将chat模型和embedding模型分离，支持了ollama和openai两种格式

.env.example

@@ -30,9 +30,26 @@ CHROMA_SERVER_PORT=8002
 CHROMA_COLLECTION_NAME=rag_collection
 
 # ============================================
-# Ollama 配置
+# LLM 配置 (用于文本生成)
+# ============================================
+# LLM provider: ollama, openai 等
+LLM_PROVIDER=ollama
+LLM_BASE_URL=http://localhost:11434/v1
+LLM_MODEL=qwen3:8b
+LLM_API_KEY=  # 如使用openai等需要API Key的服务
+
+# ============================================
+# Embedding 配置 (用于向量检索)
+# ============================================
+# Embedding provider: ollama, openai 等
+EMBEDDING_PROVIDER=ollama
+EMBEDDING_BASE_URL=http://localhost:11434
+EMBEDDING_MODEL=qwen3-embedding:0.6b
+EMBEDDING_API_KEY=  # 如使用openai等需要API Key的服务
+
+# ============================================
+# 兼容旧版本配置 (已废弃，仍可用但推荐使用上面的配置)
 # ============================================
-# OLLAMA_BASE_URL: Ollama 服务地址
 OLLAMA_BASE_URL=http://localhost:11434
 OLLAMA_MODEL=qwen3:8b
 OLLAMA_EMBEDDING_MODEL=qwen3-embedding:0.6b

config.py

@@ -138,13 +138,24 @@ class Settings(BaseSettings):
     CHROMA_DB_PATH: str = "./chroma_db"  # Only used for PersistentClient mode
     CHROMA_COLLECTION_NAME: str = "rag_collection"
     
-    # Ollama Settings
-    # Configure OLLAMA_BASE_URL in .env file based on your deployment
-    # - Local: http://localhost:11434
-    # - Remote: http://192.168.1.100:11434
+    # LLM Settings
+    # LLM provider: "ollama", "vllm", "openai", "deepseek" etc.
+    LLM_PROVIDER: str = "ollama"
+    LLM_BASE_URL: str = "http://localhost:11434"
+    LLM_MODEL: str = "qwen3:8b"
+    LLM_API_KEY: Optional[str] = None
+    
+    # Embedding Settings  
+    # Embedding provider: "ollama", "vllm", "openai" etc.
+    EMBEDDING_PROVIDER: str = "ollama"
+    EMBEDDING_BASE_URL: str = "http://localhost:11434"
+    EMBEDDING_MODEL: str = "qwen3-embedding:0.6b"
+    EMBEDDING_API_KEY: Optional[str] = None
+    
+    # Legacy Ollama Settings (for backward compatibility)
     OLLAMA_BASE_URL: str = "http://localhost:11434"
-    OLLAMA_MODEL: str = "qwen3:1.7b"  # LLM model for text generation
-    OLLAMA_EMBEDDING_MODEL: str = "qwen3-embedding:0.6b"  # Embedding model for vectorization
+    OLLAMA_MODEL: str = "qwen3:8b"
+    OLLAMA_EMBEDDING_MODEL: str = "qwen3-embedding:0.6b"
     
     # RAG Settings
     EMBEDDING_DIMENSION: int = 768
@@ -184,6 +195,71 @@ class Settings(BaseSettings):
         extra="ignore"  # Ignore extra fields in .env file that are not defined in Settings
     )
     
