ArabicNewSplits6_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: 0.6899
• Qwk: 0.2653
• Mse: 0.6899
• Rmse: 0.8306

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	3.1605	-0.0205	3.1605	1.7778
No log	0.3333	4	1.7883	0.0130	1.7883	1.3373
No log	0.5	6	0.9545	0.0236	0.9545	0.9770
No log	0.6667	8	1.1878	0.1235	1.1878	1.0899
No log	0.8333	10	0.6158	0.1373	0.6158	0.7848
No log	1.0	12	0.5992	0.0	0.5992	0.7741
No log	1.1667	14	0.5703	0.0	0.5703	0.7552
No log	1.3333	16	0.5445	0.0569	0.5445	0.7379
No log	1.5	18	0.8664	0.0288	0.8664	0.9308
No log	1.6667	20	1.1357	0.0698	1.1357	1.0657
No log	1.8333	22	1.0508	0.0698	1.0508	1.0251
No log	2.0	24	0.9151	0.0431	0.9151	0.9566
No log	2.1667	26	0.6692	0.25	0.6692	0.8180
No log	2.3333	28	0.5555	0.0	0.5555	0.7453
No log	2.5	30	0.5779	0.0	0.5779	0.7602
No log	2.6667	32	0.6066	0.0	0.6066	0.7788
No log	2.8333	34	0.5824	0.0569	0.5824	0.7631
No log	3.0	36	0.6225	-0.0556	0.6225	0.7890
No log	3.1667	38	1.1674	0.0843	1.1674	1.0804
No log	3.3333	40	1.6923	0.0617	1.6923	1.3009
No log	3.5	42	1.3382	0.0877	1.3382	1.1568
No log	3.6667	44	0.7521	0.1515	0.7521	0.8673
No log	3.8333	46	0.6022	0.0388	0.6022	0.7760
No log	4.0	48	0.6471	0.0476	0.6471	0.8044
No log	4.1667	50	0.6674	0.0569	0.6674	0.8170
No log	4.3333	52	0.6210	0.0569	0.6210	0.7881
No log	4.5	54	0.5435	0.0476	0.5435	0.7372
No log	4.6667	56	0.5913	0.1030	0.5913	0.7690
No log	4.8333	58	0.8561	0.1861	0.8561	0.9253
No log	5.0	60	1.1137	0.1191	1.1137	1.0553
No log	5.1667	62	0.9960	0.1756	0.9960	0.9980
No log	5.3333	64	0.6891	0.1765	0.6891	0.8301
No log	5.5	66	0.5479	0.0857	0.5479	0.7402
No log	5.6667	68	0.7322	0.0588	0.7322	0.8557
No log	5.8333	70	0.8445	-0.1429	0.8445	0.9190
No log	6.0	72	0.8287	-0.1667	0.8287	0.9103
No log	6.1667	74	0.7007	0.0388	0.7007	0.8371
No log	6.3333	76	0.5826	0.0857	0.5826	0.7633
No log	6.5	78	0.5748	0.1020	0.5748	0.7582
No log	6.6667	80	0.5849	0.0769	0.5849	0.7648
No log	6.8333	82	0.6411	0.2444	0.6411	0.8007
No log	7.0	84	0.6723	0.2350	0.6723	0.8199
No log	7.1667	86	0.6289	0.2273	0.6289	0.7931
No log	7.3333	88	0.6245	0.1724	0.6245	0.7903
No log	7.5	90	0.6756	0.2857	0.6756	0.8220
No log	7.6667	92	0.7378	0.2000	0.7378	0.8590
No log	7.8333	94	0.8199	0.0053	0.8199	0.9055
No log	8.0	96	0.8632	0.0151	0.8632	0.9291
No log	8.1667	98	0.8436	0.0053	0.8436	0.9185
No log	8.3333	100	0.8032	0.2000	0.8032	0.8962
No log	8.5	102	0.7482	0.1675	0.7482	0.8650
No log	8.6667	104	0.7321	0.1921	0.7321	0.8556
No log	8.8333	106	0.7217	0.2161	0.7217	0.8495
No log	9.0	108	0.7065	0.2000	0.7065	0.8405
No log	9.1667	110	0.6956	0.2390	0.6956	0.8340
No log	9.3333	112	0.6897	0.2390	0.6897	0.8305
No log	9.5	114	0.6861	0.2000	0.6861	0.8283
No log	9.6667	116	0.6862	0.2390	0.6862	0.8284
No log	9.8333	118	0.6886	0.2653	0.6886	0.8298
No log	10.0	120	0.6899	0.2653	0.6899	0.8306

Framework versions

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