TommyPanLab/AAP-SQL-Candidate-Reranker

1---2language:3- en4library_name: sentence-transformers5pipeline_tag: text-ranking6base_model: cross-encoder/ms-marco-MiniLM-L6-v27tags:8- sentence-transformers9- text-ranking10- text2sql11- schema-linking12- aap-sql13---14 15# AAP-SQL candidate reranker16 17AAP-SQL 候選重排序器是完整 AAP-SQL 設定中的 cross-encoder。它對欄位檢索器召回的候選欄位重新評分，保留前 10 個核心欄位供後續提示增強使用。18 19AAP-SQL candidate reranker is the cross-encoder used after the first stage of schema retrieval. It scores the retrieved candidate columns and retains the top 10 core columns for prompt augmentation.20 21## Model details22 23- Base model: cross-encoder/ms-marco-MiniLM-L6-v224- Training objective: BinaryCrossEntropyLoss25- Training seed: 4226- Training data: schema-ranking examples derived from the BIRD training split and schema descriptions27- Expected library: sentence-transformers>=5.1.228 29## AAP-SQL publication branch30 31The complete AAP-SQL workflow, research method terminology, BIRD directory layout, and reproduction instructions are maintained in the [GitHub publication branch](https://github.com/Tommyweige/AAP-SQL/tree/codex/final-aap-sql-experiment/AAP-SQL-Original).32 33## Use with AAP-SQL34 35Download this repository into the path expected by the final runner:36 37~~~powershell38hf download TommyPanLab/AAP-SQL-Candidate-Reranker --local-dir models/cross_encoder_schema_paper_repro39~~~40 41Direct loading:42 43~~~python44from sentence_transformers import CrossEncoder45 46model = CrossEncoder("TommyPanLab/AAP-SQL-Candidate-Reranker")47scores = model.predict([ ("user question", "table.column: column description") ])48~~~49 50## Data and license notice51 52The training examples were derived from the BIRD benchmark. Review the [BIRD project terms](https://bird-bench.github.io/) before using the model. No additional license has been declared for these fine-tuned weights; the upstream model and dataset terms still apply.