Bert Hateful Memes Expanded

🚀 bert-hateful-memes-expanded

This model is a fine - tuned version of BERT, designed to handle texts from multiple datasets related to hateful memes, providing valuable results on the evaluation set.

🚀 Quick Start

This model is a fine-tuned version of bert-base-uncased on texts from the following datasets:

• Hateful Memes, train, dev_seen and dev_unseen
• HarMeme, train, val and test
• MultiOFF, Training, Validation and Testing

It achieves the following results on the evaluation set:

• Loss: 3.7600

✨ Features

This model is fine - tuned on multiple datasets related to hateful memes, which can effectively deal with text data in these scenarios.

📚 Documentation

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

🔧 Technical Details

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3.0

Training results

No specific results provided in the original text.

Framework versions

• Transformers 4.11.0
• Pytorch 1.8.1+cu102
• Datasets 1.8.0
• Tokenizers 0.10.2

📄 License

This model is licensed under the Apache-2.0 license.