Upload 3 files

Dockerfile

@@ -0,0 +1,12 @@
+FROM python:3.11-slim
+
+WORKDIR /app
+
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+COPY app.py .
+
+EXPOSE 7860
+
+CMD ["python", "app.py"]

app.py

@@ -0,0 +1,133 @@
+# Entry point file for Hugging Face Spaces - OpenAI Compatible
+import uvicorn
+from fastapi import FastAPI, HTTPException, Request
+import requests
+from pydantic import BaseModel, Field
+from typing import Optional, List, Dict, Any, Literal
+
+app = FastAPI(title="OpenAI-Compatible Chat API", 
+              description="A FastAPI application that provides an OpenAI-compatible interface")
+
+# Models for OpenAI compatibility
+class Message(BaseModel):
+    role: str
+    content: str
+    name: Optional[str] = None
+
+class ChatCompletionRequest(BaseModel):
+    model: str = "granite-3-2-8b-instruct"
+    messages: List[Message]
+    temperature: Optional[float] = 0.7
+    top_p: Optional[float] = 0.9
+    max_tokens: Optional[int] = 2048
+    stream: Optional[bool] = False
+    
+class ChatCompletionChoice(BaseModel):
+    index: int
+    message: Message
+    finish_reason: str = "stop"
+
+class Usage(BaseModel):
+    prompt_tokens: int
+    completion_tokens: int
+    total_tokens: int
+
+class ChatCompletionResponse(BaseModel):
+    id: str
+    object: str = "chat.completion"
+    created: int
+    model: str
+    choices: List[ChatCompletionChoice]
+    usage: Usage
+
+# Custom endpoints for graniteAI
+@app.post("/v1/chat/completions", response_model=ChatCompletionResponse)
+async def chat_completion(request: ChatCompletionRequest):
+    # Forward to granite API
+    url = "https://d18n68ssusgr7r.cloudfront.net/v1/chat/completions"
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": "Bearer 89de4a8b-9dc6-4617-86a0-28690278b651"
+    }
+    
+    # Convert to GraniteAI format if needed
+    granite_data = {
+        "messages": [{"role": msg.role, "content": msg.content} for msg in request.messages],
+        "model": request.model,
+        "max_tokens": request.max_tokens,
+        "temperature": request.temperature,
+        "top_p": request.top_p
+    }
+    
+    try:
+        response = requests.post(url, headers=headers, json=granite_data)
+        response_json = response.json()
+        
+        # Format into OpenAI-compatible response
+        # This assumes the granite API returns something we can parse
+        # You may need to adjust based on actual granite response
+        
+        # Extract the assistant message
+        assistant_message = ""
+        if "choices" in response_json and len(response_json["choices"]) > 0:
+            assistant_message = response_json["choices"][0]["message"]["content"]
+        else:
+            # Fallback in case the response structure is different
+            assistant_message = str(response_json)
+        
+        # Estimate token counts (very rough estimation)
+        prompt_tokens = sum(len(msg.content.split()) for msg in request.messages)
+        completion_tokens = len(assistant_message.split())
+        
+        return ChatCompletionResponse(
+            id=f"chatcmpl-{response_json.get('id', 'unknown')}",
+            created=response_json.get("created", 0),
+            model=request.model,
+            choices=[
+                ChatCompletionChoice(
+                    index=0,
+                    message=Message(
+                        role="assistant",
+                        content=assistant_message
+                    )
+                )
+            ],
+            usage=Usage(
+                prompt_tokens=prompt_tokens,
+                completion_tokens=completion_tokens,
+                total_tokens=prompt_tokens + completion_tokens
+            )
+        )
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+# Alternative version of the endpoint that directly passes through the raw granite API response
+@app.post("/raw/chat/completions")
+async def raw_chat_completion(request: Request):
+    data = await request.json()
+    
+    # Forward to granite API
+    url = "https://d18n68ssusgr7r.cloudfront.net/v1/chat/completions"
+    headers = {
+        "Content-Type": "application/json",
+        "Authorization": "Bearer 89de4a8b-9dc6-4617-86a0-28690278b651"
+    }
+    
+    try:
+        response = requests.post(url, headers=headers, json=data)
+        return response.json()
+    except Exception as e:
+        raise HTTPException(status_code=500, detail=str(e))
+
+@app.get("/")
+async def root():
+    return {
+        "message": "Welcome to the OpenAI-Compatible Chat API",
+        "endpoints": {
+            "/v1/chat/completions": "OpenAI-compatible chat completions endpoint",
+            "/raw/chat/completions": "Direct passthrough to the granite API"
+        }
+    }
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -0,0 +1,5 @@
+fastapi==0.104.1
+uvicorn==0.23.2
+requests==2.31.0
+pydantic==2.4.2
+python-dotenv==1.0.0