arabert_baseline_mechanics_task2_fold1

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: 0.8879
• Qwk: 0.5758
• Mse: 0.8784

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	3.5975	-0.0382	3.6807
No log	0.6667	4	1.4679	0.1796	1.5092
No log	1.0	6	1.0780	0.0	1.0882
No log	1.3333	8	1.2308	0.1695	1.2193
No log	1.6667	10	1.2253	0.1695	1.2102
No log	2.0	12	0.8911	0.2451	0.8886
No log	2.3333	14	0.8137	0.3099	0.8139
No log	2.6667	16	0.8083	0.3099	0.8044
No log	3.0	18	0.9159	0.5185	0.9080
No log	3.3333	20	1.0465	0.4556	1.0353
No log	3.6667	22	0.9742	0.4731	0.9635
No log	4.0	24	0.7724	0.3897	0.7631
No log	4.3333	26	0.7534	0.3636	0.7425
No log	4.6667	28	0.7201	0.3636	0.7079
No log	5.0	30	0.8574	0.5586	0.8468
No log	5.3333	32	0.9644	0.5130	0.9546
No log	5.6667	34	0.8585	0.608	0.8487
No log	6.0	36	0.7248	0.5405	0.7112
No log	6.3333	38	0.7385	0.5405	0.7257
No log	6.6667	40	0.8072	0.5758	0.7967
No log	7.0	42	0.9702	0.6513	0.9631
No log	7.3333	44	0.9882	0.6513	0.9815
No log	7.6667	46	0.9719	0.6513	0.9651
No log	8.0	48	0.9498	0.6513	0.9425
No log	8.3333	50	0.9073	0.6513	0.8988
No log	8.6667	52	0.8907	0.5758	0.8816
No log	9.0	54	0.8763	0.5758	0.8665
No log	9.3333	56	0.8701	0.5758	0.8602
No log	9.6667	58	0.8850	0.5758	0.8754
No log	10.0	60	0.8879	0.5758	0.8784

Framework versions

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