ArabicNewSplits6_WithDuplicationsForScore5_FineTuningAraBERT_run3_AugV5_k2_task2_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: 0.8385
• Qwk: 0.5594
• Mse: 0.8385
• Rmse: 0.9157

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	0.1333	2	3.9412	0.0017	3.9412	1.9852
No log	0.2667	4	2.0766	0.0432	2.0766	1.4410
No log	0.4	6	1.3380	0.0959	1.3380	1.1567
No log	0.5333	8	0.8377	-0.0133	0.8377	0.9153
No log	0.6667	10	0.7465	0.1443	0.7465	0.8640
No log	0.8	12	0.7175	0.1934	0.7175	0.8470
No log	0.9333	14	0.7785	0.1442	0.7785	0.8823
No log	1.0667	16	0.8947	0.1276	0.8947	0.9459
No log	1.2	18	0.8350	0.0918	0.8350	0.9138
No log	1.3333	20	0.7539	0.0918	0.7539	0.8683
No log	1.4667	22	0.6500	0.1404	0.6500	0.8062
No log	1.6	24	0.6132	0.2941	0.6132	0.7830
No log	1.7333	26	0.6068	0.3361	0.6068	0.7790
No log	1.8667	28	0.5960	0.3361	0.5960	0.7720
No log	2.0	30	0.5688	0.3520	0.5688	0.7542
No log	2.1333	32	0.5999	0.2548	0.5999	0.7745
No log	2.2667	34	0.6688	0.2838	0.6688	0.8178
No log	2.4	36	0.6558	0.4058	0.6558	0.8098
No log	2.5333	38	0.5997	0.4895	0.5997	0.7744
No log	2.6667	40	0.6403	0.5050	0.6403	0.8002
No log	2.8	42	0.6230	0.5347	0.6230	0.7893
No log	2.9333	44	0.6096	0.4884	0.6096	0.7808
No log	3.0667	46	0.6223	0.5050	0.6223	0.7889
No log	3.2	48	0.6418	0.4737	0.6418	0.8011
No log	3.3333	50	0.6297	0.5559	0.6297	0.7935
No log	3.4667	52	0.5937	0.5547	0.5937	0.7705
No log	3.6	54	0.5952	0.4922	0.5952	0.7715
No log	3.7333	56	0.6508	0.5105	0.6508	0.8067
No log	3.8667	58	0.7818	0.4654	0.7818	0.8842
No log	4.0	60	0.7153	0.5371	0.7153	0.8457
No log	4.1333	62	0.6383	0.5326	0.6383	0.7989
No log	4.2667	64	0.7425	0.5231	0.7425	0.8617
No log	4.4	66	0.8569	0.5348	0.8569	0.9257
No log	4.5333	68	0.9525	0.5335	0.9525	0.9760
No log	4.6667	70	0.9923	0.5215	0.9923	0.9962
No log	4.8	72	0.9477	0.5378	0.9477	0.9735
No log	4.9333	74	0.9024	0.4834	0.9024	0.9499
No log	5.0667	76	0.9449	0.4804	0.9449	0.9721
No log	5.2	78	0.9041	0.5085	0.9041	0.9508
No log	5.3333	80	0.7393	0.5449	0.7393	0.8598
No log	5.4667	82	0.6713	0.5494	0.6713	0.8193
No log	5.6	84	0.6774	0.5481	0.6774	0.8230
No log	5.7333	86	0.6457	0.5453	0.6457	0.8036
No log	5.8667	88	0.6869	0.5855	0.6869	0.8288
No log	6.0	90	0.7758	0.5882	0.7758	0.8808
No log	6.1333	92	0.8270	0.5691	0.8270	0.9094
No log	6.2667	94	0.7774	0.5778	0.7774	0.8817
No log	6.4	96	0.6858	0.5867	0.6858	0.8281
No log	6.5333	98	0.6735	0.5337	0.6735	0.8207
No log	6.6667	100	0.6864	0.5178	0.6864	0.8285
No log	6.8	102	0.6798	0.5512	0.6798	0.8245
No log	6.9333	104	0.6915	0.5286	0.6915	0.8316
No log	7.0667	106	0.7117	0.5336	0.7117	0.8436
No log	7.2	108	0.7486	0.5281	0.7486	0.8652
No log	7.3333	110	0.7557	0.5369	0.7557	0.8693
No log	7.4667	112	0.7662	0.5573	0.7662	0.8753
No log	7.6	114	0.8247	0.5811	0.8247	0.9082
No log	7.7333	116	0.8635	0.6026	0.8635	0.9293
No log	7.8667	118	0.8598	0.6011	0.8598	0.9272
No log	8.0	120	0.8234	0.5823	0.8234	0.9074
No log	8.1333	122	0.7858	0.5837	0.7858	0.8864
No log	8.2667	124	0.7721	0.5332	0.7721	0.8787
No log	8.4	126	0.7910	0.5230	0.7910	0.8894
No log	8.5333	128	0.8224	0.5073	0.8224	0.9068
No log	8.6667	130	0.8562	0.54	0.8562	0.9253
No log	8.8	132	0.8571	0.5388	0.8571	0.9258
No log	8.9333	134	0.8415	0.5087	0.8415	0.9173
No log	9.0667	136	0.8391	0.5160	0.8391	0.9160
No log	9.2	138	0.8389	0.5439	0.8389	0.9159
No log	9.3333	140	0.8401	0.5561	0.8401	0.9166
No log	9.4667	142	0.8415	0.565	0.8415	0.9174
No log	9.6	144	0.8418	0.5583	0.8418	0.9175
No log	9.7333	146	0.8399	0.5594	0.8399	0.9165
No log	9.8667	148	0.8388	0.5594	0.8388	0.9159
No log	10.0	150	0.8385	0.5594	0.8385	0.9157

Framework versions

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