ArabicNewSplits6_WithDuplicationsForScore5_FineTuningAraBERT_run2_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: 1.1421
• Qwk: 0.4812
• Mse: 1.1421
• Rmse: 1.0687

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	3.9362	-0.0062	3.9362	1.9840
No log	0.2667	4	1.9904	0.0977	1.9904	1.4108
No log	0.4	6	1.0441	0.0666	1.0441	1.0218
No log	0.5333	8	0.7309	0.1194	0.7309	0.8550
No log	0.6667	10	0.7325	0.1404	0.7325	0.8558
No log	0.8	12	0.8429	0.0472	0.8429	0.9181
No log	0.9333	14	0.8978	-0.0775	0.8978	0.9475
No log	1.0667	16	0.7873	0.1119	0.7873	0.8873
No log	1.2	18	0.7568	0.1189	0.7568	0.8699
No log	1.3333	20	0.7336	0.1657	0.7336	0.8565
No log	1.4667	22	0.7128	0.2019	0.7128	0.8443
No log	1.6	24	0.7258	0.1546	0.7258	0.8519
No log	1.7333	26	0.7906	0.1118	0.7906	0.8892
No log	1.8667	28	0.7999	0.1652	0.7999	0.8944
No log	2.0	30	0.7153	0.2703	0.7153	0.8458
No log	2.1333	32	0.5813	0.3801	0.5813	0.7624
No log	2.2667	34	0.5966	0.3476	0.5966	0.7724
No log	2.4	36	0.5757	0.3897	0.5757	0.7587
No log	2.5333	38	0.5909	0.4719	0.5909	0.7687
No log	2.6667	40	0.6521	0.4404	0.6521	0.8076
No log	2.8	42	0.6583	0.4648	0.6583	0.8114
No log	2.9333	44	0.6864	0.5159	0.6864	0.8285
No log	3.0667	46	0.7592	0.4675	0.7592	0.8713
No log	3.2	48	0.8560	0.4557	0.8560	0.9252
No log	3.3333	50	0.9070	0.4299	0.9070	0.9523
No log	3.4667	52	0.8995	0.4630	0.8995	0.9484
No log	3.6	54	0.7296	0.5374	0.7296	0.8542
No log	3.7333	56	0.6794	0.5220	0.6794	0.8243
No log	3.8667	58	0.6453	0.5093	0.6453	0.8033
No log	4.0	60	0.7372	0.5293	0.7372	0.8586
No log	4.1333	62	0.8609	0.4943	0.8609	0.9279
No log	4.2667	64	0.7710	0.5169	0.7710	0.8781
No log	4.4	66	0.6420	0.4815	0.6420	0.8013
No log	4.5333	68	0.6617	0.5303	0.6617	0.8135
No log	4.6667	70	0.7152	0.4983	0.7152	0.8457
No log	4.8	72	0.9293	0.4855	0.9293	0.9640
No log	4.9333	74	1.0223	0.4604	1.0223	1.0111
No log	5.0667	76	0.9887	0.4799	0.9887	0.9943
No log	5.2	78	0.9308	0.4941	0.9308	0.9648
No log	5.3333	80	0.8706	0.5073	0.8706	0.9331
No log	5.4667	82	0.8465	0.5527	0.8465	0.9201
No log	5.6	84	0.8758	0.5051	0.8758	0.9358
No log	5.7333	86	1.0527	0.4979	1.0527	1.0260
No log	5.8667	88	1.1641	0.4683	1.1641	1.0789
No log	6.0	90	1.1436	0.4629	1.1436	1.0694
No log	6.1333	92	0.9992	0.5298	0.9992	0.9996
No log	6.2667	94	0.7450	0.5058	0.7450	0.8631
No log	6.4	96	0.7016	0.5510	0.7016	0.8376
No log	6.5333	98	0.7280	0.54	0.7280	0.8532
No log	6.6667	100	0.9098	0.5007	0.9098	0.9538
No log	6.8	102	1.0648	0.4903	1.0648	1.0319
No log	6.9333	104	0.9864	0.4777	0.9864	0.9932
No log	7.0667	106	0.8954	0.5311	0.8954	0.9463
No log	7.2	108	0.8353	0.5598	0.8353	0.9140
No log	7.3333	110	0.8430	0.5526	0.8430	0.9181
No log	7.4667	112	0.8475	0.5428	0.8475	0.9206
No log	7.6	114	0.9304	0.5313	0.9304	0.9646
No log	7.7333	116	1.0578	0.4826	1.0578	1.0285
No log	7.8667	118	1.2319	0.4975	1.2319	1.1099
No log	8.0	120	1.4590	0.4054	1.4590	1.2079
No log	8.1333	122	1.5472	0.4061	1.5472	1.2439
No log	8.2667	124	1.4862	0.4027	1.4862	1.2191
No log	8.4	126	1.3318	0.4530	1.3318	1.1540
No log	8.5333	128	1.1400	0.4894	1.1400	1.0677
No log	8.6667	130	0.9978	0.5197	0.9978	0.9989
No log	8.8	132	0.9552	0.5157	0.9552	0.9773
No log	8.9333	134	0.9401	0.5155	0.9401	0.9696
No log	9.0667	136	0.9700	0.5242	0.9700	0.9849
No log	9.2	138	1.0054	0.4929	1.0054	1.0027
No log	9.3333	140	1.0466	0.5004	1.0466	1.0230
No log	9.4667	142	1.0948	0.4963	1.0948	1.0463
No log	9.6	144	1.1347	0.4947	1.1347	1.0652
No log	9.7333	146	1.1428	0.4812	1.1428	1.0690
No log	9.8667	148	1.1416	0.4812	1.1416	1.0685
No log	10.0	150	1.1421	0.4812	1.1421	1.0687

Framework versions

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