LTFloat_Step_01

LTFloat_Step_01 is the 0.1 member of the Latent Tools step-helper family, and it's the one you'll actually reach for in normal image work. Where the fine steps exist for noise sweeps, 0.1 is the step size that matches how humans think about cfg, denoise, and sampler parameters. It's a calibrated constant: type a value, get the same number back as a FLOAT and a STRING, with the widget stepping by 0.1.

Sweeping cfg from 6 to 10 by 0.1 gives you forty-one runs that map the whole curve of prompt adherence versus quality - a genuinely useful experiment. A denoise sweep from 0.4 to 1.0 is six runs. A mean sweep on your noise still works, it's just chunkier. When someone says "these pack helper nodes are only for noise hobbyists," this is the one that proves them wrong: the step is generic, and 0.1 is where most of the classic parameter explorations live.

How it works

Same story as the whole family, stated plainly so the name doesn't fool you: the node does not auto-increment. It takes the value you give it, rounds to five decimals, and returns it as a FLOAT and a STRING. The step is the widget's spinner increment and the label. You bump the value between runs.

The inputs and outputs

• value - FLOAT, default 0, range ±1,000,000, step 0.1.
• FLOAT - the value.
• STRING - the value as text, for filename prefixes so every sweep output records its parameter.

Wire the FLOAT into any float input - cfg, denoise, mean, std - and the STRING into a Save Image filename. Fixed seed, one variable changed per run, compare.

Installing it

Part of xl0's Latent Tools pack. ComfyUI Manager → search Latent Tools → install → restart, or:

cd ComfyUI/custom_nodes
git clone https://github.com/xl0/latent-tools

Restart, and it's under LatentTools. The pack's only dependency is lovely-tensors (Manager installs it; manual cloners may need pip install lovely-tensors). No model downloads.