Great for testing ideas by generating a 1MP image @ 5 steps:
For editing, only ~0.56 credits/tokens
For t2i, only ~0.36 credits/tokens
Then, once the results look good, increase to 6 steps, or delete this LoRA and use Qwen-Image-2.1 at full strength 😉
These turbo LoRAs work best for generating ~1MP images; anything larger and the quality starts to break down. Below are some ideal dimensions for generating with these LoRAs on the TA/TH Creation workspace:
1024 x 1024
768 x 1152 (2:3 aspect ratio) // 1152 x 768 (3:2)
864 x 1152 (3:4) // 1152 x 864 (4:3)
864 x 1536 (9:16) // 1536 x 864 (16:9)
*******
From LoRA creator Viggle:
"We think 6 steps is close to its capacity. Since v0.2.1, every gain we found at 6 steps cost something elsewhere: sharper came with more grain, less grain came with a softer look."
"Built with Qwen. A few-step distilled student of Qwen/Qwen-Image-2.1, trained by Viggle with Distribution Matching Distillation. It does both text-to-image and instruction-driven editing with 1–3 reference images in 6 transformer passes instead of 40, with no classifier-free guidance.
About 5× faster than the 40-step base model end to end, and very competitive with it in quality: on the official Qwen examples the two are hard to tell apart on most prompts. The clearest gap is small, dense text, where the base model is still ahead (8 steps narrows it). See for yourself in the Comparison tab of the demo Space: 32 examples of the Qwen/Qwen-Image-2.1 Space, turbo in 6 steps (and 8 on the 5 dense-text examples) vs base in 40 steps, same prompt, inputs and seed, one sample each, in an image slider. Complicated edits can still fall short of the base model (Known limitations)."
For more info: Qwen-Image-2.1-viggle-turbo