arabert_cross_relevance_task2_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: 0.2539
• Qwk: 0.0
• Mse: 0.2539

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	0.8503	0.0176	0.8493
No log	0.25	4	0.3494	0.1224	0.3496
No log	0.375	6	0.5004	0.0515	0.5009
No log	0.5	8	0.3323	-0.0164	0.3328
No log	0.625	10	0.3005	0.0	0.3007
No log	0.75	12	0.3118	0.0122	0.3119
No log	0.875	14	0.3876	0.0454	0.3880
No log	1.0	16	0.4447	-0.0079	0.4455
No log	1.125	18	0.4798	-0.1341	0.4806
No log	1.25	20	0.5158	-0.0658	0.5166
No log	1.375	22	0.3941	-0.0135	0.3946
No log	1.5	24	0.3295	-0.0042	0.3298
No log	1.625	26	0.3244	0.0	0.3247
No log	1.75	28	0.3471	-0.0164	0.3476
No log	1.875	30	0.3538	-0.0042	0.3543
No log	2.0	32	0.3130	0.0	0.3134
No log	2.125	34	0.2809	0.0	0.2811
No log	2.25	36	0.2793	0.0	0.2795
No log	2.375	38	0.3047	0.0	0.3050
No log	2.5	40	0.3696	0.0582	0.3702
No log	2.625	42	0.3992	0.1381	0.3999
No log	2.75	44	0.3456	-0.0172	0.3461
No log	2.875	46	0.2805	0.0	0.2807
No log	3.0	48	0.2675	0.0	0.2675
No log	3.125	50	0.2713	0.0	0.2712
No log	3.25	52	0.2691	0.0	0.2691
No log	3.375	54	0.2726	0.0	0.2727
No log	3.5	56	0.2951	0.0	0.2954
No log	3.625	58	0.3397	0.0452	0.3402
No log	3.75	60	0.3175	0.0450	0.3180
No log	3.875	62	0.2714	0.0	0.2716
No log	4.0	64	0.2574	0.0	0.2573
No log	4.125	66	0.2608	0.0	0.2606
No log	4.25	68	0.2590	0.0	0.2590
No log	4.375	70	0.2813	0.0	0.2815
No log	4.5	72	0.3090	0.0	0.3094
No log	4.625	74	0.3106	0.0	0.3110
No log	4.75	76	0.2951	0.0	0.2954
No log	4.875	78	0.2767	0.0	0.2769
No log	5.0	80	0.2684	0.0	0.2685
No log	5.125	82	0.2825	0.0122	0.2829
No log	5.25	84	0.3266	0.0746	0.3272
No log	5.375	86	0.3486	0.0841	0.3493
No log	5.5	88	0.3299	0.0833	0.3305
No log	5.625	90	0.2905	0.0245	0.2909
No log	5.75	92	0.2675	0.0	0.2677
No log	5.875	94	0.2613	0.0	0.2614
No log	6.0	96	0.2609	0.0	0.2610
No log	6.125	98	0.2647	0.0	0.2648
No log	6.25	100	0.2690	0.0	0.2692
No log	6.375	102	0.2672	0.0	0.2675
No log	6.5	104	0.2632	0.0	0.2633
No log	6.625	106	0.2554	0.0	0.2554
No log	6.75	108	0.2637	0.0	0.2635
No log	6.875	110	0.2670	-0.0118	0.2667
No log	7.0	112	0.2544	0.0	0.2543
No log	7.125	114	0.2718	0.0245	0.2720
No log	7.25	116	0.2966	0.0161	0.2970
No log	7.375	118	0.2940	0.0161	0.2944
No log	7.5	120	0.2714	0.0245	0.2717
No log	7.625	122	0.2558	0.0	0.2559
No log	7.75	124	0.2559	0.0	0.2558
No log	7.875	126	0.2567	0.0	0.2565
No log	8.0	128	0.2560	0.0	0.2559
No log	8.125	130	0.2526	0.0	0.2526
No log	8.25	132	0.2523	0.0	0.2524
No log	8.375	134	0.2528	0.0	0.2528
No log	8.5	136	0.2536	0.0	0.2537
No log	8.625	138	0.2528	0.0	0.2529
No log	8.75	140	0.2516	0.0	0.2516
No log	8.875	142	0.2517	0.0	0.2516
No log	9.0	144	0.2521	0.0	0.2521
No log	9.125	146	0.2526	0.0	0.2526
No log	9.25	148	0.2531	0.0	0.2531
No log	9.375	150	0.2534	0.0	0.2534
No log	9.5	152	0.2534	0.0	0.2534
No log	9.625	154	0.2536	0.0	0.2536
No log	9.75	156	0.2537	0.0	0.2538
No log	9.875	158	0.2538	0.0	0.2539
No log	10.0	160	0.2539	0.0	0.2539

Framework versions

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