ArabicNewSplits6_WithDuplicationsForScore5_FineTuningAraBERT_run1_AugV5_k2_task5_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: 0.5599
• Qwk: 0.7576
• Mse: 0.5599
• Rmse: 0.7483

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	0.1818	2	2.1694	-0.0366	2.1694	1.4729
No log	0.3636	4	1.4346	0.1959	1.4346	1.1978
No log	0.5455	6	1.2657	0.2497	1.2657	1.1250
No log	0.7273	8	1.2423	0.4151	1.2423	1.1146
No log	0.9091	10	1.5205	0.3585	1.5205	1.2331
No log	1.0909	12	1.2622	0.4184	1.2622	1.1235
No log	1.2727	14	1.0897	0.3905	1.0897	1.0439
No log	1.4545	16	1.1344	0.4487	1.1344	1.0651
No log	1.6364	18	0.9575	0.4233	0.9575	0.9785
No log	1.8182	20	1.0376	0.5218	1.0376	1.0187
No log	2.0	22	1.2512	0.4901	1.2512	1.1186
No log	2.1818	24	1.0978	0.5275	1.0978	1.0478
No log	2.3636	26	0.8167	0.5936	0.8167	0.9037
No log	2.5455	28	0.9145	0.4704	0.9145	0.9563
No log	2.7273	30	1.0989	0.4569	1.0989	1.0483
No log	2.9091	32	0.9705	0.4817	0.9705	0.9851
No log	3.0909	34	0.7527	0.6103	0.7527	0.8676
No log	3.2727	36	0.8026	0.6287	0.8026	0.8959
No log	3.4545	38	0.9104	0.6462	0.9104	0.9541
No log	3.6364	40	0.8651	0.6471	0.8651	0.9301
No log	3.8182	42	0.7628	0.6837	0.7628	0.8734
No log	4.0	44	0.6630	0.6766	0.6630	0.8142
No log	4.1818	46	0.6394	0.7006	0.6394	0.7996
No log	4.3636	48	0.6506	0.6719	0.6506	0.8066
No log	4.5455	50	0.6237	0.6827	0.6237	0.7898
No log	4.7273	52	0.5885	0.7497	0.5885	0.7672
No log	4.9091	54	0.5796	0.7607	0.5796	0.7613
No log	5.0909	56	0.6241	0.7228	0.6241	0.7900
No log	5.2727	58	0.6761	0.7586	0.6761	0.8222
No log	5.4545	60	0.6633	0.7606	0.6633	0.8144
No log	5.6364	62	0.6068	0.7483	0.6068	0.7790
No log	5.8182	64	0.5693	0.7344	0.5693	0.7545
No log	6.0	66	0.6615	0.6681	0.6615	0.8133
No log	6.1818	68	0.6916	0.6510	0.6916	0.8316
No log	6.3636	70	0.6160	0.7311	0.6160	0.7849
No log	6.5455	72	0.5592	0.7629	0.5592	0.7478
No log	6.7273	74	0.5857	0.7493	0.5857	0.7653
No log	6.9091	76	0.6051	0.7458	0.6051	0.7779
No log	7.0909	78	0.5858	0.7760	0.5858	0.7654
No log	7.2727	80	0.5591	0.7746	0.5591	0.7477
No log	7.4545	82	0.5613	0.7449	0.5613	0.7492
No log	7.6364	84	0.5693	0.7436	0.5693	0.7546
No log	7.8182	86	0.5705	0.7466	0.5705	0.7553
No log	8.0	88	0.5652	0.7438	0.5652	0.7518
No log	8.1818	90	0.5581	0.7518	0.5581	0.7470
No log	8.3636	92	0.5549	0.7553	0.5549	0.7449
No log	8.5455	94	0.5592	0.7619	0.5592	0.7478
No log	8.7273	96	0.5735	0.7585	0.5735	0.7573
No log	8.9091	98	0.5814	0.7622	0.5814	0.7625
No log	9.0909	100	0.5798	0.7622	0.5798	0.7615
No log	9.2727	102	0.5714	0.7633	0.5714	0.7559
No log	9.4545	104	0.5667	0.7633	0.5667	0.7528
No log	9.6364	106	0.5621	0.7576	0.5621	0.7498
No log	9.8182	108	0.5605	0.7576	0.5605	0.7487
No log	10.0	110	0.5599	0.7576	0.5599	0.7483

Framework versions

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