import torch
from diffusers import DiffusionPipeline
# switch to "mps" for apple devices
pipe = DiffusionPipeline.from_pretrained("Qwen/Qwen-Image-2512", dtype=torch.bfloat16, device_map="cuda")
pipe.load_lora_weights("ingtellect/qwen-image-2512-fp8-scaled-comfyui")
prompt = "Astronaut in a jungle, cold color palette, muted colors, detailed, 8k"
image = pipe(prompt).images[0]Qwen-Image-2512 FP8 Scaled (ComfyUI)
Qwen-Image-2512 BF16 checkpoint converted to ComfyUI-compatible FP8 scaled format.
- Format check: passed (
_quantization_metadata, per-layerweight_scalein float32 scalar) - Scope: reference release
- No support/maintenance guarantees
Source / Conversion Code
https://github.com/ingtellect/qwen-image-fp8-scaled-comfyui-converter
- Downloads last month
- 1,386
Model tree for ingtellect/qwen-image-2512-fp8-scaled-comfyui
Base model
Qwen/Qwen-Image-2512