feat(bioflow): P0 内核移植与 8 环境镜像——副本 100 处适配零硬编码/8 镜像验证清单全过/金标 24 ID diff 空/TL 复查 4 项修正(gfpy 储备裁决+镜像备份排除+冒烟零硬编码+金标固化)

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-09-06 02:59:26 +08:00
parent 486670b75d
commit 99e774a2a0
589 changed files with 5300776 additions and 116 deletions
@@ -0,0 +1,4 @@
+from sentence_transformers import SentenceTransformer
+
+model = SentenceTransformer("BAAI/bge-m3")
+# 运行完模型就下载到本机缓存目录了（一般在 ~/.cache/huggingface）
@@ -0,0 +1,96 @@
+# list_builtin_models.py
+import os
+from pathlib import Path
+
+def list_sentence_transformers_models():
+    """查看sentence-transformers的内置模型"""
+    try:
+        import sentence_transformers
+        print("sentence-transformers版本:", sentence_transformers.__version__)
+        
+        # 查看模块目录
+        st_path = Path(sentence_transformers.__file__).parent
+        print(f"sentence-transformers安装路径: {st_path}")
+        
+        # 查看是否有预定义的模型列表
+        models_dir = st_path / "models"
+        if models_dir.exists():
+            print(f"\n模型目录内容:")
+            for item in models_dir.iterdir():
+                if item.is_dir():
+                    print(f"  📁 {item.name}")
+    except Exception as e:
+        print(f"错误: {e}")
+
+def list_available_models():
+    """尝试加载常见的预训练模型"""
+    print("\n尝试加载常见模型...")
+    
+    # 常见的sentence-transformers模型
+    common_models = [
+        # 小型模型
+        "all-MiniLM-L6-v2",
+        "all-MiniLM-L12-v2",
+        "paraphrase-MiniLM-L3-v2",
+        "paraphrase-MiniLM-L6-v2",
+        "all-distilroberta-v1",
+        "multi-qa-MiniLM-L6-cos-v1",
+        
+        # 中型模型
+        "all-mpnet-base-v2",
+        "multi-qa-mpnet-base-dot-v1",
+        "all-roberta-large-v1",
+        
+        # 多语言模型
+        "paraphrase-multilingual-MiniLM-L12-v2",
+        "paraphrase-multilingual-mpnet-base-v2",
+        
+        # BGE系列
+        "BAAI/bge-small-en-v1.5",
+        "BAAI/bge-base-en-v1.5",
+        "BAAI/bge-large-en-v1.5",
+        "BAAI/bge-m3",
+    ]
+    
+    available = []
+    
+    for model_name in common_models:
+        try:
+            from sentence_transformers import SentenceTransformer
+            # 尝试本地加载
+            model = SentenceTransformer(model_name, local_files_only=True)
+            # 测试编码
+            test_emb = model.encode(["test"], normalize_embeddings=True)
+            dim = test_emb.shape[1]
+            available.append((model_name, dim))
+            print(f"✅ {model_name} (维度: {dim})")
+        except Exception as e:
+            # 只显示成功的模型
+            pass
+    
+    return available
+
+if __name__ == "__main__":
+    print("=" * 60)
+    print("检查可用的嵌入模型")
+    print("=" * 60)
+    
+    list_sentence_transformers_models()
+    
+    print("\n" + "=" * 60)
+    print("测试模型可用性")
+    print("=" * 60)
+    
+    available_models = list_available_models()
+    
+    if available_models:
+        print(f"\n✅ 找到 {len(available_models)} 个可用模型:")
+        for model_name, dim in available_models:
+            print(f"  - {model_name} (维度: {dim})")
+        
+        # 推荐选择
+        print(f"\n🌟 推荐使用: {available_models[0][0]}")
+        print(f"   维度: {available_models[0][1]}")
+    else:
+        print("\n❌ 没有找到可用的本地模型")
+        print("需要下载模型到本地或使用网络")