api: Initial.

README.md

@@ -0,0 +1,14 @@
+---
+title: J.A.R.V.I.S. API Endpoint
+license: apache-2.0
+license_link: https://huggingface.co/hadadrjt/JARVIS/blob/main/LICENSE
+emoji: 👀
+colorFrom: green
+colorTo: green
+sdk: gradio
+sdk_version: 5.34.0
+app_file: app.py
+pinned: false
+models:
+  - hadadrjt/JARVIS
+---

app.py

@@ -0,0 +1,179 @@
+#
+# SPDX-FileCopyrightText: Hadad <hadad@linuxmail.org>
+# SPDX-License-Identifier: Apache-2.0
+#
+
+import json
+import time
+import uuid
+import uvicorn
+
+from contextlib import asynccontextmanager
+from fastapi import FastAPI, HTTPException
+from fastapi.responses import JSONResponse, StreamingResponse
+from gradio_client import Client
+from pydantic import BaseModel
+from typing import AsyncGenerator, Optional
+
+# Default AI model
+MODEL = "JARVIS: 2.1.3"
+
+# Global Gradio client instance
+jarvis: Optional[Client] = None
+
+@asynccontextmanager
+async def lifespan(app: FastAPI):
+    """
+    Initialize Gradio client at app startup.
+    """
+    global jarvis
+    print("Initializing Gradio AI client...")
+    try:
+        jarvis = Client("hadadrjt/ai")
+        print(f"Connected to Gradio AI client at: {jarvis.src}")
+
+        jarvis.predict(new=MODEL, api_name="/change_model")
+        print(f"Default model set to: {MODEL}")
+
+        yield
+    except Exception as e:
+        print(f"Error initializing Gradio client: {e}")
+        yield
+
+app = FastAPI(lifespan=lifespan)
+
+class ResponseRequest(BaseModel):
+    """
+    Request body for /v1/responses endpoint.
+    - model: AI model to use (optional).
+    - input: User input text.
+    - stream: Whether to stream response.
+    """
+    model: Optional[str] = MODEL
+    input: str
+    stream: Optional[bool] = False
+
+async def event_generator(user_input: str, model: str) -> AsyncGenerator[str, None]:
+    """
+    Stream incremental AI responses (deltas) as Server-Sent Events.
+    """
+    global jarvis
+
+    if model != MODEL:
+        jarvis.predict(new=model, api_name="/change_model")
+
+    jarvis_response = jarvis.submit(multi={"text": user_input}, api_name="/api")
+
+    buffer = ""
+
+    try:
+        for partial in jarvis_response:
+            text = partial[0][0][1]
+
+            if text.startswith(buffer):
+                delta = text[len(buffer):]
+            else:
+                delta = text
+
+            buffer = text
+
+            # Skip empty chunks
+            if delta == "":
+                continue
+
+            chunk = {
+                "id": f"chatcmpl-{uuid.uuid4().hex[:8]}",
+                "object": "chat.completion.chunk",
+                "created": int(time.time()),
+                "model": model,
+                "choices": [
+                    {
+                        "index": 0,
+                        "delta": {"content": delta},
+                        "finish_reason": None
+                    }
+                ]
+            }
+
+            yield f"data: {json.dumps(chunk)}\n\n"
+
+        # Final chunk to signal completion
+        done_chunk = {
+            "id": f"chatcmpl-{uuid.uuid4().hex[:8]}",
+            "object": "chat.completion.chunk",
+            "created": int(time.time()),
+            "model": model,
+            "choices": [
+                {
+                    "index": 0,
+                    "delta": {},
+                    "finish_reason": "stop"
+                }
+            ]
+        }
+        yield f"data: {json.dumps(done_chunk)}\n\n"
+
+    except Exception as e:
+        error_chunk = {
+            "error": {"message": f"Streaming error: {str(e)}"}
+        }
+        yield f"data: {json.dumps(error_chunk)}\n\n"
+
+@app.post("/v1/responses")
+async def responses(req: ResponseRequest):
+    """
+    Main endpoint to get AI response.
+    Supports streaming or full JSON response.
+    """
+    global jarvis
+
+    if jarvis is None:
+        raise HTTPException(status_code=503, detail="AI service not initialized or failed to connect.")
+
+    user_input = req.input
+    model = req.model or MODEL
+
+    if req.stream:
+        return StreamingResponse(event_generator(user_input, model), media_type="text/event-stream")
+
+    if model != MODEL:
+        jarvis.predict(new=model, api_name="/change_model")
+
+    jarvis_response = jarvis.submit(multi={"text": user_input}, api_name="/api")
+
+    buffer = ""
+    for partial in jarvis_response:
+        text = partial[0][0][1]
+        buffer = text
+
+    response = {
+        "id": f"chatcmpl-{uuid.uuid4().hex[:8]}",
+        "object": "chat.completion",
+        "created": int(time.time()),
+        "model": model,
+        "choices": [
+            {
+                "index": 0,
+                "message": {
+                    "role": "assistant",
+                    "content": buffer
+                },
+                "finish_reason": "stop"
+            }
+        ]
+    }
+
+    return JSONResponse(response)
+
+@app.get("/")
+def root():
+    """
+    Health check endpoint.
+    """
+    if jarvis:
+        return {"status": "API is running", "jarvis_service": True}
+    else:
+        return {"status": "API is running", "jarvis_service": False, "message": "AI service not ready."}
+
+if __name__ == "__main__":
+    uvicorn.run(app, host="0.0.0.0", port=7860)

requirements.txt

@@ -0,0 +1,4 @@
+fastapi
+uvicorn
+gradio_client
+rich