ArabicNewSplits5_FineTuningAraBERT_run3_AugV5_k1_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: 0.9056
• Qwk: 0.5054
• Mse: 0.9056
• Rmse: 0.9516

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	0.2	2	3.9906	0.0073	3.9906	1.9976
No log	0.4	4	2.1019	0.0432	2.1019	1.4498
No log	0.6	6	1.2545	0.0874	1.2545	1.1201
No log	0.8	8	0.9057	0.1679	0.9057	0.9517
No log	1.0	10	0.7165	0.2203	0.7165	0.8465
No log	1.2	12	0.6782	0.2414	0.6782	0.8235
No log	1.4	14	0.6686	0.2076	0.6686	0.8177
No log	1.6	16	0.7710	0.2599	0.7710	0.8781
No log	1.8	18	0.8177	0.3383	0.8177	0.9042
No log	2.0	20	0.8220	0.3640	0.8220	0.9066
No log	2.2	22	1.0171	0.2832	1.0171	1.0085
No log	2.4	24	0.8842	0.4010	0.8842	0.9403
No log	2.6	26	0.6205	0.4700	0.6205	0.7877
No log	2.8	28	0.6283	0.3826	0.6283	0.7926
No log	3.0	30	0.6533	0.3865	0.6533	0.8083
No log	3.2	32	0.6671	0.4984	0.6671	0.8167
No log	3.4	34	0.9910	0.4253	0.9910	0.9955
No log	3.6	36	1.0287	0.3973	1.0287	1.0143
No log	3.8	38	0.8114	0.5410	0.8114	0.9008
No log	4.0	40	0.7221	0.5083	0.7221	0.8497
No log	4.2	42	0.7996	0.4294	0.7996	0.8942
No log	4.4	44	0.8573	0.4131	0.8573	0.9259
No log	4.6	46	0.8773	0.5083	0.8773	0.9366
No log	4.8	48	0.9856	0.4749	0.9856	0.9928
No log	5.0	50	1.0273	0.4656	1.0273	1.0136
No log	5.2	52	1.0428	0.4479	1.0428	1.0212
No log	5.4	54	0.9264	0.4912	0.9264	0.9625
No log	5.6	56	0.8727	0.5069	0.8727	0.9342
No log	5.8	58	0.9282	0.4506	0.9282	0.9634
No log	6.0	60	0.9331	0.4432	0.9331	0.9660
No log	6.2	62	0.9129	0.4923	0.9129	0.9555
No log	6.4	64	0.8859	0.5198	0.8859	0.9412
No log	6.6	66	0.8706	0.5223	0.8706	0.9331
No log	6.8	68	0.8724	0.5018	0.8724	0.9340
No log	7.0	70	0.8922	0.4914	0.8922	0.9445
No log	7.2	72	0.9041	0.4891	0.9041	0.9508
No log	7.4	74	0.9034	0.5111	0.9034	0.9504
No log	7.6	76	0.9201	0.5184	0.9201	0.9592
No log	7.8	78	0.9357	0.4810	0.9357	0.9673
No log	8.0	80	0.9322	0.5042	0.9322	0.9655
No log	8.2	82	0.9252	0.5049	0.9252	0.9619
No log	8.4	84	0.9177	0.5121	0.9177	0.9579
No log	8.6	86	0.9070	0.5204	0.9070	0.9524
No log	8.8	88	0.8995	0.5123	0.8995	0.9484
No log	9.0	90	0.8955	0.4770	0.8955	0.9463
No log	9.2	92	0.9001	0.5194	0.9001	0.9488
No log	9.4	94	0.9051	0.5047	0.9051	0.9514
No log	9.6	96	0.9069	0.5054	0.9069	0.9523
No log	9.8	98	0.9063	0.5054	0.9063	0.9520
No log	10.0	100	0.9056	0.5054	0.9056	0.9516

Framework versions

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