Modelpublic

cslin612/masked-lm-tpu

sourceHugging Facemitupdated 2y agoView on Hugging Face

0likes7downloads

Model Card

cslin612/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: 9.9462
—Train Accuracy: 0.0015
—Validation Loss: 9.8661
—Validation Accuracy: 0.0096
—Epoch: 8

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', 'learningrate': {'module': 'transformers.optimizationtf', 'classname': 'WarmUp', 'config': {'initiallearningrate': 0.0001, 'decayschedulefn': {'module': 'keras.optimizers.schedules', 'classname': 'PolynomialDecay', 'config': {'initiallearningrate': 0.0001, 'decaysteps': 22325, 'endlearningrate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}, 'registeredname': None}, 'warmupsteps': 1175, 'power': 1.0, 'name': None}, 'registeredname': 'WarmUp'}, 'decay': 0.0, 'beta1': 0.9, 'beta2': 0.999, 'epsilon': 1e-08, 'amsgrad': False, 'weightdecayrate': 0.001}
—training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
10.2730	0.0	10.2892	0.0	0
10.2704	0.0	10.2654	0.0	1
10.2570	0.0	10.2421	0.0	2
10.2259	0.0000	10.2044	0.0	3
10.1972	0.0000	10.1556	0.0	4
10.1408	0.0	10.0927	0.0000	5
10.0955	0.0000	10.0258	0.0001	6
10.0190	0.0001	9.9450	0.0008	7
9.9462	0.0015	9.8661	0.0096	8

Framework versions

—Transformers 4.31.0
—TensorFlow 2.15.0
—Tokenizers 0.13.3