ArabicNewSplits_FineTuningAraBERT_AugV5_k3_task2_organization_fold1

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: 1.2089
• Qwk: 0.1650
• Mse: 1.2089
• Rmse: 1.0995

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	0.1667	2	2.7988	0.0223	2.7988	1.6730
No log	0.3333	4	2.1895	0.0013	2.1895	1.4797
No log	0.5	6	1.6534	-0.0561	1.6534	1.2858
No log	0.6667	8	1.3308	-0.0576	1.3308	1.1536
No log	0.8333	10	1.1537	0.0	1.1537	1.0741
No log	1.0	12	1.0321	0.0	1.0321	1.0159
No log	1.1667	14	0.9171	0.0	0.9171	0.9577
No log	1.3333	16	0.8777	0.0	0.8777	0.9368
No log	1.5	18	0.8187	0.0	0.8187	0.9048
No log	1.6667	20	0.7407	-0.0343	0.7407	0.8606
No log	1.8333	22	0.7213	-0.0343	0.7213	0.8493
No log	2.0	24	0.7162	-0.0514	0.7162	0.8463
No log	2.1667	26	0.7517	-0.0343	0.7517	0.8670
No log	2.3333	28	0.8221	0.0	0.8221	0.9067
No log	2.5	30	0.9091	0.0	0.9091	0.9535
No log	2.6667	32	0.8390	-0.0172	0.8390	0.9160
No log	2.8333	34	0.7597	-0.0132	0.7597	0.8716
No log	3.0	36	0.6810	0.0586	0.6810	0.8253
No log	3.1667	38	0.6696	0.0960	0.6696	0.8183
No log	3.3333	40	0.6696	0.1124	0.6696	0.8183
No log	3.5	42	0.6862	0.0586	0.6862	0.8284
No log	3.6667	44	0.7170	0.1360	0.7170	0.8468
No log	3.8333	46	0.7453	0.0706	0.7453	0.8633
No log	4.0	48	0.7558	0.0706	0.7558	0.8694
No log	4.1667	50	0.6905	0.0383	0.6905	0.8310
No log	4.3333	52	0.6712	0.0634	0.6712	0.8193
No log	4.5	54	0.6990	0.2057	0.6990	0.8361
No log	4.6667	56	0.7240	0.1764	0.7240	0.8509
No log	4.8333	58	0.7116	0.2028	0.7116	0.8436
No log	5.0	60	0.7025	0.0865	0.7025	0.8381
No log	5.1667	62	0.6988	0.0994	0.6988	0.8360
No log	5.3333	64	0.6903	0.0960	0.6903	0.8308
No log	5.5	66	0.6812	0.0994	0.6812	0.8253
No log	5.6667	68	0.7267	0.1866	0.7267	0.8525
No log	5.8333	70	0.8736	0.1764	0.8736	0.9347
No log	6.0	72	0.9794	0.1255	0.9794	0.9897
No log	6.1667	74	1.0014	0.0906	1.0014	1.0007
No log	6.3333	76	0.9560	0.1380	0.9560	0.9777
No log	6.5	78	0.8962	0.2352	0.8962	0.9467
No log	6.6667	80	0.8621	0.1771	0.8621	0.9285
No log	6.8333	82	0.8460	0.1687	0.8460	0.9198
No log	7.0	84	0.8421	0.1212	0.8421	0.9176
No log	7.1667	86	0.8589	0.1212	0.8589	0.9268
No log	7.3333	88	0.9129	0.1077	0.9129	0.9555
No log	7.5	90	1.0367	0.1534	1.0367	1.0182
No log	7.6667	92	1.1517	0.1544	1.1517	1.0732
No log	7.8333	94	1.1765	0.1544	1.1765	1.0847
No log	8.0	96	1.1597	0.1544	1.1597	1.0769
No log	8.1667	98	1.1773	0.1544	1.1773	1.0850
No log	8.3333	100	1.1680	0.1657	1.1680	1.0807
No log	8.5	102	1.1279	0.1636	1.1279	1.0620
No log	8.6667	104	1.1335	0.1636	1.1335	1.0646
No log	8.8333	106	1.1676	0.1636	1.1676	1.0806
No log	9.0	108	1.2096	0.1599	1.2096	1.0998
No log	9.1667	110	1.2308	0.1412	1.2308	1.1094
No log	9.3333	112	1.2353	0.1412	1.2353	1.1114
No log	9.5	114	1.2265	0.1412	1.2265	1.1075
No log	9.6667	116	1.2184	0.1412	1.2184	1.1038
No log	9.8333	118	1.2121	0.1412	1.2121	1.1010
No log	10.0	120	1.2089	0.1650	1.2089	1.0995

Framework versions

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