nikoslefkos/rebert_v3

nikoslefkos/rebert_trex_reformed_v3

This model is a fine-tuned version of distilbert-base-cased on trex for 250 labels. It achieves the following results on the evaluation set:

• Train Loss: 0.4216
• Train Accuracy: 0.8541
• Validation Loss: 0.8042
• Validation Accuracy: 0.7628
• Epoch: 4

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': 0.01, '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': 1e-05, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

Train Loss	Train Accuracy	Validation Loss	Validation Accuracy	Epoch
1.4168	0.6536	0.9242	0.7294	0
0.8272	0.7506	0.8106	0.7534	1
0.6786	0.7826	0.7871	0.7587	2
0.5718	0.8100	0.7981	0.7571	3
0.4216	0.8541	0.8042	0.7628	4

Framework versions

• Transformers 4.34.0
• TensorFlow 2.13.0
• Datasets 2.14.5
• Tokenizers 0.14.1