ArabicNewSplits6_WithDuplicationsForScore5_FineTuningAraBERT_run2_AugV5_k1_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: 0.6825
• Qwk: 0.6894
• Mse: 0.6825
• Rmse: 0.8261

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	0.25	2	4.8181	-0.0214	4.8181	2.1950
No log	0.5	4	3.0472	0.0775	3.0472	1.7456
No log	0.75	6	1.7749	0.0950	1.7749	1.3322
No log	1.0	8	1.2721	0.2081	1.2721	1.1279
No log	1.25	10	1.0288	0.2586	1.0288	1.0143
No log	1.5	12	1.0643	0.2689	1.0643	1.0317
No log	1.75	14	1.1435	0.2587	1.1435	1.0693
No log	2.0	16	1.0403	0.2535	1.0403	1.0200
No log	2.25	18	0.9411	0.4160	0.9411	0.9701
No log	2.5	20	0.9341	0.4316	0.9341	0.9665
No log	2.75	22	0.8895	0.4966	0.8895	0.9431
No log	3.0	24	0.7714	0.5662	0.7714	0.8783
No log	3.25	26	0.8362	0.5832	0.8362	0.9144
No log	3.5	28	0.8891	0.5890	0.8891	0.9429
No log	3.75	30	0.7211	0.6294	0.7211	0.8492
No log	4.0	32	0.6850	0.6875	0.6850	0.8276
No log	4.25	34	0.7811	0.6358	0.7811	0.8838
No log	4.5	36	0.7747	0.6509	0.7747	0.8802
No log	4.75	38	0.7769	0.6630	0.7769	0.8814
No log	5.0	40	0.7258	0.6832	0.7258	0.8519
No log	5.25	42	0.7599	0.6653	0.7599	0.8717
No log	5.5	44	0.7971	0.6426	0.7971	0.8928
No log	5.75	46	0.7474	0.6758	0.7474	0.8645
No log	6.0	48	0.6884	0.7054	0.6884	0.8297
No log	6.25	50	0.7031	0.7066	0.7031	0.8385
No log	6.5	52	0.6788	0.7117	0.6788	0.8239
No log	6.75	54	0.6686	0.7221	0.6686	0.8177
No log	7.0	56	0.6695	0.7261	0.6695	0.8182
No log	7.25	58	0.6645	0.6965	0.6645	0.8152
No log	7.5	60	0.7008	0.6547	0.7008	0.8371
No log	7.75	62	0.6921	0.6547	0.6921	0.8319
No log	8.0	64	0.6828	0.6547	0.6828	0.8263
No log	8.25	66	0.6972	0.6547	0.6972	0.8350
No log	8.5	68	0.7154	0.6585	0.7154	0.8458
No log	8.75	70	0.7211	0.6768	0.7211	0.8492
No log	9.0	72	0.7272	0.6689	0.7272	0.8528
No log	9.25	74	0.7081	0.6797	0.7081	0.8415
No log	9.5	76	0.6927	0.6882	0.6927	0.8323
No log	9.75	78	0.6847	0.6894	0.6847	0.8274
No log	10.0	80	0.6825	0.6894	0.6825	0.8261

Framework versions

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