sugardata 0.0.3

Generates synthetic datasets tailored for transformer-based models

sugardata

sugardata is a Python package that helps you generate synthetic datasets for NLP tasks, enabling experimentation, prototyping, and training when labeled data is limited or unavailable.

Installation

You can install sugardata via pip:

pip install sugardata

Quick Start

sugardata supports multiple NLP tasks out of the box.

Task 1: Sentiment Analysis

import sugardata as su

results = su.generate_sentiments(concept="online shopping", n_samples=100)

Task 2: Aspect-Based Sentiment Analysis (ABSA)

import sugardata as su

results = su.generate_aspect_sentiments(concept="smartphones", aspects=["battery life", "camera", "price"], n_samples=100)

To learn more about configuration options, advanced parameters, and integration tips, please visit tutorials.