roberta-base model is fine tuned on news article abstracts and headlines.

Respective labels and thier categories

• 0: 'ARTS & CULTURE',
• 1: 'BUSINESS & FINANCES',
• 2: 'COMEDY',
• 3: 'CRIME',
• 4: 'DIVORCE',
• 5: 'EDUCATION',
• 6: 'ENTERTAINMENT',
• 7: 'ENVIRONMENT',
• 8: 'FOOD & DRINK',
• 9: 'GROUPS VOICES',
• 10: 'HOME & LIVING',
• 11: 'IMPACT',
• 12: 'MEDIA',
• 13: 'MISCELLANEOUS',
• 14: 'PARENTING',
• 15: 'POLITICS',
• 16: 'RELIGION',
• 17: 'SCIENCE & TECH',
• 18: 'SPORTS',
• 19: 'STYLE & BEAUTY',
• 20: 'TRAVEL',
• 21: 'U.S. NEWS',
• 22: 'WEDDINGS',
• 23: 'WEIRD NEWS',
• 24: 'WELLNESS',
• 25: 'WOMEN',
• 26: 'WORLD NEWS'