hate-social-network-adversarial

<!-- This model card has been generated automatically according to the information Keras had access to. You should probably proofread and complete it, then remove this comment. -->

hate-social-network-adversarial

Este modelo es una version of esmarquez17/fine-tunning-roberta-bne-hate-offensive on an unknown dataset. evaluado con un conjunto de datos semEvaml-2019 con generación de datos adversariales:

Model description

• Modelo Base con fine-tunning de Roberta-BNE en un corpus de guiones de teatro
• Modelo entrenado con un conjunto de datos adversarios propuestos

Training and evaluation data

• Entrenado en corpus base SemEval-spanish
• Validado SemEval-spanish
• Testeado en corpus: HATERNET y HATECHECK

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• optimizer: {'name': 'Adam', 'learning_rate': {'module': 'keras.optimizers.schedules', 'class_name': 'PolynomialDecay', 'config': {'initial_learning_rate': 1e-05, 'decay_steps': 9385, 'end_learning_rate': 0.0, 'power': 1.0, 'cycle': False, 'name': None}}
• training_precision: float32

Training results

Training { Exactitud 0.9702 Precisión 0.9622 F1-score 0.9615 Recall 0.9609 } Validacion { Exactitud 0.8520 Precisión 0.8558 F1-score 0.8279 Recall 0.8018 }

Framework versions

• Transformers 4.35.2
• TensorFlow 2.14.0
• Datasets 2.15.0
• Tokenizers 0.15.0