No log	5.5	22	0.7554	0.7041	0.7554	0.8691
No log	6.0	24	0.9363	0.6586	0.9363	0.9676
No log	6.5	26	0.9602	0.6462	0.9602	0.9799
No log	7.0	28	0.8795	0.6641	0.8795	0.9378
No log	7.5	30	0.8847	0.6631	0.8847	0.9406
No log	8.0	32	0.8917	0.6699	0.8917	0.9443
No log	8.5	34	0.9017	0.6699	0.9017	0.9496
No log	9.0	36	0.9023	0.6699	0.9023	0.9499
No log	9.5	38	0.8919	0.6826	0.8919	0.9444
No log	10.0	40	0.8934	0.6871	0.8934	0.9452

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.0+cu118
• Datasets 2.21.0
• Tokenizers 0.19.1