pcmill/masked-lm-tpu

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

• Train Loss: 9.9285
• Train Accuracy: 0.0044
• Validation Loss: 9.8057
• Validation Accuracy: 0.0197
• Epoch: 9

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': 'AdamWeightDecay', 'learning_rate': {'module': 'transformers.optimization_tf', 'class_name': 'WarmUp', 'config': {'initial_learning_rate': 0.0001, 'decay_schedule_fn': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 0.0001, 'decay_steps': 22325, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registered_name': None}, 'warmup_steps': 1175, 'power': 1.0, 'name': None}, 'registered_name': 'WarmUp'}, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weight_decay_rate': 0.001}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
10.3759	0.0000	10.3822	0.0	0
10.3601	0.0000	10.3615	0.0	1
10.3529	0.0	10.3315	0.0000	2
10.3210	0.0	10.2957	0.0000	3
10.2824	0.0	10.2382	0.0	4
10.2333	0.0	10.1677	0.0	5
10.1625	0.0	10.0990	0.0	6
10.1024	0.0000	10.0062	0.0001	7
10.0126	0.0004	9.9072	0.0058	8
9.9285	0.0044	9.8057	0.0197	9

Framework versions

• Transformers 4.38.1
• TensorFlow 2.15.0
• Tokenizers 0.15.2