arabert_baseline_grammar_task2_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: 1.2738
• Qwk: 0.0491
• Mse: 1.2508

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: 16
• eval_batch_size: 16
• 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
No log	0.3333	2	4.9005	0.0126	4.8502
No log	0.6667	4	2.4699	0.0387	2.4471
No log	1.0	6	1.4800	-0.0982	1.4570
No log	1.3333	8	1.0496	0.0541	1.0294
No log	1.6667	10	1.2284	0.0	1.2139
No log	2.0	12	1.4327	0.0	1.4198
No log	2.3333	14	1.4042	0.0	1.3911
No log	2.6667	16	1.1140	-0.0161	1.0974
No log	3.0	18	1.0138	0.1083	0.9870
No log	3.3333	20	1.0084	0.2200	0.9831
No log	3.6667	22	1.0806	0.0696	1.0592
No log	4.0	24	1.2446	-0.1170	1.2246
No log	4.3333	26	1.4549	-0.0960	1.4445
No log	4.6667	28	1.3332	0.0585	1.3205
No log	5.0	30	1.0953	0.0870	1.0641
No log	5.3333	32	1.0705	0.0616	1.0381
No log	5.6667	34	1.1503	0.1064	1.1300
No log	6.0	36	1.2818	0.0585	1.2670
No log	6.3333	38	1.3332	-0.0161	1.3186
No log	6.6667	40	1.2686	0.0491	1.2516
No log	7.0	42	1.1870	0.0491	1.1658
No log	7.3333	44	1.1491	0.0491	1.1243
No log	7.6667	46	1.1946	0.0491	1.1709
No log	8.0	48	1.3196	0.0491	1.3003
No log	8.3333	50	1.3925	0.0541	1.3747
No log	8.6667	52	1.3818	0.0541	1.3633
No log	9.0	54	1.3491	0.0491	1.3293
No log	9.3333	56	1.3127	0.0491	1.2916
No log	9.6667	58	1.2864	0.0491	1.2640
No log	10.0	60	1.2738	0.0491	1.2508

Framework versions

• Transformers 4.44.0
• Pytorch 2.4.0
• Datasets 2.21.0
• Tokenizers 0.19.1