AAP-SQL column retriever
AAP-SQL 欄位檢索器是完整 AAP-SQL 設定中的雙編碼模型。它從完整資料庫結構與欄位描述中召回最多 50 個候選欄位，供後續重排序器選出核心欄位。
AAP-SQL column retriever is the bi-encoder used in the full AAP-SQL workflow. It performs the first stage of schema retrieval and returns up to 50 candidate columns for candidate reranking.
Model details
- Base model: sentence-transformers/all-mpnet-base-v2
- Training objective: MultipleNegativesRankingLoss
- Training seed: 42
- Training data: column-retrieval examples derived from the BIRD training split and schema descriptions
- Expected library: sentence-transformers>=5.1.2
AAP-SQL publication branch
The complete AAP-SQL workflow, research method terminology, BIRD directory layout, and reproduction instructions are maintained in the GitHub publication branch.
Use with AAP-SQL
Download this repository into the path expected by the final runner:
hf download TommyPanLab/AAP-SQL-Column-Retriever --local-dir models/column_retriever_v3_paper_earlystop
Direct loading:
from sentence_transformers import SentenceTransformer
model = SentenceTransformer("TommyPanLab/AAP-SQL-Column-Retriever")
embeddings = model.encode(["table.column: column description"])
Data and license notice
The training examples were derived from the BIRD benchmark. Review the BIRD project terms before using the model. No additional license has been declared for these fine-tuned weights; the upstream model and dataset terms still apply.