+    def get_llm_config(self) -> Dict[str, Any]:
+        """
+        Get LLM configuration with backward compatibility for legacy OLLAMA_* settings.
+        
+        Returns:
+            Dictionary with provider, base_url, model, and api_key
+        """
+        if self.LLM_PROVIDER == "ollama":
+            return {
+                "provider": "ollama",
+                "base_url": self.LLM_BASE_URL or self.OLLAMA_BASE_URL,
+                "model": self.LLM_MODEL or self.OLLAMA_MODEL,
+                "api_key": self.LLM_API_KEY
+            }
+        elif self.LLM_PROVIDER == "openai":
+            return {
+                "provider": "openai",
+                "base_url": self.LLM_BASE_URL or "https://api.openai.com/v1",
+                "model": self.LLM_MODEL,
+                "api_key": self.LLM_API_KEY
+            }
+        else:
+            return {
+                "provider": self.LLM_PROVIDER,
+                "base_url": self.LLM_BASE_URL,
+                "model": self.LLM_MODEL,
+                "api_key": self.LLM_API_KEY
+            }
+    
+    def get_embedding_config(self) -> Dict[str, Any]:
+        """
+        Get Embedding configuration with backward compatibility for legacy OLLAMA_* settings.
+        
+        Returns:
+            Dictionary with provider, base_url, model, and api_key
+        """
+        if self.EMBEDDING_PROVIDER == "ollama":
+            return {
+                "provider": "ollama",
+                "base_url": self.EMBEDDING_BASE_URL or self.OLLAMA_BASE_URL,
+                "model": self.EMBEDDING_MODEL or self.OLLAMA_EMBEDDING_MODEL,
+                "api_key": self.EMBEDDING_API_KEY
+            }
+        elif self.EMBEDDING_PROVIDER == "vllm":
+            return {
+                "provider": "vllm",
+                "base_url": self.EMBEDDING_BASE_URL,
+                "model": self.EMBEDDING_MODEL,
+                "api_key": self.EMBEDDING_API_KEY
+            }
+        elif self.EMBEDDING_PROVIDER == "openai":
+            return {
+                "provider": "openai",
+                "base_url": self.EMBEDDING_BASE_URL or "https://api.openai.com/v1",
+                "model": self.EMBEDDING_MODEL,
+                "api_key": self.EMBEDDING_API_KEY
+            }
+        else:
+            return {
+                "provider": self.EMBEDDING_PROVIDER,
+                "base_url": self.EMBEDDING_BASE_URL,
+                "model": self.EMBEDDING_MODEL,
+                "api_key": self.EMBEDDING_API_KEY
+            }
+    
     def get_data_sources(self) -> List[BaseDataSourceConfig]:
         """
         Get list of data source configurations

rag/rag_engine.py

@@ -5,6 +5,7 @@ from llama_index.core.query_engine import RetrieverQueryEngine
 from llama_index.core.response_synthesizers import ResponseMode
 from llama_index.core.base.response.schema import StreamingResponse
 from llama_index.llms.ollama import Ollama
+from llama_index.llms.openai import OpenAI
 from llama_index.core import PromptTemplate
 from typing import AsyncIterator, Optional, Tuple
 import asyncio
@@ -82,42 +83,71 @@ class RAGEngine:
     """Main RAG engine for query processing"""
     
     @staticmethod
-    def check_ollama_connection() -> Tuple[bool, str]:
+    def check_llm_connection(provider: str, base_url: str, model: str, api_key: Optional[str] = None) -> Tuple[bool, str]:
         """
-        Check if Ollama server is accessible and connection can be established
+        Check if LLM server is accessible and connection can be established
         
+        Args:
+            provider: LLM provider (ollama, vllm, openai)
+            base_url: Base URL for the LLM service
+            model: Model name
+            api_key: Optional API key for authentication
+            
         Returns:
             Tuple of (is_connected: bool, error_message: str)
             If connected, error_message will be empty string
         """
         try:
-            logger.info(f"Checking Ollama connection to {settings.OLLAMA_BASE_URL}...")
+            headers = {}
+            if api_key:
+                headers["Authorization"] = f"Bearer {api_key}"
             
