{"slug": "truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step", "title": "TrueRelight: Directional & Physical Relighting LoRA for FLUX.3 Image Klein (16-step distilled, Open Weights + 280-Pair Dataset + Interactive Space)", "summary": "A developer publishing as pandacooming released TrueRelight, an open-source directional and physical relighting LoRA for Black Forest Labs' FLUX.3 Image Klein, trained on the flux-3-image-klein-base model and evaluated on the 16-step distilled flux-3-image-klein. The LoRA was trained on 280 pixel-aligned <ref_image_0> pairs at rank=32 and 1024px, and the author reports that popping the prompt_upsampler block during 16-step inference lets TrueRelight reverse light direction instead of producing the flat color wash seen with prompt-only relighting. The release includes a 160.7 MB pytorch_lora_weights.safetensors file, a 280-pair Hugging Face dataset, a demo notebook, and ComfyUI compatibility at strength 0.8–1.0 with 16 steps, CFG 1.0, and euler/simple samplers.", "body_md": "Hi everyone! I just open-sourced **TrueRelight**, a physical & directional relighting LoRA trained on `black-forest-labs/flux-3-image-klein-base` and evaluated on the 16-step distilled `black-forest-labs/flux-3-image-klein`.\n\nPlain prompt-only relighting on distilled FLUX.3 Image Klein often behaves like a flat color wash (e.g., tinting existing shadows orange/yellow instead of reversing light direction). By training on **280 pixel-aligned `<ref_image_0>` pairs** (`rank=32`, `1024px`) and **popping `prompt_upsampler`** during 16-step inference, TrueRelight can:\n\n`pytorch_lora_weights.safetensors`, 160.7 MB):` https://huggingface.co/datasets/pandacooming/FLUX-3-Image-Klein-TrueRelight-Dataset``https://huggingface.co/pandacooming/FLUX-3-Image-Klein-TrueRelight-LoRA/blob/main/FLUX_3_Klein_TrueRelight_Demo.ipynb``diffusers`)\nPop `prompt_upsampler` before running the 16-step pipeline so the `<ref_image_0>` grammar reaches the LoRA cleanly:\n\n```\nblocks = ModularPipeline.from_pretrained(\"black-forest-labs/flux-3-image-klein\").blocks\nblocks.sub_blocks.pop(\"prompt_upsampler\", None)\npipe = blocks.init_pipeline(\"black-forest-labs/flux-3-image-klein\")\n```\n\n(Also compatible with `flux-3-image-klein-comfy` at strength `0.8–1.0`, 16 steps, CFG 1.0, `euler`/` simple`, skipping the prompt upsampler.)\n\nWould love feedback from anyone testing it on portraits, tricky shadows, or ComfyUI workflows!", "url": "https://wpnews.pro/news/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step", "canonical_source": "https://discuss.huggingface.co/t/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step-distilled-open-weights-280-pair-dataset-interactive-space/191946#post_1", "published_at": "2026-11-16 13:39:12+00:00", "updated_at": "2026-11-16 13:50:15.161108+00:00", "lang": "en", "topics": ["generative-ai", "ai-tools", "ai-research", "large-language-models"], "entities": ["TrueRelight", "FLUX.3 Image Klein", "Black Forest Labs", "pandacooming", "Hugging Face", "ComfyUI", "diffusers"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step", "markdown": "https://wpnews.pro/news/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step.md", "text": "https://wpnews.pro/news/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step.txt", "jsonld": "https://wpnews.pro/news/truerelight-directional-physical-relighting-lora-for-flux-3-image-klein-16-step.jsonld"}}