Skip to content

A Django library that leverages LLMs to enhance data quality through semantic validation and automatic data cleaning.

License

Notifications You must be signed in to change notification settings

Mazafard/Django-AI-Validator

Repository files navigation

Django AI Validator

PyPI version License: MIT codecov Documentation

Documentation: https://mazafard.github.io/django-ai-validator/

A Django library for AI-powered data validation and cleaning.

Features

  • Semantic Validation: Validate data based on meaning using LLMs.
  • Automated Cleaning: Automatically clean and normalize data.
  • Admin Integration: Bulk actions and status indicators in Django Admin.
  • Asynchronous Support: Offload LLM calls to Celery tasks.
  • Multiple Providers: Support for OpenAI, Anthropic, Gemini, and Ollama.

Installation

pip install django-ai-validator

Configuration

Add to INSTALLED_APPS:

INSTALLED_APPS = [
    ...
    'django_ai_validator',
]

Set your API key:

# Defaults to OpenAIClient
AI_CLEANER_LLM_CLIENT = 'django_ai_validator.llm.client.OpenAIClient'
OPENAI_API_KEY = "your-api-key"

Usage

Semantic Validation

from django.db import models
from django_ai_validator.validators import AISemanticValidator

class Product(models.Model):
    description = models.TextField(
        validators=[
            AISemanticValidator(
                prompt_template="Ensure the description is marketing-friendly and contains no offensive language."
            )
        ]
    )

Automatic Cleaning

from django_ai_validator.fields import AICleanedField

class UserProfile(models.Model):
    bio = AICleanedField(
        cleaning_prompt="Fix grammar and remove personal contact info."
    )

Admin Integration

from django.contrib import admin
from django_ai_validator.admin import AIAdminMixin
from .models import MyModel

@admin.register(MyModel)
class MyModelAdmin(AIAdminMixin, admin.ModelAdmin):
    list_display = ['content', 'is_dirty']
    actions = ['run_ai_cleanup_on_selected']

Manual Validation

To check if user input contains bad words:

from django_ai_validator.validators import AISemanticValidator

class Comment(models.Model):
    text = models.TextField(
        validators=[
            AISemanticValidator(
                prompt_template="Check if this text contains bad words. Return VALID if not."
            )
        ]
    )

Asynchronous Cleaning

To automatically remove personal information:

from django_ai_validator.fields import AICleanedField

class Feedback(models.Model):
    message = AICleanedField(
        cleaning_prompt="Remove PII.",
    )

Star History

Star History Chart

About

A Django library that leverages LLMs to enhance data quality through semantic validation and automatic data cleaning.

Resources

License

Contributing

Stars

Watchers

Forks

Packages

No packages published

Contributors 2

  •  
  •