README.md

metadata

title: Multimodal A/B Test Predictor
emoji: 🚀
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: 4.44.0
app_file: app.py
pinned: false

🚀 Multimodal A/B Test Predictor

Overview

Advanced A/B testing outcome predictor using multimodal AI analysis combining:

• 🖼️ Image Analysis: Visual features from control & variant images
• 📝 OCR Text Extraction: Automatically extracts and analyzes text from images
• 📊 Categorical Features: Business context provided via API (industry, page type, etc.)
• 🎯 Confidence Scores: Based on training data statistics and historical accuracy

🎯 Direct Input Architecture

Image + Categorical Data

• Accepts control and variant images directly via API
• Requires categorical inputs: Business Model, Customer Type, Conversion Type, Industry, Page Type
• Fast, efficient predictions without external API dependencies
• All processing happens locally on GPU

🎯 Features

Direct Prediction with Categorical Data

• Upload control & variant images
• Provide categorical business context data
• Fast prediction with comprehensive confidence analysis

Enhanced Results

• Winner Prediction: Variant vs Control with probability
• Model Confidence: Accuracy percentage from training data
• Training Data Count: Number of samples model trained on for this category
• Historical Win/Loss: Real A/B test outcome statistics for this category
• Confidence Source: Industry + Page Type combination used for scoring

🔧 Setup

Model Files

• model/multimodal_gated_model_2.7_GGG.pth: Enhanced multimodal model (789MB)
• model/multimodal_cat_mappings_GGG.json: Category mappings

🚀 Technical Architecture

Model: SupervisedSiameseMultimodal (GGG Enhanced)

• Vision: ViT (Vision Transformer) for image features
• Text: DistilBERT for OCR text processing
• Fusion: Gated fusion with directional features
• Categories: Embedding layers for categorical features
• Architecture: BatchNorm + Fusion Block + Enhanced Prediction Head

Confidence Scoring

• Based on Industry + Page Type combinations
• Uses holdout statistics with average 160 samples per combination
• Much more reliable than full 5-feature combinations

📊 Performance

• Multimodal Analysis: Images + Text + Categories
• GPU Accelerated: Fast predictions (2-4 seconds average)
• High Accuracy: Enhanced GGG architecture with real training data
• No External Dependencies: All processing done locally

🎯 Use Cases

• A/B Test Prediction: Predict winners before running tests with provided context
• Batch Processing: Process multiple tests efficiently from CSV
• Confidence Assessment: Understand prediction reliability based on historical data
• API Integration: Easy integration with external systems

📡 API Usage

Quick Start

from gradio_client import Client

client = Client("SpiralyzeLLC/ABTestPredictor")
result = client.predict(
    "control.jpg", "variant.jpg",
    "SaaS", "B2B", "High-Intent Lead Gen",
    "B2B Software & Tech", "Awareness & Discovery",
    api_name="/predict_with_categorical_data"
)

📚 Documentation

• QUICK_START_GUIDE.md - Get started in 3 steps
• API_DOCUMENTATION.md - Complete API reference with examples in Python, JavaScript, and cURL
• API_USAGE_UPDATED.md - Technical details and migration notes

🌐 Web Interface

Use the interactive interface: https://huggingface.co/spaces/SpiralyzeLLC/ABTestPredictor

---

Built with ❤️ using Gradio, PyTorch, Transformers, and Hugging Face.