library_name: transformers tags:

Model Card for roberta-base-allsides

This model is a fine-tuned version of FacebookAI/roberta-base on the liyucheng/allsides dataset. It is trained to classify U.S. news articles and story summaries into political bias categories: left, center, or right.

---

Model Details

Model Description

• Developed by: [Your Name or Organization]
• Funded by [optional]: [Optional]
• Shared by: [Your HF username]
• Model type: Transformer-based text classifier
• Language(s): English
• License: Apache 2.0
• Finetuned from model: FacebookAI/roberta-base

Model Sources

• Repository: [Insert link to your GitHub or HF repo if public]
• Dataset: liyucheng/allsides

---

Uses

Direct Use

Downstream Use

Out-of-Scope Use

---

Bias, Risks, and Limitations

Recommendations

Use this model as an analytical aid, not a ground truth source. Always review outputs critically, especially in high-stakes applications (e.g., journalism, fact-checking).

---

How to Get Started with the Model

from transformers import pipeline

classifier = pipeline("text-classification", model="your-username/roberta-base-allsides", tokenizer="your-username/roberta-base-allsides")
classifier("Biden proposes new economic plan amid GOP criticism")