app.py

import os
import requests
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import JSONResponse
from pydantic import BaseModel, Field
import gradio as gr

# ---------------- CONFIG ----------------
MODEL_ID = os.getenv("MODEL_ID", "HuggingFaceH4/zephyr-7b-beta")
HF_TOKEN = os.getenv("HF_TOKEN")  # must be set in Space secrets

API_URL = f"https://api-inference.huggingface.co/models/{MODEL_ID}"
HEADERS = {"Authorization": f"Bearer {HF_TOKEN}"}

app = FastAPI(title="Edu Assistant API 🎉")

# ---------------- CORS ----------------
app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=False,
    allow_methods=["*"],
    allow_headers=["*"],
)

# ---------------- MODELS ----------------
class ChatRequest(BaseModel):
    user: str
    temperature: float = Field(default=0.7, ge=0.0, le=2.0)
    max_tokens: int = Field(default=512, gt=1, le=1024)
    style: str = Field(default="detailed", pattern="^(short|detailed|no_explanation)$")

# ---------------- HELPERS ----------------
def build_prompt(user: str, style: str) -> str:
    style_note = {
        "short": "Answer briefly.",
        "detailed": "Explain step by step, with details.",
        "no_explanation": "Just give the answer, no explanation."
    }[style]

    return (
        "You are a helpful multilingual tutor. Use emojis to make responses engaging.\n\n"
        f"User: {user}\n"
        f"Instruction: {style_note}\n\n"
        "Assistant:"
    )

def query_hf(prompt: str, max_tokens: int, temperature: float) -> str:
    payload = {
        "inputs": prompt,
        "parameters": {
            "max_new_tokens": max_tokens,
            "temperature": temperature,
            "top_p": 0.9,
            "do_sample": True,
            "return_full_text": False
        }
    }
    response = requests.post(API_URL, headers=HEADERS, json=payload, timeout=60)
    if response.status_code != 200:
        raise RuntimeError(f"HF API Error {response.status_code}: {response.text}")
    data = response.json()
    if isinstance(data, list) and "generated_text" in data[0]:
        return data[0]["generated_text"].strip()
    else:
        return str(data)

# ---------------- FASTAPI ENDPOINT ----------------
@app.post("/chat")
def chat(req: ChatRequest):
    try:
        prompt = build_prompt(req.user, req.style)
        text = query_hf(prompt, req.max_tokens, req.temperature)
        return {"answer": "💡 " + text}
    except RuntimeError as e:
        return JSONResponse({"error": str(e)}, status_code=502)

# ---------------- GRADIO WRAPPER ----------------
def gradio_chat(user, temperature=0.7, max_tokens=512, style="detailed"):
    prompt = build_prompt(user, style)
    try:
        text = query_hf(prompt, max_tokens, temperature)
        return text
    except Exception as e:
        return f"[error] {str(e)}"

# Create Gradio interface
iface = gr.Interface(
    fn=gradio_chat,
    inputs=[
        gr.Textbox(label="User Message"),
        gr.Slider(0, 2, value=0.7, label="Temperature"),
        gr.Slider(1, 1024, value=512, label="Max Tokens"),
        gr.Dropdown(["short", "detailed", "no_explanation"], value="detailed", label="Style")
    ],
    outputs=gr.Textbox(label="Assistant Response"),
    title="Edu Assistant API 🎉"
)

# Launch Gradio on HF Spaces
iface.launch(server_name="0.0.0.0", server_port=7860, share=True)