From 5c05617e79184ad3406c75cd1038cb7c6ea063de Mon Sep 17 00:00:00 2001
From: letai2001 <nguyenletai16122001@gmail.com>
Date: Mon, 6 Oct 2025 23:37:59 +0700
Subject: [PATCH] s2

---
 utils/api_llm.py         | 12 ++++++++
 utils/llm_client_vllm.py | 66 ++++++++++++++++++++++++++++++++++++++++
 2 files changed, 78 insertions(+)
 create mode 100644 utils/api_llm.py
 create mode 100644 utils/llm_client_vllm.py

diff --git a/utils/api_llm.py b/utils/api_llm.py
new file mode 100644
index 0000000..366915b
--- /dev/null
+++ b/utils/api_llm.py
@@ -0,0 +1,12 @@
+import google.generativeai as genai
+import time
+genai.configure(api_key="AIzaSyDWqNUBKhaZjbFI8CW52_hKr46JtWABkGU")
+
+model = genai.GenerativeModel("models/gemini-2.0-flash-001")
+start_time = time.time()
+response = model.generate_content("Xin chào, bạn là ai?")
+print(response.text)
+print("Thời gian thực thi: %.2f giây" % (time.time() - start_time))
+# for m in genai.list_models():
+#     if "generateContent" in m.supported_generation_methods:
+#         print(m.name)
diff --git a/utils/llm_client_vllm.py b/utils/llm_client_vllm.py
new file mode 100644
index 0000000..b071e85
--- /dev/null
+++ b/utils/llm_client_vllm.py
@@ -0,0 +1,66 @@
+# -*- coding: utf-8 -*-
+"""
+LLM Client — tương tác với vLLM server
+-------------------------------------
+- Dùng HTTP endpoint (OpenAI-compatible) của vLLM.
+- Nhận prompt đã được build từ PromptBuilder.
+- Gọi model sinh câu trả lời.
+"""
+
+import requests
+from typing import Optional, Dict, Any
+from src.core.config import VLLM_URL, VLLM_MODEL, LLM_TEMPERATURE, LLM_MAX_TOKENS
+
+
+class LLMClient:
+    """
+    Gọi model vLLM theo giao thức OpenAI-compatible.
+    """
+    def __init__(
+        self,
+        base_url: str = VLLM_URL,
+        model: str = VLLM_MODEL,
+        temperature: float = LLM_TEMPERATURE,
+        max_tokens: int = LLM_MAX_TOKENS
+    ):
+        self.base_url = base_url.rstrip("/")
+        self.model = model
+        self.temperature = temperature
+        self.max_tokens = max_tokens
+
+    def generate(self, prompt: str, system_prompt: Optional[str] = None) -> str:
+        """
+        Gửi prompt tới vLLM API và trả về câu trả lời.
+        """
+        url = f"{self.base_url}/v1/chat/completions"
+
+        messages = []
+        if system_prompt:
+            messages.append({"role": "system", "content": system_prompt})
+        messages.append({"role": "user", "content": prompt})
+
+        payload: Dict[str, Any] = {
+            "model": self.model,
+            "messages": messages,
+            "temperature": self.temperature,
+            "max_tokens": self.max_tokens,
+            "stream": False,
+        }
+
+        try:
+            response = requests.post(url, json=payload, timeout=60)
+            response.raise_for_status()
+            data = response.json()
+            return data["choices"][0]["message"]["content"].strip()
+        except Exception as e:
+            print(f"[LLMClient] ❌ Error calling vLLM: {e}")
+            return f"[Error] {e}"
+
+
+# ---- Test nhanh ----
+if __name__ == "__main__":
+    client = LLMClient()
+    prompt = "Viết đoạn mô tả ngắn về hành tinh Sao Hỏa."
+    answer = client.generate(prompt)
+    print("🪐 Kết quả từ LLM:")
+    print(answer)