arabert_cross_relevance_task6_fold4

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: 0.2650
• Qwk: 0.3603
• Mse: 0.2650

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: 64
• eval_batch_size: 64
• 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
No log	0.125	2	0.5208	0.1951	0.5208
No log	0.25	4	0.5167	0.1743	0.5167
No log	0.375	6	0.4163	0.2124	0.4163
No log	0.5	8	0.3864	0.2931	0.3864
No log	0.625	10	0.3687	0.3169	0.3687
No log	0.75	12	0.3018	0.3399	0.3018
No log	0.875	14	0.3682	0.2424	0.3682
No log	1.0	16	0.4150	0.2379	0.4150
No log	1.125	18	0.3832	0.2446	0.3832
No log	1.25	20	0.3293	0.3301	0.3293
No log	1.375	22	0.2923	0.4174	0.2923
No log	1.5	24	0.2780	0.3858	0.2780
No log	1.625	26	0.2804	0.3481	0.2804
No log	1.75	28	0.2982	0.3235	0.2982
No log	1.875	30	0.3255	0.3037	0.3255
No log	2.0	32	0.3232	0.3223	0.3232
No log	2.125	34	0.3010	0.3265	0.3010
No log	2.25	36	0.2972	0.3385	0.2972
No log	2.375	38	0.2953	0.3744	0.2953
No log	2.5	40	0.2914	0.3362	0.2914
No log	2.625	42	0.2775	0.3356	0.2775
No log	2.75	44	0.2717	0.3488	0.2717
No log	2.875	46	0.2682	0.3516	0.2682
No log	3.0	48	0.2717	0.3659	0.2717
No log	3.125	50	0.2762	0.4330	0.2762
No log	3.25	52	0.2974	0.4860	0.2974
No log	3.375	54	0.3155	0.4427	0.3155
No log	3.5	56	0.2887	0.3242	0.2887
No log	3.625	58	0.2903	0.2974	0.2903
No log	3.75	60	0.2966	0.3011	0.2966
No log	3.875	62	0.3108	0.3189	0.3108
No log	4.0	64	0.3002	0.3729	0.3002
No log	4.125	66	0.3025	0.4184	0.3025
No log	4.25	68	0.2768	0.4398	0.2768
No log	4.375	70	0.2840	0.4468	0.2840
No log	4.5	72	0.2805	0.4330	0.2805
No log	4.625	74	0.2739	0.3745	0.2739
No log	4.75	76	0.2753	0.3395	0.2753
No log	4.875	78	0.2722	0.3153	0.2722
No log	5.0	80	0.2714	0.3164	0.2714
No log	5.125	82	0.2809	0.3280	0.2809
No log	5.25	84	0.2844	0.3541	0.2844
No log	5.375	86	0.2818	0.3541	0.2818
No log	5.5	88	0.2663	0.3932	0.2663
No log	5.625	90	0.2624	0.3872	0.2624
No log	5.75	92	0.2639	0.3941	0.2639
No log	5.875	94	0.2817	0.3331	0.2817
No log	6.0	96	0.3073	0.3225	0.3073
No log	6.125	98	0.3094	0.3274	0.3094
No log	6.25	100	0.2931	0.3261	0.2931
No log	6.375	102	0.2777	0.3289	0.2777
No log	6.5	104	0.2693	0.3611	0.2693
No log	6.625	106	0.2685	0.3510	0.2685
No log	6.75	108	0.2718	0.3630	0.2718
No log	6.875	110	0.2628	0.3742	0.2628
No log	7.0	112	0.2633	0.3688	0.2633
No log	7.125	114	0.2771	0.3635	0.2771
No log	7.25	116	0.2913	0.3936	0.2913
No log	7.375	118	0.2849	0.4111	0.2849
No log	7.5	120	0.2790	0.4053	0.2790
No log	7.625	122	0.2834	0.4053	0.2834
No log	7.75	124	0.2822	0.3877	0.2822
No log	7.875	126	0.2858	0.3818	0.2858
No log	8.0	128	0.2754	0.3690	0.2754
No log	8.125	130	0.2679	0.3880	0.2679
No log	8.25	132	0.2657	0.4144	0.2657
No log	8.375	134	0.2684	0.4198	0.2684
No log	8.5	136	0.2677	0.4184	0.2677
No log	8.625	138	0.2659	0.4077	0.2659
No log	8.75	140	0.2673	0.3853	0.2673
No log	8.875	142	0.2664	0.3686	0.2664
No log	9.0	144	0.2685	0.3582	0.2685
No log	9.125	146	0.2677	0.3582	0.2677
No log	9.25	148	0.2671	0.3582	0.2671
No log	9.375	150	0.2663	0.3582	0.2663
No log	9.5	152	0.2659	0.3526	0.2659
No log	9.625	154	0.2647	0.3697	0.2647
No log	9.75	156	0.2651	0.3705	0.2651
No log	9.875	158	0.2648	0.3705	0.2648
No log	10.0	160	0.2650	0.3603	0.2650

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1