Merge pull request #316 from magicyuan876/patch-1

使用AzureOpenAI实现，支持RPM/TPM限制。修复原先429响应即抛出异常的问题

examples/lightrag_azure_openai_demo.py

@@ -4,8 +4,8 @@ from lightrag import LightRAG, QueryParam
 from lightrag.utils import EmbeddingFunc
 import numpy as np
 from dotenv import load_dotenv
-import aiohttp
 import logging
+from openai import AzureOpenAI
 
 logging.basicConfig(level=logging.INFO)
 
@@ -32,11 +32,11 @@ os.mkdir(WORKING_DIR)
 async def llm_model_func(
     prompt, system_prompt=None, history_messages=[], **kwargs
 ) -> str:
-    headers = {
+    client = AzureOpenAI(
-        "Content-Type": "application/json",
+        api_key=AZURE_OPENAI_API_KEY,
-        "api-key": AZURE_OPENAI_API_KEY,
+        api_version=AZURE_OPENAI_API_VERSION,
-    }
+        azure_endpoint=AZURE_OPENAI_ENDPOINT,
-    endpoint = f"{AZURE_OPENAI_ENDPOINT}openai/deployments/{AZURE_OPENAI_DEPLOYMENT}/chat/completions?api-version={AZURE_OPENAI_API_VERSION}"
+    )
 
     messages = []
     if system_prompt:
@@ -45,41 +45,26 @@ async def llm_model_func(
         messages.extend(history_messages)
     messages.append({"role": "user", "content": prompt})
 
-    payload = {
+    chat_completion = client.chat.completions.create(
-        "messages": messages,
+        model=AZURE_OPENAI_DEPLOYMENT,  # model = "deployment_name".
-        "temperature": kwargs.get("temperature", 0),
+        messages=messages,
-        "top_p": kwargs.get("top_p", 1),
+        temperature=kwargs.get("temperature", 0),
-        "n": kwargs.get("n", 1),
+        top_p=kwargs.get("top_p", 1),
-    }
+        n=kwargs.get("n", 1),
-
+    )
-    async with aiohttp.ClientSession() as session:
+    return chat_completion.choices[0].message.content
-        async with session.post(endpoint, headers=headers, json=payload) as response:
-            if response.status != 200:
-                raise ValueError(
-                    f"Request failed with status {response.status}: {await response.text()}"
-                )
-            result = await response.json()
-            return result["choices"][0]["message"]["content"]
 
 
 async def embedding_func(texts: list[str]) -> np.ndarray:
-    headers = {
+    client = AzureOpenAI(
-        "Content-Type": "application/json",
+        api_key=AZURE_OPENAI_API_KEY,
-        "api-key": AZURE_OPENAI_API_KEY,
+        api_version=AZURE_EMBEDDING_API_VERSION,
-    }
+        azure_endpoint=AZURE_OPENAI_ENDPOINT,
-    endpoint = f"{AZURE_OPENAI_ENDPOINT}openai/deployments/{AZURE_EMBEDDING_DEPLOYMENT}/embeddings?api-version={AZURE_EMBEDDING_API_VERSION}"
+    )
+    embedding = client.embeddings.create(model=AZURE_EMBEDDING_DEPLOYMENT, input=texts)
 
-    payload = {"input": texts}
+    embeddings = [item.embedding for item in embedding.data]
-
+    return np.array(embeddings)
-    async with aiohttp.ClientSession() as session:
-        async with session.post(endpoint, headers=headers, json=payload) as response:
-            if response.status != 200:
-                raise ValueError(
-                    f"Request failed with status {response.status}: {await response.text()}"
-                )
-            result = await response.json()
-            embeddings = [item["embedding"] for item in result["data"]]
-            return np.array(embeddings)
 
 
 async def test_funcs():