Rauf Ibishov/Projects/Cross-Encoder Re-ranking Service/Cross-Encoder Re-ranking ServiceCross-Encoder LLM FastAPI PyTorchMicroservice wrapping a cross-encoder LLM and Lightweight Re-ranking Transformer (LRT) to re-score BM25/dense first-stage candidates.Benchmarked latency vs. precision trade-offs across pipeline stages to guide production deployment decisions.Links#GitHub