ArabicNewSplits6_FineTuningAraBERT_run3_AugV5_k3_task3_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: 0.9390
• Qwk: 0.2000
• Mse: 0.9390
• Rmse: 0.9690

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	0.125	2	3.0932	-0.0238	3.0932	1.7587
No log	0.25	4	1.4104	0.0255	1.4104	1.1876
No log	0.375	6	0.7489	0.0588	0.7489	0.8654
No log	0.5	8	0.6723	0.1345	0.6723	0.8199
No log	0.625	10	0.6006	0.0569	0.6006	0.7750
No log	0.75	12	0.5864	0.0569	0.5864	0.7658
No log	0.875	14	0.6283	0.2308	0.6283	0.7927
No log	1.0	16	0.8652	0.0823	0.8652	0.9302
No log	1.125	18	0.6676	0.1724	0.6676	0.8171
No log	1.25	20	0.6142	-0.0732	0.6142	0.7837
No log	1.375	22	0.6304	-0.0732	0.6304	0.7940
No log	1.5	24	0.6883	-0.0133	0.6883	0.8297
No log	1.625	26	0.6700	0.0	0.6700	0.8185
No log	1.75	28	0.6536	-0.0794	0.6536	0.8085
No log	1.875	30	0.6686	-0.0081	0.6686	0.8177
No log	2.0	32	0.6496	0.1282	0.6496	0.8060
No log	2.125	34	0.9088	0.1150	0.9088	0.9533
No log	2.25	36	0.7310	0.1648	0.7310	0.8550
No log	2.375	38	0.7294	0.0303	0.7294	0.8541
No log	2.5	40	0.7875	-0.0081	0.7875	0.8874
No log	2.625	42	0.7300	0.0476	0.7300	0.8544
No log	2.75	44	0.6438	-0.0435	0.6438	0.8024
No log	2.875	46	0.7291	0.2444	0.7291	0.8539
No log	3.0	48	0.6863	0.1429	0.6863	0.8284
No log	3.125	50	0.6180	0.0303	0.6180	0.7862
No log	3.25	52	0.5979	0.0303	0.5979	0.7732
No log	3.375	54	0.6595	0.2251	0.6595	0.8121
No log	3.5	56	0.6868	0.2487	0.6868	0.8288
No log	3.625	58	0.5663	0.1020	0.5663	0.7526
No log	3.75	60	0.6383	0.1533	0.6383	0.7990
No log	3.875	62	0.7027	0.1813	0.7027	0.8382
No log	4.0	64	0.7696	0.2421	0.7696	0.8772
No log	4.125	66	0.6654	0.1818	0.6654	0.8157
No log	4.25	68	0.7220	0.2621	0.7220	0.8497
No log	4.375	70	0.6577	0.2251	0.6577	0.8110
No log	4.5	72	0.6697	0.1807	0.6697	0.8184
No log	4.625	74	0.7483	0.1323	0.7483	0.8651
No log	4.75	76	0.6575	0.1111	0.6575	0.8108
No log	4.875	78	0.6422	0.2179	0.6422	0.8014
No log	5.0	80	0.8053	0.2137	0.8053	0.8974
No log	5.125	82	0.8208	0.1861	0.8208	0.9060
No log	5.25	84	0.8139	0.1790	0.8139	0.9022
No log	5.375	86	0.8721	0.1535	0.8721	0.9338
No log	5.5	88	0.9447	0.1873	0.9447	0.9719
No log	5.625	90	1.0101	0.1756	1.0101	1.0051
No log	5.75	92	0.8783	0.1811	0.8783	0.9372
No log	5.875	94	0.7400	0.2727	0.7400	0.8602
No log	6.0	96	0.7356	0.3303	0.7356	0.8577
No log	6.125	98	0.7805	0.2000	0.7805	0.8835
No log	6.25	100	1.1496	0.0355	1.1496	1.0722
No log	6.375	102	1.3139	0.0307	1.3139	1.1462
No log	6.5	104	1.1224	0.0418	1.1224	1.0594
No log	6.625	106	0.7571	0.2489	0.7571	0.8701
No log	6.75	108	0.6865	0.2072	0.6865	0.8286
No log	6.875	110	0.8479	0.3220	0.8479	0.9208
No log	7.0	112	1.2002	0.1111	1.2002	1.0955
No log	7.125	114	1.2852	0.0556	1.2852	1.1337
No log	7.25	116	1.2347	0.0556	1.2347	1.1112
No log	7.375	118	0.9844	0.1027	0.9844	0.9922
No log	7.5	120	0.6955	0.2233	0.6955	0.8340
No log	7.625	122	0.6003	0.1739	0.6003	0.7748
No log	7.75	124	0.6015	0.1739	0.6015	0.7756
No log	7.875	126	0.6648	0.1323	0.6648	0.8153
No log	8.0	128	0.8106	0.2511	0.8106	0.9003
No log	8.125	130	1.0146	0.1290	1.0146	1.0073
No log	8.25	132	1.0888	0.1062	1.0888	1.0435
No log	8.375	134	1.0415	0.1318	1.0415	1.0205
No log	8.5	136	0.8897	0.2134	0.8897	0.9433
No log	8.625	138	0.7472	0.2269	0.7472	0.8644
No log	8.75	140	0.6317	0.2000	0.6317	0.7948
No log	8.875	142	0.6041	0.1556	0.6041	0.7772
No log	9.0	144	0.6116	0.2000	0.6116	0.7820
No log	9.125	146	0.6515	0.2372	0.6515	0.8072
No log	9.25	148	0.7231	0.2605	0.7231	0.8503
No log	9.375	150	0.8159	0.2885	0.8159	0.9033
No log	9.5	152	0.9018	0.2320	0.9018	0.9496
No log	9.625	154	0.9407	0.2000	0.9407	0.9699
No log	9.75	156	0.9465	0.2000	0.9465	0.9729
No log	9.875	158	0.9394	0.2000	0.9394	0.9692
No log	10.0	160	0.9390	0.2000	0.9390	0.9690

Framework versions

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