HybridServe: Efficient Serving of Large AI Models with Confidence-Based Cascade Routing

Re-assign community