Bert Base Uncased Goemotions Ekman Finetuned

🚀 Go-Emotion Text Classification Model

A text classification model based on BERT, designed for classifying emotions in text using the Go-Emotions dataset.

🚀 Quick Start

This model is a text classification model that can classify emotions in text. It is based on the uncased BERT pretrained model and is trained on the Go-Emotions dataset.

✨ Features

• Based on the uncased BERT pretrained model with a linear output layer.
• Added several commonly-used emoji and tokens to the special token list of the tokenizer.
• Did label smoothing while training.
• Used weighted loss and focal loss to help the cases which trained badly.

📦 Installation

No specific installation steps are provided in the original README.

💻 Usage Examples

No code examples are provided in the original README.

📚 Documentation

Model Information

Widget

You can test the model with the following text:

"Thanks for giving advice to the people who need it! 👌🙏"

🔧 Technical Details

• The model is built on the uncased BERT pretrained model, with a linear output layer added for classification.
• The tokenizer's special token list is extended with several commonly-used emoji and tokens.
• During training, label smoothing is applied to prevent overfitting.
• Weighted loss and focal loss are used to address the issue of poorly-trained cases.

📄 License

This project is licensed under the MIT License.