feat: enhance LlamaEdgeClient initialization with additional parameters and update run_mcp_server example

Author: Acuspeedster

app/llm_client.py

@@ -1,20 +1,29 @@
 import os
 import json
-from typing import Dict, List, Optional, Union
-from openai import OpenAI
+import requests
+from typing import List, Dict, Any, Optional
+from openai import OpenAI  # Add this import
 
 class LlamaEdgeClient:
     """Client for interacting with LlamaEdge OpenAI-compatible API"""
 
-    def __init__(self, api_key=None):
+    def __init__(self, api_key=None, api_base=None, model=None, embed_model=None):
+        """Initialize LlamaEdgeClient with API credentials
+        
+        Args:
+            api_key: API key for LLM service
+            api_base: Base URL for API (overrides LLM_API_BASE env var)
+            model: Model name (overrides LLM_MODEL env var)
+            embed_model: Embedding model name (overrides LLM_EMBED_MODEL env var)
+        """
         self.api_key = api_key or os.getenv("LLM_API_KEY")
         if not self.api_key:
             raise ValueError("API key is required")
 
-        # Use environment variables with defaults
-        self.base_url = os.getenv("LLM_API_BASE", "http://localhost:8080/v1")
-        self.llm_model = os.getenv("LLM_MODEL", "Qwen2.5-Coder-3B-Instruct")
-        self.llm_embed_model = os.getenv("LLM_EMBED_MODEL", "gte-Qwen2-1.5B-instruct")  # Fixed variable name
+        # Use provided parameters with fallback to environment variables
+        self.base_url = api_base or os.getenv("LLM_API_BASE", "http://localhost:8080/v1")
+        self.llm_model = model or os.getenv("LLM_MODEL", "Qwen2.5-Coder-3B-Instruct")
+        self.llm_embed_model = embed_model or os.getenv("LLM_EMBED_MODEL", "gte-Qwen2-1.5B-instruct")  # Fixed variable name
 
         # Initialize OpenAI client with custom base URL
         self.client = OpenAI(

app/vector_store.py

@@ -2,7 +2,8 @@
 import uuid
 from typing import Dict, List, Optional, Any
 from qdrant_client import QdrantClient
-from qdrant_client.http import models
+from qdrant_client.http import models as qmodels
+from qdrant_client import models  # Add this import
 
 class QdrantStore:
     """Interface for Qdrant vector database"""

examples/run_mcp_server.py

@@ -18,12 +18,11 @@ def main():
     llm_api_base = os.getenv("LLM_API_BASE", "https://coder.gaia.domains/v1")
     llm_model = os.getenv("LLM_MODEL", "Qwen2.5-Coder-32B-Instruct-Q5_K_M")
 
-    llm_client = LlamaEdgeClient(api_key=api_key, api_base=llm_api_base, model=llm_model)
-    
-    # Get Qdrant host and port from environment variables (for Docker)
+    # Get Qdrant connection details
     qdrant_host = os.getenv("QDRANT_HOST", "localhost")
     qdrant_port = int(os.getenv("QDRANT_PORT", "6333"))
 
+    llm_client = LlamaEdgeClient(api_key=api_key, api_base=llm_api_base, model=llm_model)
     vector_store = QdrantStore(embedding_size=llm_embed_size, host=qdrant_host, port=qdrant_port)
     vector_store.create_collection("project_examples")
     vector_store.create_collection("error_examples")