Datasets:
Download construction/retrieve_candidates.py from leequeue/MIPR: direct link, hf CLI and curl.
- Browser
- Download file 4.69 kB
-
https://huggingface.co/datasets/leequeue/MIPR/resolve/main/construction/retrieve_candidates.py
- Command line
-
hf download hf://datasets/leequeue/MIPR/construction/retrieve_candidates.py
-
curl -L -o retrieve_candidates.py https://huggingface.co/datasets/leequeue/MIPR/resolve/main/construction/retrieve_candidates.py
4.69 kB
| """商品词走三路检索,意图走文本融合和多模态两路;每源默认 top-5。 | |
| search(mode, queries, top_k) 由调用方提供,不在本文件实现检索后端。 | |
| queries 为 {query_id: text},返回 {query_id: [(item_id, score), ...]},按分数降序。 | |
| mode: bm25(jieba 分词);embedding(Qwen3-Embedding-0.6B 与 | |
| BGE-small-zh-v1.5 的 RRF 融合);qwen3_vl(Qwen3-VL-Embedding-2B)。 | |
| """ | |
| import argparse | |
| import json | |
| from pathlib import Path | |
| from utils import deduplicate, load_embeddings, read_jsonl, write_jsonl | |
| MODES = ("bm25", "embedding", "qwen3_vl") | |
| def retrieve_candidates(input_path, output_path, search, top_k=5): | |
| with open(input_path, encoding="utf-8") as f: | |
| entries = [json.loads(line) for line in f if line.strip()] | |
| keywords = dict.fromkeys(kw for row in entries for kw in row["product_keywords"]) | |
| intents = dict.fromkeys(row["scene"]["intent"] for row in entries if row["scene"]["intent"]) | |
| kw_to_qid = {kw: f"kw_{i}" for i, kw in enumerate(keywords)} | |
| intent_to_qid = {intent: f"int_{i}" for i, intent in enumerate(intents)} | |
| kw_queries = {qid: kw for kw, qid in kw_to_qid.items()} | |
| intent_queries = {qid: intent for intent, qid in intent_to_qid.items()} | |
| mode_results = {} | |
| for mode in MODES: | |
| queries = kw_queries if mode == "bm25" else {**kw_queries, **intent_queries} | |
| mode_results[mode] = search(mode, queries, top_k=top_k) | |
| output = [assemble_sources(row, mode_results, kw_to_qid, intent_to_qid, top_k) | |
| for row in entries] | |
| Path(output_path).parent.mkdir(parents=True, exist_ok=True) | |
| with open(output_path, "w", encoding="utf-8") as f: | |
| for row in output: | |
| f.write(json.dumps(row, ensure_ascii=False) + "\n") | |
| return output | |
| def assemble_sources(row, mode_results, kw_to_qid, intent_to_qid, top_k=5): | |
| searches = [(kw, kw_to_qid[kw], "keyword") for kw in row["product_keywords"]] | |
| intent = row["scene"]["intent"] | |
| if intent: | |
| searches.append((intent, intent_to_qid[intent], "intent")) | |
| sources = [] | |
| for text, qid, kind in searches: | |
| for mode in MODES: | |
| if kind == "intent" and mode == "bm25": | |
| continue | |
| hits = mode_results[mode].get(qid, [])[:top_k] | |
| sources.append({ | |
| "source": f"{mode}:{text if kind == 'keyword' else 'intent'}", | |
| "mode": mode, "query_text": text, "query_type": kind, | |
| "items": [{"item_id": did, "score": score, "rank": rank} | |
| for rank, (did, score) in enumerate(hits, 1)], | |
| }) | |
| return {"idx": row["idx"], "query": row["query"], "image_url": row["image_url"], | |
| "scene": row["scene"], "product_keywords": row["product_keywords"], | |
| "source_results": sources} | |
| # 叶类目内去重,随后合并并重定向各 query 的候选。 | |
| def redirect_candidates(entries, product_mapping): | |
| """将检索源结果合并为 query-product 对,重定向商品并合并来源。""" | |
| pairs = {} | |
| for row in entries: | |
| for source in row["source_results"]: | |
| for hit in source["items"]: | |
| item_id = product_mapping[str(hit["item_id"])] | |
| key = (str(row["idx"]), item_id) | |
| pair = pairs.setdefault(key, {"idx": key[0], "item_id": item_id, "retrieved_by": []}) | |
| if source["source"] not in pair["retrieved_by"]: | |
| pair["retrieved_by"].append(source["source"]) | |
| return list(pairs.values()) | |
| def run_product_dedup(args): | |
| items = read_jsonl(args.input) | |
| emb = load_embeddings(args.emb_path, [row["item_id"] for row in items]) | |
| kept, mapping = deduplicate(items, emb, "product", args.threshold) | |
| pairs = redirect_candidates(read_jsonl(args.sources), mapping) | |
| out = Path(args.output_dir) | |
| write_jsonl(out / "deduplicated.jsonl", kept) | |
| write_jsonl(out / "pairs.jsonl", pairs) | |
| (out / "id_mapping.json").write_text(json.dumps(mapping, ensure_ascii=False, indent=2), encoding="utf-8") | |
| print(f"商品去重:{len(items)} → {len(kept)}") | |
| def main(): | |
| p = argparse.ArgumentParser(description="候选商品去重及 query-product 对重定向") | |
| p.add_argument("--input", required=True, help="召回商品并集 JSONL") | |
| p.add_argument("--emb-path", required=True, help="Qwen3-VL-Embedding-8B 图文 NPZ: emb, ids") | |
| p.add_argument("--sources", required=True, help="retrieve_candidates 输出的 source_results") | |
| p.add_argument("--output-dir", required=True) | |
| p.add_argument("--threshold", type=float, default=0.60) | |
| run_product_dedup(p.parse_args()) | |
| if __name__ == "__main__": | |
| main() | |