# Demo: NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4 – transformers pipeline (image-text-to-text)
import torch
from transformers import pipeline

REPO = "nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4"
pipe = pipeline("image-text-to-text", model=REPO, device_map="auto", torch_dtype="auto", trust_remote_code=True)
result = pipe(text=[{"role": "user", "content": [{"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/bee.jpg"}, {"type": "text", "text": "Beschreibe das Bild auf Deutsch."}]}], max_new_tokens=200)
print(result)