|
|
--- |
|
|
license: apache-2.0 |
|
|
datasets: |
|
|
- gplsi/SocialTOX |
|
|
language: |
|
|
- es |
|
|
metrics: |
|
|
- accuracy |
|
|
- f1 |
|
|
- precision |
|
|
- recall |
|
|
base_model: |
|
|
- BSC-LT/roberta-base-bne |
|
|
pipeline_tag: text-classification |
|
|
--- |
|
|
|
|
|
# 🧠 Toxicity_model_RoBERTa-base-bne– Spanish Toxicity Classifier Binary (Fine-tuned) |
|
|
|
|
|
## 📌 Model Description |
|
|
|
|
|
This model is a fine-tuned version** of `RoBERTa-base-bne`, specifically trained to classify the toxicity level of **Spanish-language user comments on news articles**. It distinguishes between two categories: |
|
|
|
|
|
- **Non-toxic** |
|
|
- **Toxic** |
|
|
|
|
|
--- |
|
|
|
|
|
## 📂 Training Data |
|
|
|
|
|
The model was fine-tuned on the **[SocialTOX dataset](https://huggingface.co/datasets/gplsi/SocialTOX)**, a collection of Spanish-language comments annotated for varying levels of toxicity. These comments come from news platforms and represent real-world scenarios of online discourse. In this case, a Binary classifier was developed, where the classes \textit{Slightly toxic} and \textit{Toxic} were merged into a single \textit{Toxic} category. |
|
|
|
|
|
--- |
|
|
|
|
|
## Training hyperparameters |
|
|
- epochs: 10 |
|
|
- learning_rate: 2.45e-6 |
|
|
- beta1: 0.9 |
|
|
- beta2: 0.95 |
|
|
- Adam_epsilon: 1.00e-8 |
|
|
- weight_decay: 0 |
|
|
- batch_size: 16 |
|
|
- max_seq_length: 512 |
