ArabicNewSplits6_WithDuplicationsForScore5_FineTuningAraBERT_run3_AugV5_k1_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: 1.0494
• Qwk: 0.5948
• Mse: 1.0494
• Rmse: 1.0244

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	0.3333	2	2.3584	0.0263	2.3584	1.5357
No log	0.6667	4	1.5726	0.1917	1.5726	1.2541
No log	1.0	6	1.2525	0.2342	1.2525	1.1192
No log	1.3333	8	1.3271	0.1475	1.3271	1.1520
No log	1.6667	10	1.4177	0.3543	1.4177	1.1907
No log	2.0	12	1.7354	0.3403	1.7354	1.3174
No log	2.3333	14	2.1633	0.2262	2.1633	1.4708
No log	2.6667	16	2.5027	0.2288	2.5027	1.5820
No log	3.0	18	2.3444	0.1914	2.3444	1.5312
No log	3.3333	20	1.9998	0.2155	1.9998	1.4142
No log	3.6667	22	1.8190	0.2521	1.8190	1.3487
No log	4.0	24	1.7885	0.3205	1.7885	1.3373
No log	4.3333	26	1.6569	0.3668	1.6569	1.2872
No log	4.6667	28	1.6328	0.3958	1.6328	1.2778
No log	5.0	30	1.6401	0.3848	1.6401	1.2806
No log	5.3333	32	1.5225	0.4138	1.5225	1.2339
No log	5.6667	34	1.4126	0.4624	1.4126	1.1885
No log	6.0	36	1.2496	0.4841	1.2496	1.1178
No log	6.3333	38	1.1718	0.4900	1.1718	1.0825
No log	6.6667	40	1.1720	0.5302	1.1720	1.0826
No log	7.0	42	1.1444	0.5629	1.1444	1.0698
No log	7.3333	44	1.1238	0.5803	1.1238	1.0601
No log	7.6667	46	1.1178	0.5791	1.1178	1.0573
No log	8.0	48	1.0614	0.5884	1.0614	1.0302
No log	8.3333	50	1.0351	0.5859	1.0351	1.0174
No log	8.6667	52	1.0004	0.5915	1.0004	1.0002
No log	9.0	54	0.9981	0.5915	0.9981	0.9990
No log	9.3333	56	1.0201	0.6002	1.0201	1.0100
No log	9.6667	58	1.0448	0.5948	1.0448	1.0222
No log	10.0	60	1.0494	0.5948	1.0494	1.0244

Framework versions

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