Text-to-Image
Diffusers
Safetensors
QwenImagePipeline
nvidia
ModelOpt
Qwen-Image
quantized
FP4
fp4
8-bit precision
Instructions to use nvidia/Qwen-Image-NVFP4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use nvidia/Qwen-Image-NVFP4 with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("nvidia/Qwen-Image-NVFP4", dtype=torch.bfloat16, device_map="cuda") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- Draw Things
- DiffusionBee
Download serving/visualgen-eager.yaml from nvidia/Qwen-Image-NVFP4: direct link, hf CLI and curl.
- Browser
- Download file 199 Bytes
-
https://huggingface.co/nvidia/Qwen-Image-NVFP4/resolve/main/serving/visualgen-eager.yaml
- Command line
-
hf download hf://nvidia/Qwen-Image-NVFP4/serving/visualgen-eager.yaml
-
curl -L -o visualgen-eager.yaml https://huggingface.co/nvidia/Qwen-Image-NVFP4/resolve/main/serving/visualgen-eager.yaml
199 Bytes
| attention_config: | |
| backend: VANILLA | |
| parallel_config: | |
| cfg_size: 1 | |
| ulysses_size: 1 | |
| torch_compile_config: | |
| enable: false | |
| cuda_graph_config: | |
| enable: false | |
| compilation_config: | |
| skip_warmup: true | |