User trains open-source Gaussian Splash LoRA for Qwen-Image 2.1, prompted instead of plug-and-play
Spy4killer · reddit · 2026-10-26
Reddit user Spy4killer loved the Gaussian Splash LoRA from Qwen-Image-Edit-2511 and wanted the same effect on Qwen-Image 2.1, so he trained his own and released it.
Key points:
- New LoRA is on GitHub (aaljman/splat-repair-lora); the original 2511 version is on HuggingFace (dx8152)
- Unlike the plug-and-play 2511 LoRA, this version is triggered via prompting, trading convenience for more control — fans of the old plug-and-play flow may be disappointed
More from Multimodal
- Magnific teases October 28 release: 'We're making it move' — cuenca · 2026-10-27
- Midjourney --sref demo: retro Americana Batman art series — gizakdag · 2026-10-27
- AI video is a new medium, not cinema or animation — and it's perfect for depicting dreams — Kyrannio · 2026-10-27
- Dad turns his toddler's 31-year-old quote into a song via SUNO kids' choir — NathanpmYoung · 2026-10-27
- Band's local-AI music video hits 200k views: 900 generations kept 64, 3-second clips beat face drift — Hwang0500 · 2026-10-27
- Adobe Firefly video editor walkthrough: one-sentence Quick Cut turns 32s into 20s — LudovicCreator · 2026-10-26