DP Lora Strength Controller

Stacking several LoRAs means several strength sliders scattered across your graph - one per LoraLoader node, each a separate float widget you have to hunt down when you want to rebalance the mix. This node puts all five in one place: five strength dials, one output, meant to pair with the pack's multi-LoRA loader (DP Five Lora, per the README) so you can tune a five-LoRA stack from a single compact node instead of chasing widgets across the canvas.

How it works

There's not much mechanism to explain - it's five independent FLOAT widgets, each 0 to 3, bundled into a single custom output type (LORA_STRENGTHS) that a matching loader node reads. The 0-3 range is worth noting: LoRA strength usually lives around 0.6-1.0 for a "normal" application, but a range up to 3 leaves room for deliberately overdriving an underpowered or subtly-trained LoRA well past its intended weight, which is sometimes exactly what you want for a LoRA that's too weak at 1.0 to have any visible effect.

Because the output is a custom LORA_STRENGTHS type rather than five separate FLOAT wires, it's built specifically to plug into whatever DP loader node expects that bundled type - it isn't a generic strength control you can wire into an arbitrary LoraLoader, which typically wants a plain FLOAT per input instead.

Inputs and outputs

Five required inputs, all identical in shape:

• lora_01_strength through lora_05_strength - FLOAT, 0 to 3, step 0.01, default 1 each.

One output:

• strengths - a LORA_STRENGTHS bundle carrying all five values, meant to feed into a matching multi-LoRA loader node in the same pack.

How to install it

Search ComfyUI-Desert-Pixel-Nodes in ComfyUI Manager, or:

cd ComfyUI/custom_nodes
git clone https://github.com/DesertPixelAi/ComfyUI-Desert-Pixel-Nodes

Restart, find it under DP/utils. No models or extra dependencies - this node holds numbers, it doesn't load anything itself.

Common issues & troubleshooting

Nothing will connect to the strengths output. The LORA_STRENGTHS output is a custom type specific to this pack - it only plugs into a node built to accept it, which per the README's node list is DP Five Lora (or its random-strength sibling). A standard ComfyUI LoraLoader node has its own individual FLOAT strength inputs and won't accept this bundled type at all; if you're trying to wire this into a stock loader, that's the mismatch.

Changing a value on this node doesn't seem to affect a specific LoRA. Double-check the numbering lines up - lora_01_strength corresponds to whichever LoRA slot 1 is on the paired loader node, and it's easy to lose track of which index maps to which file once you've got five LoRAs loaded and only numbers to go by. Label your LoRA loader slots (or keep a note) if you're juggling more than two or three at once.

A LoRA has no visible effect even at strength 3. Cranking strength up to the node's ceiling doesn't fix a LoRA that's fundamentally mismatched to your base model or trained on a very different concept than what you're prompting for - if 1.0 does nothing, going to 3.0 usually amplifies noise rather than revealing a hidden effect. Confirm the LoRA is actually compatible with your checkpoint's architecture before assuming strength is the problem.

Five slots is more than you need. If you're only running two or three LoRAs, the unused slots at their default of 1 don't hurt anything as long as the paired loader node doesn't have a LoRA plugged into that slot - but if your workflow genuinely never needs five, a plain FLOAT primitive per LoRA loader might be less overhead than adopting this node's fixed five-slot shape.