I made a LoRA loader that schedules strength against the exact <code>SIGMAS</code> tensor used by the sampler. The main feature is explicit mode. Instead of relying on a start/end percentage or a preset curve, you enter one strength for every model-call sigma:eg. sigmas:<pre><code>1.0000, 0.9567, 0.9045, 0.8403, 0.7595, 0.6546, 0.5128, 0.3109, 0.0000</code></pre>explicit_strengths example (obv this is not a recommended set this is just a demo of the node's capability)<span style="background-color:rgb(0, 19, 41);color:rgb(162, 195, 254);font-family:-apple-system, BlinkMacSystemFont, &quot;Segoe UI&quot;, Roboto, &quot;Helvetica Neue&quot;, Arial, &quot;Apple Color Emoji&quot;, &quot;Segoe UI Emoji&quot;, &quot;Segoe UI Symbol&quot;, sans-serif;font-size:14px"> :<pre><code>0.75, 0.75, 0.75, 0.75, 1.20, 1.40, 1.20, 1.30</code></pre>The terminal zero is not a model call, so nine sigma values require eight strengths.This makes it possible to increase, reduce, or maintain LoRA influence at specific parts of a custom schedule instead of applying one strength across the entire generation. Intermediate sampler evaluations are interpolated in sigma space.The node also includes linear, cosine, smoothstep, and power curves when exact manual control is unnecessary. It does not merge the LoRA into the base weights and does not modify the supplied sigma schedule. Developed initially for my refusal reduction lora for Krea2 as I noticed one universal strength was not cutting it for me, but the loader itself is not Krea-specific; compatibility depends on ComfyUI recognizing the model's LoRA keys.GitHub: <a target="_blank" rel="ugc" href="https://github.com/capitan01R/ComfyUI-SigmaSync-LoRA">https://github.com/capitan01R/ComfyUI-SigmaSync-LoRA</a>

Files with SHA256: 49e3d089377653bcb3ce292b66ce167fcda9a2dc2a1d9e00c9e8c7cc8996c7fb