Save 4 hours! We auto-setup your workflow! Free!

Drop your workflow.json — we handle every dependency, custom node, and model. Just open the link and run.

Auto-Setup Workflow Json (Free) Now!
ComfyUI > Nodes > ComfyUI-Nunchaku-Klein > Nunchaku FLUX.2 Klein LoRA Loader

ComfyUI Node: Nunchaku FLUX.2 Klein LoRA Loader

Class Name

NunchakuKleinLoraLoader

Category
loaders
Author
tom-m-2020 (Account age: 0days)
Extension
ComfyUI-Nunchaku-Klein
Latest Updated
2026-11-03
Github Stars
0.0K

How to Install ComfyUI-Nunchaku-Klein

Install this extension via the ComfyUI Manager by searching for ComfyUI-Nunchaku-Klein
  • 1. Click the Manager button in the main menu
  • 2. Select Custom Nodes Manager button
  • 3. Enter ComfyUI-Nunchaku-Klein in the search bar
After installation, click the Restart button to restart ComfyUI. Then, manually refresh your browser to clear the cache and access the updated list of nodes.

Visit ComfyUI Online for ready-to-use ComfyUI environment

  • Free trial available
  • 16GB VRAM to 80GB VRAM GPU machines
  • 400+ preloaded models/nodes
  • Freedom to upload custom models/nodes
  • 200+ ready-to-run workflows
  • 100% private workspace with up to 200GB storage
  • Dedicated Support

Run ComfyUI Online

Nunchaku FLUX.2 Klein LoRA Loader Description

Loads and validates a FLUX.2 Klein LoRA at adjustable strength onto a cloned model branch.

Nunchaku FLUX.2 Klein LoRA Loader:

Loads a compatible FLUX.2 Klein LoRA for use on a particular model branch. A LoRA is a set of additional weights that can influence an image model’s results without replacing the base model. This node checks that the LoRA contains usable weights and that their dimensions match the Klein model before adding it to the branch’s execution state. Because it clones the model rather than changing the incoming branch in place, you can build separate workflow paths with different LoRA choices or strengths.

Nunchaku FLUX.2 Klein LoRA Loader Input Parameters:

model

The MODEL to which the LoRA will be added. It must come from NunchakuKleinModelLoader; a different MODEL type will be rejected. If the incoming model branch already has Klein LoRAs assigned, this node preserves those assignments and adds the selected LoRA after them. There is no default, minimum, or maximum value: connect a compatible MODEL.

lora_name

Selects the LoRA file to load. The available options come from the filenames in your configured loras locations, so the exact list depends on the files in your installation. The node reads the selected file and checks its metadata, weights, and compatibility with the connected Klein model. No fixed default, minimum, or maximum is specified.

strength

Sets the strength stored for the selected LoRA. The default is 1.0; the control has a minimum of -10.0, a maximum of 10.0, and steps of 0.05. A value of 0.0 skips adding this LoRA, although the node still requires a compatible MODEL and returns a cloned branch. Nonzero values add the LoRA to that branch; the source does not define a particular visual effect for negative values, so judge those settings by your results. The value must be finite.

Nunchaku FLUX.2 Klein LoRA Loader Output Parameters:

model

A cloned MODEL branch carrying the LoRA assignments for subsequent execution. At a nonzero strength, it contains the incoming branch’s existing Klein LoRA assignments followed by the newly selected one. At 0.0, it retains only the inherited assignments. Connect this output to the next step of the workflow you want to use with those settings; other branches from the original input are not changed by this node.

Nunchaku FLUX.2 Klein LoRA Loader Usage Tips:

  • Connect a MODEL from NunchakuKleinModelLoader. A general MODEL from another loader is not interchangeable here.
  • Choose a FLUX.2 Klein LoRA compatible with the connected model. The loader checks weight pairs and model dimensions, so a file appearing in the menu does not by itself guarantee compatibility.
  • Start with the default strength of 1.0 and adjust it while comparing results. Use 0.0 when you want this node to pass along a branch without adding its selected LoRA.
  • Chain multiple instances when you want to add multiple LoRAs to one branch. Each instance preserves the assignments inherited from its input.
  • Branch from an earlier MODEL output when comparing different LoRA setups. Each branch receives its own assignments.
  • The node keeps one parsed LoRA in its local cache. Reusing the same file may avoid parsing it again, while switching among several files can require repeated loading.

