Update main.py

main.py

@@ -7,6 +7,7 @@ import os
 import time
 import httpx
 import json
+import asyncio
 from dotenv import load_dotenv
 
 load_dotenv()
@@ -129,6 +130,7 @@ async def create_completion(model: str, content: str, return_stream: bool, retry
     token = await request_token()
     
     try:
+        # Fixed: Don't pass stream=True to AsyncClient.post() - it's not supported
         async with httpx.AsyncClient() as client:
             response = await client.post(
                 "https://duckduckgo.com/duckchat/v1/chat",
@@ -141,8 +143,7 @@ async def create_completion(model: str, content: str, return_stream: bool, retry
                 json={
                     "model": model,
                     "messages": [{"role": "user", "content": content}]
-                },
-                stream=True
+                }
             )
 
             if response.status_code != 200:
@@ -165,8 +166,9 @@ async def process_stream(model: str, response, return_stream: bool):
         nonlocal buffer, full_text
         
         # Process chunks as they arrive
-        async for chunk in response.aiter_bytes():
-            chunk_str = chunk.decode('utf-8').strip()
+        # Use response.aiter_text() instead of aiter_bytes() for easier text handling
+        async for chunk_str in response.aiter_text():
+            chunk_str = chunk_str.strip()
             
             # Handle buffer from previous chunk if needed
             if buffer:
@@ -208,8 +210,8 @@ async def process_stream(model: str, response, return_stream: bool):
     if return_stream:
         return StreamingResponse(generate_stream(), media_type="text/event-stream")
     else:
-        # For non-streaming, consume the generator and return complete response
-        async for _ in generate_stream():
+        # For non-streaming, collect all the text
+        async for chunk in generate_stream():
             pass  # Just collecting text in full_text
         
         return JSONResponse(content=create_complete_response(full_text, model))
@@ -263,9 +265,6 @@ def create_complete_response(text: str, model: str) -> dict:
         ],
     }
 
-# Only needed for retry delays
-import asyncio
-
 if __name__ == "__main__":
     import uvicorn
     uvicorn.run("app:app", host="0.0.0.0", port=7860, reload=True)