AIbase
EN
Home
FRIDA
F

FRIDA

Developed by ai-forever
FRIDA is a multi-task evaluation model that supports various natural language processing tasks, including classification, clustering, and re-ranking.
Text Embedding
Transformers
 Supports Multiple LanguagesOpen Source License:MIT#Multi-label Classification#Russian Re-ranking#High-precision Sentiment Analysis
Downloads 9,491
Release Time : 12/26/2024

Model Overview

FRIDA is a versatile model suitable for multiple natural language processing tasks such as text classification, sentiment analysis, clustering, and re-ranking.

Model Features

Multi-task Support
Supports various natural language processing tasks, including classification, clustering, and re-ranking.
High Performance
Performs excellently on multiple datasets, especially in headline classification and inappropriate content classification tasks.
Multilingual Support
Primarily supports Russian but also performs well in multilingual tasks.

Model Capabilities

Text Classification
Sentiment Analysis
Clustering
Re-ranking
Multi-label Classification

Use Cases

Content Moderation
Inappropriate Content Detection
Used to detect and classify inappropriate content, such as hate speech or harmful information.
Accuracy 78.33%, F1-score 78.20%
Sentiment Analysis
Movie Review Sentiment Classification
Used to analyze the sentiment orientation of movie reviews.
Accuracy 70.47%, F1-score 65.84%
Information Retrieval
MIRACL Re-ranking
Used for document re-ranking tasks in information retrieval.
NDCG@10 66.04%, MAP@10 60.21%
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase