modified

app.py

@@ -5,6 +5,7 @@ from vllm import AsyncLLMEngine, SamplingParams
 from vllm.engine.arg_utils import AsyncEngineArgs
 import asyncio
 import json
+import uuid
 
 app = FastAPI()
 
@@ -32,8 +33,10 @@ async def generate_stream(prompt: str, max_tokens: int, temperature: float):
         temperature=temperature,
         max_tokens=max_tokens
     )
+
+    request_id = str(uuid.uuid4())
     
-    async for output in engine.generate(prompt, sampling_params, 1):  # True enables streaming
+    async for output in engine.generate(prompt, sampling_params, request_id=request_id):  # True enables streaming
         yield f"data: {json.dumps({'text': output.outputs[0].text})}\n\n"
     
     yield "data: [DONE]\n\n"