ArabicNewSplits5_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: 0.7619
• Qwk: 0.3080
• Mse: 0.7619
• Rmse: 0.8729

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	3.4300	-0.0160	3.4300	1.8520
No log	0.2667	4	1.8515	-0.0130	1.8515	1.3607
No log	0.4	6	1.2353	0.0588	1.2353	1.1115
No log	0.5333	8	0.9459	0.0335	0.9459	0.9726
No log	0.6667	10	1.1278	0.0255	1.1278	1.0620
No log	0.8	12	1.9985	0.0788	1.9985	1.4137
No log	0.9333	14	1.4061	0.0255	1.4061	1.1858
No log	1.0667	16	1.2688	0.0	1.2688	1.1264
No log	1.2	18	1.0546	0.0	1.0546	1.0269
No log	1.3333	20	0.8166	0.0698	0.8166	0.9037
No log	1.4667	22	0.7717	0.1486	0.7717	0.8785
No log	1.6	24	0.6851	0.3086	0.6851	0.8277
No log	1.7333	26	0.6032	0.0909	0.6032	0.7767
No log	1.8667	28	0.6480	0.2941	0.6480	0.8050
No log	2.0	30	0.7680	0.1861	0.7680	0.8763
No log	2.1333	32	0.7549	0.1861	0.7549	0.8688
No log	2.2667	34	0.8014	0.1934	0.8014	0.8952
No log	2.4	36	0.7583	0.1861	0.7583	0.8708
No log	2.5333	38	0.6885	0.2688	0.6885	0.8298
No log	2.6667	40	0.6992	0.2593	0.6992	0.8362
No log	2.8	42	0.6328	0.1282	0.6328	0.7955
No log	2.9333	44	0.6889	0.0545	0.6889	0.8300
No log	3.0667	46	0.6816	0.0545	0.6816	0.8256
No log	3.2	48	0.6693	-0.0719	0.6693	0.8181
No log	3.3333	50	0.6656	0.0145	0.6656	0.8159
No log	3.4667	52	0.7642	0.0455	0.7642	0.8742
No log	3.6	54	0.7436	-0.0299	0.7436	0.8623
No log	3.7333	56	0.7400	0.0388	0.7400	0.8602
No log	3.8667	58	0.7491	-0.0303	0.7491	0.8655
No log	4.0	60	0.8199	0.0439	0.8199	0.9055
No log	4.1333	62	0.9522	0.0909	0.9522	0.9758
No log	4.2667	64	0.7046	0.0256	0.7046	0.8394
No log	4.4	66	0.7353	0.0388	0.7353	0.8575
No log	4.5333	68	0.7482	0.1655	0.7482	0.8650
No log	4.6667	70	0.7252	0.2000	0.7252	0.8516
No log	4.8	72	1.0154	0.2441	1.0154	1.0077
No log	4.9333	74	1.0923	0.2548	1.0923	1.0451
No log	5.0667	76	1.0033	0.2191	1.0033	1.0016
No log	5.2	78	0.6948	0.2233	0.6948	0.8336
No log	5.3333	80	0.7000	0.2661	0.7000	0.8367
No log	5.4667	82	0.8677	0.0894	0.8677	0.9315
No log	5.6	84	0.9197	0.0722	0.9197	0.9590
No log	5.7333	86	0.7566	0.2432	0.7566	0.8698
No log	5.8667	88	0.6486	0.3153	0.6486	0.8053
No log	6.0	90	0.7266	0.3188	0.7266	0.8524
No log	6.1333	92	0.6351	0.3448	0.6351	0.7969
No log	6.2667	94	0.6751	0.3846	0.6751	0.8216
No log	6.4	96	0.6798	0.3171	0.6798	0.8245
No log	6.5333	98	0.6090	0.3684	0.6090	0.7804
No log	6.6667	100	0.6550	0.4087	0.6550	0.8093
No log	6.8	102	0.6531	0.3991	0.6531	0.8081
No log	6.9333	104	0.5942	0.3645	0.5942	0.7708
No log	7.0667	106	0.6400	0.3744	0.6400	0.8000
No log	7.2	108	0.6507	0.2676	0.6507	0.8067
No log	7.3333	110	0.6583	0.2676	0.6583	0.8113
No log	7.4667	112	0.6763	0.2676	0.6763	0.8224
No log	7.6	114	0.6567	0.3422	0.6567	0.8103
No log	7.7333	116	0.6407	0.3939	0.6407	0.8004
No log	7.8667	118	0.6331	0.3480	0.6331	0.7956
No log	8.0	120	0.6429	0.3504	0.6429	0.8018
No log	8.1333	122	0.6617	0.3939	0.6617	0.8134
No log	8.2667	124	0.6955	0.2963	0.6955	0.8340
No log	8.4	126	0.6852	0.3739	0.6852	0.8278
No log	8.5333	128	0.6644	0.3418	0.6644	0.8151
No log	8.6667	130	0.6714	0.3333	0.6714	0.8194
No log	8.8	132	0.6998	0.3220	0.6998	0.8366
No log	8.9333	134	0.7623	0.2920	0.7623	0.8731
No log	9.0667	136	0.8286	0.2711	0.8286	0.9103
No log	9.2	138	0.8591	0.2432	0.8591	0.9269
No log	9.3333	140	0.8670	0.2432	0.8670	0.9311
No log	9.4667	142	0.8509	0.2432	0.8509	0.9225
No log	9.6	144	0.8185	0.2711	0.8185	0.9047
No log	9.7333	146	0.7897	0.2554	0.7897	0.8886
No log	9.8667	148	0.7689	0.3080	0.7689	0.8769
No log	10.0	150	0.7619	0.3080	0.7619	0.8729

Framework versions

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