Arabic_FineTuningAraBERT_noAug_task3_organization_fold0

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: 0.9261
• Qwk: -0.1786
• Mse: 0.9261
• Rmse: 0.9623

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	0.3333	2	5.0795	-0.0373	5.0795	2.2538
No log	0.6667	4	3.7935	-0.0292	3.7935	1.9477
No log	1.0	6	1.3848	-0.0722	1.3848	1.1768
No log	1.3333	8	0.8755	-0.0087	0.8755	0.9357
No log	1.6667	10	0.8313	-0.1786	0.8313	0.9118
No log	2.0	12	0.9317	-0.1786	0.9317	0.9653
No log	2.3333	14	0.8420	0.0	0.8420	0.9176
No log	2.6667	16	0.7754	0.0	0.7754	0.8805
No log	3.0	18	0.7847	0.2029	0.7847	0.8858
No log	3.3333	20	0.8265	-0.0087	0.8265	0.9091
No log	3.6667	22	0.8339	0.0737	0.8339	0.9132
No log	4.0	24	0.8201	-0.2384	0.8201	0.9056
No log	4.3333	26	0.8259	0.0	0.8259	0.9088
No log	4.6667	28	0.9062	0.0	0.9062	0.9519
No log	5.0	30	0.8545	0.0	0.8545	0.9244
No log	5.3333	32	0.8293	-0.0565	0.8293	0.9106
No log	5.6667	34	0.8200	-0.0732	0.8200	0.9056
No log	6.0	36	0.9028	0.0	0.9028	0.9502
No log	6.3333	38	1.0402	0.0	1.0402	1.0199
No log	6.6667	40	1.0327	0.0	1.0327	1.0162
No log	7.0	42	0.9744	0.0	0.9744	0.9871
No log	7.3333	44	0.8963	0.0	0.8963	0.9467
No log	7.6667	46	0.8552	0.0530	0.8552	0.9248
No log	8.0	48	0.8597	0.0530	0.8597	0.9272
No log	8.3333	50	0.8662	0.0530	0.8662	0.9307
No log	8.6667	52	0.8815	-0.0732	0.8815	0.9389
No log	9.0	54	0.9075	-0.1786	0.9075	0.9526
No log	9.3333	56	0.9180	-0.1786	0.9180	0.9581
No log	9.6667	58	0.9234	-0.1786	0.9234	0.9609
No log	10.0	60	0.9261	-0.1786	0.9261	0.9623

Framework versions

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