eraydikyologlu/bert_ayt_biyoloji

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: 0.1157
• Train Accuracy: 0.9868
• Validation Loss: 0.0668
• Validation Accuracy: 0.9915
• Epoch: 19

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': 4014, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 446, '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
3.7422	0.0744	3.2078	0.3239	0
2.3051	0.5032	1.6116	0.5895	1
1.5086	0.6218	1.2025	0.6903	2
1.2218	0.6717	0.9914	0.7259	3
1.0446	0.7138	0.8631	0.7543	4
0.9093	0.7422	0.7421	0.7926	5
0.7923	0.7759	0.6106	0.8310	6
0.6796	0.8119	0.4951	0.8523	7
0.5777	0.8433	0.4144	0.8949	8
0.4702	0.8819	0.3143	0.9134	9
0.3878	0.9091	0.2503	0.9446	10
0.3213	0.9313	0.2066	0.9517	11
0.2680	0.9462	0.1635	0.9602	12
0.2231	0.9584	0.1286	0.9702	13
0.1887	0.9699	0.1108	0.9744	14
0.1619	0.9759	0.0936	0.9815	15
0.1419	0.9800	0.0798	0.9886	16
0.1296	0.9840	0.0733	0.9886	17
0.1185	0.9860	0.0689	0.9886	18
0.1157	0.9868	0.0668	0.9915	19

Framework versions

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