SADAF_test3

This model is a fine-tuned version of aubmindlab/bert-base-arabertv02 on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 1.0061
Macro F1: 0.7951
Precision: 0.7874
Recall: 0.8073
Kappa: 0.7169
Accuracy: 0.8073

Model description

Relation identification for explicit dataset

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: 3e-05
train_batch_size: 32
eval_batch_size: 32
seed: 25
gradient_accumulation_steps: 2
total_train_batch_size: 64
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Macro F1	Precision	Recall	Kappa	Accuracy
No log	1.0	76	1.0304	0.6632	0.6222	0.7436	0.5881	0.7436
No log	2.0	152	0.8135	0.7191	0.6933	0.7800	0.6519	0.7800
No log	3.0	228	0.7417	0.7715	0.7663	0.8007	0.6973	0.8007
No log	4.0	304	0.7449	0.7807	0.7704	0.7957	0.6999	0.7957
No log	5.0	380	0.7447	0.7874	0.7770	0.8089	0.7128	0.8089
No log	6.0	456	0.8034	0.7654	0.7599	0.7750	0.6761	0.7750
0.7186	7.0	532	0.8874	0.7672	0.7669	0.7750	0.6785	0.7750
0.7186	8.0	608	0.8737	0.7830	0.7729	0.7974	0.7030	0.7974
0.7186	9.0	684	0.8964	0.7785	0.7675	0.7924	0.6978	0.7924
0.7186	10.0	760	0.9368	0.7863	0.7761	0.7998	0.7071	0.7998
0.7186	11.0	836	0.9717	0.7897	0.7803	0.8040	0.7119	0.8040
0.7186	12.0	912	0.9876	0.7883	0.7810	0.8007	0.7086	0.8007
0.7186	13.0	988	0.9893	0.7893	0.7812	0.8023	0.7106	0.8023
0.1542	14.0	1064	0.9999	0.7917	0.7841	0.8023	0.7109	0.8023
0.1542	15.0	1140	1.0061	0.7951	0.7874	0.8073	0.7169	0.8073

Framework versions

Transformers 4.28.1
Pytorch 2.0.0+cu118
Tokenizers 0.13.3