ArabicNewSplits4_WithDuplicationsForScore5_FineTuningAraBERT_run2_AugV5_k2_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: 0.8949
• Qwk: 0.2258
• Mse: 0.8949
• Rmse: 0.9460

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	0.1667	2	3.5466	-0.0066	3.5466	1.8832
No log	0.3333	4	1.9614	-0.0370	1.9614	1.4005
No log	0.5	6	1.0359	0.0118	1.0359	1.0178
No log	0.6667	8	0.8280	0.1515	0.8280	0.9100
No log	0.8333	10	0.5973	0.2418	0.5973	0.7729
No log	1.0	12	0.5649	0.0476	0.5649	0.7516
No log	1.1667	14	0.7553	0.0080	0.7553	0.8691
No log	1.3333	16	0.7238	0.0080	0.7238	0.8507
No log	1.5	18	0.6471	0.0	0.6471	0.8044
No log	1.6667	20	0.5890	0.0	0.5890	0.7675
No log	1.8333	22	0.5679	-0.0081	0.5679	0.7536
No log	2.0	24	0.6321	0.2350	0.6321	0.7951
No log	2.1667	26	0.7411	0.0918	0.7411	0.8609
No log	2.3333	28	0.7005	0.1919	0.7005	0.8370
No log	2.5	30	0.5896	0.2308	0.5896	0.7678
No log	2.6667	32	0.6600	0.0	0.6600	0.8124
No log	2.8333	34	0.7186	-0.0732	0.7186	0.8477
No log	3.0	36	0.6905	0.0	0.6905	0.8310
No log	3.1667	38	0.5803	-0.0233	0.5803	0.7618
No log	3.3333	40	0.6392	0.2184	0.6392	0.7995
No log	3.5	42	0.9462	0.0476	0.9462	0.9727
No log	3.6667	44	0.7689	0.1841	0.7689	0.8769
No log	3.8333	46	0.5934	0.1020	0.5934	0.7703
No log	4.0	48	0.6640	0.0769	0.6640	0.8149
No log	4.1667	50	0.6332	0.0728	0.6332	0.7957
No log	4.3333	52	0.6439	0.0769	0.6439	0.8025
No log	4.5	54	0.5927	0.1020	0.5927	0.7699
No log	4.6667	56	0.5994	0.3103	0.5994	0.7742
No log	4.8333	58	0.5805	0.0850	0.5805	0.7619
No log	5.0	60	0.6371	0.1304	0.6371	0.7982
No log	5.1667	62	0.6370	0.1304	0.6370	0.7981
No log	5.3333	64	0.6429	0.2189	0.6429	0.8018
No log	5.5	66	0.7203	0.2233	0.7203	0.8487
No log	5.6667	68	0.6949	0.2811	0.6949	0.8336
No log	5.8333	70	0.6353	0.3469	0.6353	0.7970
No log	6.0	72	0.6875	0.3363	0.6875	0.8291
No log	6.1667	74	0.7795	0.1730	0.7795	0.8829
No log	6.3333	76	0.7757	0.2333	0.7757	0.8807
No log	6.5	78	0.8028	0.2405	0.8028	0.8960
No log	6.6667	80	0.9168	0.2353	0.9168	0.9575
No log	6.8333	82	0.9588	0.2353	0.9588	0.9792
No log	7.0	84	1.0012	0.2340	1.0012	1.0006
No log	7.1667	86	1.0044	0.2340	1.0044	1.0022
No log	7.3333	88	1.1181	0.1515	1.1181	1.0574
No log	7.5	90	1.1214	0.1788	1.1214	1.0590
No log	7.6667	92	0.9673	0.2659	0.9673	0.9835
No log	7.8333	94	0.8592	0.2863	0.8592	0.9269
No log	8.0	96	0.9054	0.2314	0.9054	0.9515
No log	8.1667	98	1.0474	0.2340	1.0474	1.0234
No log	8.3333	100	1.0908	0.1781	1.0908	1.0444
No log	8.5	102	1.2143	0.1304	1.2143	1.1020
No log	8.6667	104	1.2963	0.1373	1.2963	1.1385
No log	8.8333	106	1.2818	0.1362	1.2818	1.1322
No log	9.0	108	1.1869	0.1523	1.1869	1.0895
No log	9.1667	110	1.0807	0.2347	1.0807	1.0396
No log	9.3333	112	0.9881	0.2296	0.9881	0.9940
No log	9.5	114	0.9212	0.2558	0.9212	0.9598
No log	9.6667	116	0.8890	0.2263	0.8890	0.9429
No log	9.8333	118	0.8888	0.2263	0.8888	0.9428
No log	10.0	120	0.8949	0.2258	0.8949	0.9460

Framework versions

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