Spaces:
Runtime error
Runtime error
Commit History
-
Mapping of grouped metadata
5b49b49 -
Deploy: Optimize for Hugging Face Spaces - remove AI APIs and unnecessary files
ec4c5af -
build error - fix 1
e93a798 -
Take input from API instead of AI
e749f25
-
% change
e43b46b -
Convert all JSON keys to camelCase for consistency
84c9af7 -
Update JSON format: camelCase keys, split accuracy into discrete fields
87dad1f -
Update JSON structure: Remove winner field, rename fields, add new prediction metrics
fba0a01
-
Clean API: Remove emojis, add colleague integration guide
05bf721 -
...
61e48f5
-
Update: Re-enable GPU decorators with optimized durations (50s/60s)
4a19fdb -
Temporary fix: Disable GPU to avoid quota limits (runs on CPU)
55530e3 -
Fix: Update to Gemini 2.5 Flash, increase token limit, add .env support
8f319c2 -
Fix: Reduce GPU duration to 60s for Hugging Face free tier and add local frontend
a4f25de
-
Add comprehensive API integration guide
- Document all methods for integrating the model (Gradio Client, HTTP API, iframe, React) - Include code examples in Python, JavaScript, cURL - Add response format documentation - Include best practices and rate limit guidelines - Provide full working examples for web app integrationbffa639 -
Fix confidence scores loading by using JSON instead of Node.js
BREAKING CHANGE: Switch from parsing JS file to using pure JSON - Created confidence_scores.json from confidence_scores.js - Removed Node.js dependency for loading confidence scores - Added extensive debugging for file path resolution - Added fallback path handling - This eliminates the subprocess reliability issues in HF Spaces The confidence scores should now load reliably without depending on Node.js subprocess execution.cfff171 -
Fix confidence scores loading with absolute path
- Use absolute path to confidence_scores.js instead of relative path - Add detailed debugging for file path resolution - Add better error reporting for subprocess failures - This should fix the issue with confidence scores not loading in HF Spacese87ec54 -
Fix missing model confidence metrics with enhanced debugging
- Added comprehensive error handling to predict_single function - Added detailed debugging logs for confidence data loading - Fixed indentation issues that could cause incomplete results - Enhanced get_confidence_data function with detailed logging - Now properly displays Model Confidence, Training Data, Historical Accuracy, and Win/Loss Ratiofef82a4 -
...
774e5c0 -
🎯 Optimize model loading - use local model file
✅ Model File Confirmed: - GGG model file (789MB) successfully uploaded to Space - Located at model/multimodal_gated_model_2.7_GGG.pth - Quick upload likely due to HF deduplication or Git LFS efficiency 🔧 Optimized Loading Logic: - Prioritize local model file if available - Only download from Model Hub if local file missing - Faster startup when model is already in Space - Fallback to Model Hub download if needed 🚀 Performance Benefits: - No unnecessary downloads when model file present - Faster app startup with local model access - Reduced bandwidth usage - Better user experience with quicker load times The app will now use the uploaded 789MB GGG model directly!81aa4c1 -
✅ Add Hugging Face Spaces configuration
🔧 Configuration Fix: - Added required YAML front matter to README.md - Set title: Multimodal A/B Test Predictor - Set emoji: 🚀 and blue to purple gradient - Configured Gradio SDK with version 4.44.0 - Set app_file: app.py 📁 Added Missing Files: - model/multimodal_cat_mappings_GGG.json for category mappings - Complete file structure for Space deployment 🚀 Ready for Launch: - All configuration requirements met - Space should build and deploy successfully - Enhanced dual-AI A/B test predictor ready159cf40 -
Upload multimodal_gated_model_2.7_GGG.pth
Verified signaturebf6c91a -
🚀 Complete A/B Test Predictor with Enhanced Dual-AI Analysis
✨ Features: - 🤖 Dual-AI powered analysis (Perplexity + Gemini Pro) - 🎯 359 A/B test pattern detection with rich context - 📊 Confidence scoring based on training statistics - 🔍 Enhanced GGG model architecture with real trained weights - 📝 OCR text extraction and multimodal fusion 🔧 Technical Stack: - SupervisedSiameseMultimodal with GGG enhancements - Perplexity Sonar Reasoning Pro for business categorization - Gemini Pro Vision for visual pattern detection - Industry + Page Type confidence scoring (avg 160 samples) - Hugging Face Model Hub integration for large model files 📊 Model Architecture: - Vision: ViT (google/vit-base-patch16-224-in21k) - Text: DistilBERT (distilbert-base-uncased) - Fusion: Gated fusion with directional features - Categories: 5 categorical features with embeddings - Enhanced: BatchNorm + Fusion Block + Role Embedding 🎯 Capabilities: - Smart auto-prediction with zero manual input - Manual category selection for precise control - Batch prediction from C…37380c1 -
initial commit
Verified signaturefa5735b