Claude ff58ff0189 rerank: use an example where the cross-encoder fixes the order in 02, correct the download size ...

Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01Cfo2K4u4rF18QsPXPFnytf

2026-10-24 18:03:45 +00:00

..

01-pair-encoding.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

02-one-question-by-hand.txt

rerank: use an example where the cross-encoder fixes the order in 02, correct the download size

2026-10-24 18:03:45 +00:00

03-quality-by-pool-size.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

04-quality-by-question-kind.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

05-helped-and-hurt.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

06-latency-by-candidates-and-batch.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

07-latency-against-quality.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

08-post-processor-in-the-advisor.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

09-post-processor-edge-cases.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

10-trap-missing-input.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

11-trap-pair-order.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

12-trap-score-scale.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

13-trap-truncation.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00

14-trap-small-pool.txt

Add rerank module: bi-encoder retrieval plus an ONNX Runtime cross-encoder reranker behind a Spring AI DocumentPostProcessor; hit@k and MRR by pool size and question kind, latency by candidates, batch size, threads and int8, and five reproduced traps. Local MiniLM models, no database or API key.

2026-10-24 18:00:05 +00:00