Update app.py

app.py

@@ -1,35 +1,20 @@
 import os
-from fastapi import FastAPI, HTTPException
+from fastapi import FastAPI, Request
 from fastapi.responses import StreamingResponse
 from openai import AsyncOpenAI
-from pydantic import BaseModel
-import httpx
+import asyncio
 
-# Initialize FastAPI app
 app = FastAPI()
 
-# Define request body model for the prompt
-class PromptRequest(BaseModel):
-    prompt: str
+async def generate_ai_response(prompt: str):
+    # Get GitHub token from environment variable
+    token = os.getenv("GITHUB_TOKEN")
+    endpoint = "https://models.github.ai/inference"
+    model = "openai/gpt-4-1-mini"  # Fixed typo in model name (was gpt-4.1-mini)
 
-# Initialize OpenAI client
-token = os.getenv("GITHUB_TOKEN")
-if not token:
-    raise ValueError("GITHUB_TOKEN environment variable not set")
+    # Initialize OpenAI client
+    client = AsyncOpenAI(base_url=endpoint, api_key=token)
 
-# Use environment variables for endpoint and model, with fallbacks
-endpoint = os.getenv("API_ENDPOINT", "https://api.openai.com/v1")  # Fallback to OpenAI-compatible endpoint
-model = os.getenv("MODEL_NAME", "gpt-4o-mini")  # Default to a known model
-
-# Initialize AsyncOpenAI with a custom HTTP client to avoid proxies issue
-client = AsyncOpenAI(
-    base_url=endpoint,
-    api_key=token,
-    http_client=httpx.AsyncClient()  # Explicitly use httpx.AsyncClient without proxies
-)
-
-# Async generator to stream chunks
-async def stream_response(prompt: str):
     try:
         # Create streaming chat completion
         stream = await client.chat.completions.create(
@@ -43,28 +28,21 @@ async def stream_response(prompt: str):
             stream=True
         )
 
-        # Yield each chunk as it arrives
+        # Process the stream
         async for chunk in stream:
             if chunk.choices and len(chunk.choices) > 0:
                 content = chunk.choices[0].delta.content or ""
-                if content:
-                    yield content
+                yield content
 
     except Exception as err:
-        yield f"Error: {str(err)}"
+        yield f"The sample encountered an error: {err}"
 
-# Endpoint to handle prompt and stream response
 @app.post("/generate")
-async def generate_response(request: PromptRequest):
-    try:
-        return StreamingResponse(
-            stream_response(request.prompt),
-            media_type="text/event-stream"
-        )
-    except Exception as err:
-        raise HTTPException(status_code=500, detail=f"Server error: {str(err)}")
-
-# Health check endpoint for Hugging Face Spaces
-@app.get("/")
-async def health_check():
-    return {"status": "healthy"}
+async def generate_response(request: Request):
+    data = await request.json()
+    prompt = data.get("prompt", "what is ai")  # Default prompt if none provided
+    
+    return StreamingResponse(
+        generate_ai_response(prompt),
+        media_type="text/event-stream"
+    )