Requirements

This pipeline relies on Qwen3VLForConditionalGeneration / Qwen3VLProcessor. Due to upstream changes in transformers >= 5.0, you must pin:

pip install "transformers==4.57.3"                                                                                                                          

Using transformers >= 5.0 will produce visible block-pattern artifacts in the generated image.

DreamLite

ByteDance's UNet-based text-to-image and image-edit diffusion model. 3-branch dual-CFG design, runs at 1024×1024.

import torch
from diffusers import DreamLitePipeline

pipe = DreamLitePipeline.from_pretrained(
    "carlofkl/DreamLite-base", torch_dtype=torch.bfloat16
).to("cuda")
image = pipe("a corgi astronaut", num_inference_steps=28).images[0]

License: CC BY-NC 4.0 (non-commercial). A full model card will be added once the diffusers integration PR is merged.