Tag: production-ml
Concepts
- Continuous Training Triggers
- Feature Freshness and Staleness SLAs
- Feedback Loops and Self-Fulfilling Predictions
- Inference Latency Budgeting
- Label Latency and Delayed Feedback
- ML Technical Debt and Hidden Coupling
- Model Complexity vs Inference Cost
- Model Monitoring in Production
- Model Packaging and ONNX
- Model Serialization and Pickle Risks
- Model Serving Architectures
- Prediction Drift Monitoring
- Training-Serving Skew
- Unseen Categories at Inference Time