| import os |
| from fastapi import FastAPI, Request |
| from fastapi.responses import HTMLResponse |
| from openai import OpenAI |
|
|
| app = FastAPI() |
|
|
| MODEL = "nousresearch/hermes-2-pro-llama-3-8b" |
|
|
| def get_client(): |
| key = os.getenv("NVIDIA_API_KEY") |
| if not key: |
| raise RuntimeError("❌ NVIDIA_API_KEY missing. Add it in Settings → Variables & secrets → Factory reboot.") |
| return OpenAI(base_url="https://integrate.api.nvidia.com/v1", api_key=key) |
|
|
| @app.get("/", response_class=HTMLResponse) |
| async def home(): |
| return """ |
| <!DOCTYPE html> |
| <html><head><title>Hermes NIM</title> |
| <style> |
| body{font-family:system-ui,sans-serif;max-width:700px;margin:40px auto;padding:20px;background:#0f0f0f;color:#e0e0e0} |
| textarea{width:100%;height:120px;padding:12px;background:#1a1a1a;color:#fff;border:1px solid #333;border-radius:8px;font-size:16px} |
| button{margin-top:12px;padding:12px 24px;background:#7c3aed;color:#fff;border:none;border-radius:8px;font-size:16px;cursor:pointer} |
| button:hover{background:#6d28d9} |
| pre{background:#161616;padding:16px;border-radius:8px;white-space:pre-wrap;margin-top:16px;line-height:1.5} |
| </style></head><body> |
| <h2>🟣 Hermes + NVIDIA NIM</h2> |
| <textarea id="p" placeholder="Type your prompt..."></textarea><br> |
| <button onclick="run()">Run Hermes</button> |
| <pre id="out">Response will appear here...</pre> |
| <script> |
| async function run(){ |
| const o=document.getElementById('out'); o.textContent='⏳ Running...'; |
| try{ |
| const r=await fetch('/run',{method:'POST',headers:{'Content-Type':'application/json'},body:JSON.stringify({prompt:document.getElementById('p').value})}); |
| const d=await r.json(); o.textContent=d.output||d.detail||JSON.stringify(d); |
| }catch(e){o.textContent='❌ '+e.message} |
| } |
| </script></body></html> |
| """ |
|
|
| @app.post("/run") |
| async def run_hermes(req: Request): |
| try: |
| data = await req.json() |
| prompt = data.get("prompt", "") |
| if not prompt: |
| return {"error": "Missing prompt"} |
| client = get_client() |
| res = client.chat.completions.create( |
| model=MODEL, |
| messages=[{"role": "user", "content": prompt}], |
| temperature=0.7, |
| max_tokens=1024 |
| ) |
| return {"status": "done", "output": res.choices[0].message.content} |
| except Exception as e: |
| return {"status": "error", "detail": str(e)} |
|
|
|
|
