Fizgig 6.6 adds edit-LoRA training for Qwen Image 2.1, learns grading from 40 pairs
shootthesound · reddit · 2026-09-29
Fizgig 6.6.0 can now train edit LoRAs for Qwen Image 2.1: feed it original/edited photo pairs and it learns to apply that edit to new images. The author trained a film-grade LoRA on 40 before/after pairs using the Qwen 2.1 Fast preset (rank 8) plus Edit LoRA — it changes color and tone only and generalizes to unseen photos. Two Edit presets ship: rank 8 for most edits, rank 16 for trickier ones. Release notes on GitHub.
More from Multimodal
- Testing Qwen Image 2.1 and Minimax H3 on anime-to-photo conversion and film photo restoration — Ant_6431 · 2026-11-16
- Second take of a 'Cubee run' video generated with Minimax H3 — JahJedi · 2026-11-16
- Non-artist builds 14-chapter hand-painted storybook game with Fable 5.2 in 4 hours — jradoff · 2026-11-16
- Subject LoRAs on Minimax H3 Image learn poor likenesses while identical datasets excel on Qwen 2.1 — Ad_Organic · 2026-11-16
- Seedance-generated anime train scene stuns Reddit with cinematic quality — Ok-Solution-7889 · 2026-11-16
- AI-generated 'Men in Black' reimagined in 1977 goes viral — techhalla · 2026-11-16