eraydikyologlu/bert_ayt_fizik

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

• Train Loss: 0.2037
• Train Accuracy: 0.9634
• Validation Loss: 0.1170
• Validation Accuracy: 0.9784
• Epoch: 18

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:

• optimizer: {'name': 'Adam', 'weight_decay': None, 'clipnorm': None, 'global_clipnorm': None, 'clipvalue': None, 'use_ema': False, 'ema_momentum': 0.99, 'ema_overwrite_frequency': None, 'jit_compile': True, 'is_legacy_optimizer': False, 'learning_rate': {'module': 'transformers.optimization_tf', 'class_name': 'WarmUp', 'config': {'initial_learning_rate': 2e-05, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 2e-05, 'decay_steps': 4770, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 530, 'power': 1.0, 'name': None}, 'registered_name': 'WarmUp'}, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
4.5820	0.0303	4.3223	0.0817	0
3.4110	0.2978	2.3701	0.4760	1
2.0594	0.5300	1.5347	0.5938	2
1.4984	0.6083	1.1782	0.6526	3
1.2008	0.6594	0.9504	0.7043	4
1.0088	0.7080	0.7924	0.7536	5
0.8641	0.7486	0.6628	0.8089	6
0.7482	0.7838	0.5492	0.8522	7
0.6515	0.8144	0.4472	0.8786	8
0.5631	0.8435	0.3810	0.8966	9
0.4869	0.8695	0.3191	0.9062	10
0.4241	0.8928	0.2604	0.9291	11
0.3696	0.9075	0.2225	0.9519	12
0.3252	0.9258	0.1905	0.9591	13
0.2845	0.9367	0.1612	0.9736	14
0.2607	0.9423	0.1430	0.9820	15
0.2336	0.9545	0.1307	0.9772	16
0.2150	0.9586	0.1225	0.9748	17
0.2037	0.9634	0.1170	0.9784	18

Framework versions

• Transformers 4.52.4
• TensorFlow 2.18.0
• Datasets 2.14.4
• Tokenizers 0.21.1