modified

app.py

@@ -33,7 +33,7 @@ async def generate_stream(prompt: str, max_tokens: int, temperature: float):
         max_tokens=max_tokens
     )
     
-    async for output in engine.generate(prompt, sampling_params):  # True enables streaming
+    async for output in engine.generate(prompt, sampling_params, 1):  # True enables streaming
         yield f"data: {json.dumps({'text': output.outputs[0].text})}\n\n"
     
     yield "data: [DONE]\n\n"