YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Banking SMS Parser - Working Model 🎉
✅ TRAINING SUCCESSFUL!
- Final Loss: 0.94 (60% improvement from 2.38)
- Shape Errors: FIXED - Perfect tensor dimensions
- Training Time: ~3 minutes
- Parameters: 74,429
📊 Model Specifications
- Architecture: LSTM-based Language Model
- Vocabulary: 61 banking-specific tokens
- Categories: 11 core banking categories
- Sequence Length: 12 tokens (fixed)
- Batch Size: 8
🎯 Supported Categories
- Automobile and Transportation (UBER, OLA, TAXI, PETROL)
- Food and Drinks (SWIGGY, ZOMATO, DOMINOS, KFC)
- Shopping (AMAZON, FLIPKART, MYNTRA)
- Entertainment (NETFLIX, SPOTIFY)
- Health and Wellness (HOSPITAL, PHARMACY, APOLLO)
- Bills (ELECTRICITY, WATER, BROADBAND)
- Mobile (RECHARGE, AIRTEL, JIO)
- Cash Transactions (ATM, WITHDRAWAL)
- Salary (SALARY, BONUS, PAYROLL)
- Reversal and Refunds (CASHBACK, REFUND)
- Self Transfer (UPI, NEFT, TRANSFER)
🚀 Usage
📈 Training Progress
- Epoch 1: Loss 2.38
- Epoch 2: Loss 1.25
- Epoch 3: Loss 0.94 ✅
🔧 Technical Details
- No shape errors during training
- Perfect tensor dimension matching
- LSTM architecture for sequence modeling
- Custom banking vocabulary
- Rule-based parsing fallback
🎉 SUCCESS METRICS
- ✅ Training completed without errors
- ✅ Model saved successfully
- ✅ Loss convergence achieved
- ✅ Ready for production testing
Built with ❤️ for banking SMS parsing!
- Downloads last month
- -
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support