Ino Get Sampler Config

A sampler isn't one setting, it's five that only make sense together: sampler, scheduler, steps, CFG, denoise. That's why workflow after workflow ships with the settings scattered across the canvas. Ino Get Sampler Config is the Ino Nodes pack's attempt to bundle them - it takes a model config string, applies your overrides, and builds the actual sampler pieces ComfyUI's advanced sampling nodes need: a guider, a sampler, and a sigmas schedule.

It comes from the Ino Nodes pack (nobandegani) and sits at the sampling end of the pack's config-driven pipeline. It pairs with Ino Get Model Config - that node hands you the config JSON, this one turns it into something a sampler can run.

How it works

The inputs that matter:

• config - the model config JSON string (from Ino Get Model Config, or hand-built).
• model, positive, negative - the three graph inputs any sampler needs: the loaded model and your two conditionings.
• use_cfg, cfg, sampler_name, scheduler_name, steps, denoise - the overrides. Every one of them has a default that means "keep whatever the config says" (unset for the dropdowns, -1 for the numbers). Only set an override when you want to beat the config's preset.

Internally it merges your overrides into the config, then builds three outputs:

• guider - the guidance wrapper for the model (CFG-based, or a no-CFG guider when the config calls for it, like Flux's guidance-3.5 setup).
• sampler - the sampling algorithm.
• sigmas - the noise schedule.
• old_config / new_config - the config before and after the override merge, so you can see exactly what changed.

These three objects plug into ComfyUI's advanced sampler nodes (SamplerCustomAdvanced and friends), which expect a guider, sampler, and sigmas as separate inputs. That's the pattern: Ino builds the components, the stock sampler runs them.

Where it fits

The win is consistency. Your model config says "euler / simple / 30 steps / guidance 3.5" - this node honors it, and the overrides let you tweak a single run without editing the config. It's the difference between a workflow that silently runs with stale settings and one where the settings travel with the model.

The honest caveat: this is a power-user node. It assumes you know what a guider/sampler/sigmas combination is and which advanced sampler node consumes it. If you've never touched SamplerCustomAdvanced, this will feel like plumbing with no taps. Start from the built-in KSampler; come back to this when you want the settings to live in one string.

Installing it

It ships with the whole Ino Nodes pack. ComfyUI Manager: search "ComfyUI Ino Nodes", install, restart. Or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/nobandegani/ComfyUI-InoNodes.git
cd comfyui_ino_nodes
pip install -r requirements.txt

Restart after. Pack-wide caveats apply: it's built on ComfyUI's V3 schema (README wants v0.18.1+), and requirements.txt installs inopyutils, the helper library every node imports.

Common issues

If you get None out of the sampler outputs, the first thing to check is enabled - when disabled the node returns empty outputs by design. A string_to_dict failure in the console means the config string wasn't valid JSON, so verify you're feeding it a config, not a file path. And remember the -1/unset defaults mean "don't touch this setting," so a "wrong" step count is usually the config's value, not a bug. Search "Ino Get Sampler Config" in the node menu (grouped under InoSamplerHelper).