Do I need training data for text classification?

No, you can use the zero-shot classification feature for instant labeling. However, for higher accuracy in specific domains, you can train a custom model with as few as 8-64 examples.

How do I manage multiple trained models?

The skill includes endpoints to list all trained classifiers, delete specific models, and predict using any model by name, allowing for a full lifecycle management workflow.

What kind of tasks is this skill best for?

It is ideal for sentiment analysis, topic labeling, support ticket categorization, and any task that involves organizing text into predefined categories.

What is SetFit and why is it used here?

SetFit is a framework for few-shot fine-tuning of Sentence Transformers. It provides high accuracy on classification tasks even with very small datasets, making it ideal for quick custom deployments.

Text Classification & Labeling

Name: Text Classification & Labeling
Author: llama-farm

byllama-farm

0•

データサイエンスとML

Classifies and categorizes text using zero-shot learning or custom-trained SetFit models for high-accuracy document labeling.

This skill provides a comprehensive toolkit for text classification within the LlamaFarm ecosystem. It offers two distinct approaches: instant zero-shot classification for broad categories and quick prototyping, and high-precision custom model training using SetFit for domain-specific tasks with as few as eight examples per class. Whether you need to automate sentiment analysis, route support tickets, or organize large document sets, this skill simplifies the process of training, deploying, and managing text classifiers directly through Claude Code.

主な機能

01Comprehensive model lifecycle management including training and prediction

02Few-shot learning with SetFit for high-accuracy domain-specific models

03Support for sentence-transformer base models for efficient processing

04Zero-shot classification for instant labeling without training data

05Automated sentiment analysis and topic categorization

060 GitHub stars

ユースケース

01Tagging and organizing large repositories of documentation

02Automating support ticket routing based on issue type

03Extracting sentiment and themes from user feedback or reviews

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add llama-farm/claude-code-marketplace text-classification

For use in Claude.ai and ChatGPT

Download Skill