-            # Test connection by calling Ollama API
-            with httpx.Client(timeout=10.0) as client:
-                response = client.get(f"{settings.OLLAMA_BASE_URL}/api/tags")
-                if response.status_code == 200:
-                    models = response.json().get("models", [])
-                    model_names = [m.get("name", "unknown") for m in models]
-                    logger.info(f"✓ Ollama connection successful (found {len(models)} model(s): {', '.join(model_names[:3])}{'...' if len(model_names) > 3 else ''})")
+            if provider == "ollama":
+                logger.info(f"Checking Ollama connection to {base_url}...")
+                with httpx.Client(timeout=10.0) as client:
+                    response = client.get(f"{base_url}/api/tags")
+                    if response.status_code == 200:
+                        models = response.json().get("models", [])
+                        model_names = [m.get("name", "unknown") for m in models]
+                        logger.info(f"✓ Ollama connection successful (found {len(models)} model(s): {', '.join(model_names[:3])}{'...' if len(model_names) > 3 else ''})")
+                        return True, ""
+                    else:
+                        error_message = f"Ollama API returned status {response.status_code}: {response.text}"
+                        logger.error(f"✗ Ollama connection failed: {error_message}")
+                        return False, error_message
+            elif provider == "openai":
+                logger.info(f"Checking OpenAI-compatible API connection to {base_url}...")
+                if not api_key:
+                    logger.warning("OpenAI provider requires API key, skipping connection check")
                     return True, ""
-                else:
-                    error_message = f"Ollama API returned status {response.status_code}: {response.text}"
-                    logger.error(f"✗ Ollama connection failed: {error_message}")
-                    return False, error_message
+                with httpx.Client(timeout=10.0, headers=headers) as client:
+                    response = client.get(f"{base_url}/models")
+                    if response.status_code == 200:
+                        models = response.json().get("data", [])
+                        model_names = [m.get("id", "unknown") for m in models]
+                        logger.info(f"✓ OpenAI-compatible API connection successful (found {len(models)} model(s): {', '.join(model_names[:3])}{'...' if len(model_names) > 3 else ''})")
+                        return True, ""
+                    else:
+                        error_message = f"OpenAI API returned status {response.status_code}: {response.text}"
+                        logger.error(f"✗ OpenAI connection failed: {error_message}")
+                        return False, error_message
+            else:
+                logger.warning(f"Unknown LLM provider: {provider}, skipping connection check")
+                return True, ""
+                
         except httpx.ConnectError as e:
-            error_message = f"Cannot connect to Ollama server at {settings.OLLAMA_BASE_URL}. " \
-                          f"Please check if Ollama server is running and accessible."
-            logger.error(f"✗ Ollama connection failed: {error_message}")
+            error_message = f"Cannot connect to {provider} server at {base_url}. " \
+                          f"Please check if server is running and accessible."
+            logger.error(f"✗ {provider} connection failed: {error_message}")
             return False, error_message
         except httpx.TimeoutException:
-            error_message = f"Connection to Ollama server at {settings.OLLAMA_BASE_URL} timed out. " \
-                          f"Please check if Ollama server is running and accessible."
-            logger.error(f"✗ Ollama connection failed: {error_message}")
+            error_message = f"Connection to {provider} server at {base_url} timed out. " \
+                          f"Please check if server is running and accessible."
+            logger.error(f"✗ {provider} connection failed: {error_message}")
             return False, error_message
         except Exception as e:
-            error_message = f"Unexpected error while checking Ollama connection: {str(e)}"
-            logger.error(f"✗ Ollama connection check failed: {error_message}")
+            error_message = f"Unexpected error while checking {provider} connection: {str(e)}"
+            logger.error(f"✗ {provider} connection check failed: {error_message}")
             return False, error_message
     
     def __init__(
@@ -129,34 +159,56 @@ class RAGEngine:
         request_timeout: float = 120.0,
     ):
         self.vector_store_manager = vector_store_manager
-        # Configurable LLM / prompt parameters
         self._user_prompt_template = prompt_template
         self._system_prompt = system_prompt
         self._temperature = temperature
         self._request_timeout = request_timeout
         
