umutarpayy/difficulty_classification_bert

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: 0.0379
• Train Accuracy: 0.9861
• Validation Loss: 0.0148
• Validation Accuracy: 0.9939
• Epoch: 12

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': 3e-05, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 3e-05, 'decay_steps': 167884, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 18653, '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
1.1802	0.5355	1.0560	0.5836	0
1.0386	0.5899	0.8960	0.6624	1
0.8708	0.6629	0.6267	0.7682	2
0.6683	0.7491	0.3956	0.8641	3
0.4929	0.8192	0.2585	0.9208	4
0.3597	0.8705	0.1497	0.9550	5
0.2605	0.9074	0.0988	0.9707	6
0.1864	0.9347	0.0676	0.9793	7
0.1359	0.9529	0.0393	0.9880	8
0.0986	0.9658	0.0281	0.9912	9
0.0717	0.9754	0.0208	0.9928	10
0.0519	0.9815	0.0168	0.9935	11
0.0379	0.9861	0.0148	0.9939	12

Framework versions

• Transformers 4.51.3
• TensorFlow 2.18.0
• Datasets 3.5.1
• Tokenizers 0.21.1