Hate_speech_multiclass_BERT

Model Card Template for Hate Speech Classification Models

Model Name

Hate speech multi class classification - BERT

---

Model Overview

This model is fine-tuned on a BERT to classify hate speech in a multi-class problem. The three classes include:

• Racism
• Sexism
• Ableism

The model is designed for use in moderation systems, social media analysis, and academic research.

---

Intended Use

• Applications: Detection and classification of hate speech in text.
• Target Users: Developers, researchers, and organizations working on content moderation or hate speech analysis.
• Domains: Social media platforms, forums, and text-based content sources.

---

Metrics and Performance

• Evaluation Metrics:
  • Accuracy: 88.5696%
  • Precision: 88.8252
  • Recall: 88.5696
  • F1-Score: 88.4993%

---

Training Details

• Base Model: BERT-base-uncased
• Fine-tuning Framework: PyTorch
• Epochs: 20
• Batch Size: 8
• Learning Rate: 5 x 10-6
• Hardware Used: Intel(R) Core(TM) i7-12700 CPU at 2.10GHz, 32 GB RAM, and Nvidia GeForce GTX 16660 SUPER.

Licensing and Citation

• License: [Insert license type, e.g., Apache 2.0]
• Citation: Please cite the following if you use this model:

  @article{YourCitationHere,
    title={Title of your paper or repository},
    author={Your Name},
    year={Year},
    journal={Venue},
    url={Link to resource}
  }