Instructions to use lightx2v/Qwen-Image-Lightning with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusers
How to use lightx2v/Qwen-Image-Lightning with Diffusers:
pip install -U diffusers transformers accelerate
import torch from diffusers import DiffusionPipeline # switch to "mps" for apple devices pipe = DiffusionPipeline.from_pretrained("Qwen/Qwen-Image", dtype=torch.bfloat16, device_map="cuda") pipe.load_lora_weights("lightx2v/Qwen-Image-Lightning") prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k" image = pipe(prompt).images[0] - Inference
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- Draw Things
- DiffusionBee
Quantization issue with particles on loras, especially with 1.1 and 2.0 versions.
#20
by anyMODE - opened
Anything granular seems to become a large blob, motion trails and dust start to look like rocks or sparks.
realism, An amateur photograph on analog film of a BMW M3 in white with rally decals going at speed on a gravel track through a forest, motion blur and trails of gravel spit up behind it, the number plate reads "2 FAST" and the deca over the window reads "DAZ & SHAZ"
All of these are the same seed. Euler / Simple. Run on the bf16 base model.
4 Step 2.0 @ 8 steps
8 step 1.0 @ 8 steps
8 step 1.1 @ 8 steps
8 step 2.0 @ 8 steps