-        # Check Ollama connection before initializing
-        is_connected, error_message = self.check_ollama_connection()
+        llm_config = settings.get_llm_config()
+        provider = llm_config["provider"]
+        base_url = llm_config["base_url"]
+        model = llm_config["model"]
+        api_key = llm_config["api_key"]
+        
+        is_connected, error_message = self.check_llm_connection(provider, base_url, model, api_key)
         if not is_connected:
             error_msg = (
-                f"Error: Cannot connect to Ollama server.\n"
+                f"Error: Cannot connect to {provider} server.\n"
                 f"  {error_message}\n"
-                f"Connection info: {settings.OLLAMA_BASE_URL}\n"
+                f"Connection info: {base_url}\n"
                 f"Please check:\n"
-                f"  1. Ollama server is running\n"
-                f"  2. Ollama server is accessible from this host\n"
-                f"  3. OLLAMA_BASE_URL is correctly configured\n"
-                f"  4. Firewall rules allow connection to Ollama port"
+                f"  1. {provider} server is running\n"
+                f"  2. {provider} server is accessible from this host\n"
+                f"  3. LLM_BASE_URL is correctly configured\n"
+                f"  4. Firewall rules allow connection"
             )
             logger.error(error_msg)
             raise RuntimeError(error_msg)
         
-        self.llm = Ollama(
-            model=settings.OLLAMA_MODEL,
-            base_url=settings.OLLAMA_BASE_URL,
-            temperature=self._temperature,
-            request_timeout=self._request_timeout,
-        )
+        if provider == "ollama":
+            self.llm = Ollama(
+                model=model,
+                base_url=base_url,
+                temperature=self._temperature,
+                request_timeout=self._request_timeout,
+            )
+        elif provider == "openai":
+            self.llm = OpenAI(
+                model=model,
+                base_url=base_url,
+                api_key=api_key,
+                temperature=self._temperature,
+                timeout=self._request_timeout,
+            )
+        else:
+            self.llm = Ollama(
+                model=model,
+                base_url=base_url,
+                temperature=self._temperature,
+                request_timeout=self._request_timeout,
+            )
+        
+        self._llm_provider = provider
 
     def extract_text_from_chunk(self, chunk) -> Optional[str]:
         if hasattr(chunk, 'delta'):

rag/vector_store.py

@@ -11,6 +11,7 @@ from chromadb.config import Settings as ChromaSettings
 from llama_index.vector_stores.chroma import ChromaVectorStore
 from llama_index.core import VectorStoreIndex, StorageContext
 from llama_index.embeddings.ollama import OllamaEmbedding
+from llama_index.embeddings.openai import OpenAIEmbedding
 from loguru import logger
 from config import settings
 
@@ -106,11 +107,31 @@ class VectorStoreManager:
                     metadata={"hnsw:space": "cosine"}
                 )
                 
-                # Initialize embedding model
-                self.embed_model = OllamaEmbedding(
-                    model_name=settings.OLLAMA_EMBEDDING_MODEL,
-                    base_url=settings.OLLAMA_BASE_URL
-                )
+                # Initialize embedding model based on provider
+                embed_config = settings.get_embedding_config()
+                provider = embed_config["provider"]
+                base_url = embed_config["base_url"]
+                model = embed_config["model"]
+                api_key = embed_config["api_key"]
+                
+                if provider == "ollama":
+                    self.embed_model = OllamaEmbedding(
+                        model_name=model,
+                        base_url=base_url
+                    )
+                elif provider == "openai":
+                    self.embed_model = OpenAIEmbedding(
+                        model=model,
+                        base_url=base_url,
+                        api_key=api_key
+                    )
+                else:
+                    self.embed_model = OllamaEmbedding(
+                        model_name=model,
+                        base_url=base_url
+                    )
+                
+                logger.info(f"Initialized {provider} embedding model: {model}")
                 
                 # Create ChromaVectorStore
                 self.vector_store = ChromaVectorStore(chroma_collection=self.collection)