Nunchaku FLUX.2 Klein LoRA Loader Common Errors and Solutions:

Nunchaku Klein LoRA loading requires a MODEL from NunchakuKleinModelLoader.

  • Explanation: The connected MODEL does not contain the Klein adapter required by this node.
  • Solution: Connect a MODEL produced by NunchakuKleinModelLoader, including when you set strength to 0.0.

Nunchaku Klein MODEL has invalid transformer options.

  • Explanation: The connected model does not have the transformer options structure this node needs to store branch-specific LoRA assignments.
  • Solution: Check that the MODEL comes from the intended Klein loader and has not been altered by an incompatible model-processing node.

LORA_SPEC_OPTION must contain a tuple of KleinLoraSpec.

  • Explanation: The model’s existing Klein LoRA assignments have an unexpected format. The actual message displays the internal option name in place of LORA_SPEC_OPTION.
  • Solution: Try the node directly after NunchakuKleinModelLoader, then reconnect other model-processing nodes one at a time to identify where the incompatible assignments appear.

LoRA strength must be finite, got <value>.

  • Explanation: The strength is not a valid finite number, such as when a dynamic value evaluates to NaN or infinity.
  • Solution: Enter an ordinary numeric strength within the control’s -10.0 to 10.0 range, or correct the value-generating node.

The installed Nunchaku FLUX.2 backend does not expose its LoRA conversion parser.

  • Explanation: The installed backend does not provide the parser this loader uses to recognize and validate Klein LoRA weights.
  • Solution: Check that your Nunchaku backend installation is compatible with this node and reinstall or update the matching components together.

LoRA <path> contains no tensors.

  • Explanation: The selected file loaded without any weights for the node to use.
  • Solution: Select a valid LoRA weight file or obtain a fresh copy of the intended file.

LoRA <path> contains non-tensor weights.

  • Explanation: At least one entry in the selected file is not a weight tensor in the format this loader expects.
  • Solution: Select a compatible FLUX.2 Klein LoRA file or obtain a correctly exported version.

LoRA <path> is not supported: <details>.

  • Explanation: The loader could read the file, but its weight structure could not be normalized into a supported Klein LoRA format. The details following the colon identify the specific problem.
  • Solution: Read those details, confirm that the file is intended for this Klein model and loader, and obtain a compatible export if necessary.

LoRA <path> has no weights recognized by the installed Nunchaku FLUX.2 Klein parser.

  • Explanation: The backend parser found no weights it can apply from the selected file.
  • Solution: Check that you selected a FLUX.2 Klein LoRA rather than a LoRA for another model family, and confirm that the backend version supports its format.

LoRA <path> produced incomplete Nunchaku weight pairs.

  • Explanation: The parsed weights do not form the complete pairs required by the loader.
  • Solution: Replace the file with a complete, correctly exported LoRA; simply changing strength will not repair missing weights.

LoRA <path> is missing the pair for <weight name>.

  • Explanation: One weight matrix has no matching partner.
  • Solution: Obtain an intact copy of the LoRA or re-export it with all required weights.

LoRA pair <name> must contain two matrices, got <shapes>.

  • Explanation: A pair of weights has the wrong number of dimensions for this loader.
  • Solution: Use a LoRA exported in a supported Klein format; this cannot be corrected with a node setting.

LoRA pair <name> has inconsistent rank: <shapes>.

  • Explanation: The two matrices in a weight pair do not agree on their shared size.
  • Solution: Replace or re-export the LoRA with a valid matching pair.

LoRA pair <name> is incompatible with Klein <profile>: got <shapes>, expected input/output dimensions <dimensions>.

  • Explanation: The weights target dimensions that differ from those of the connected Klein model.
  • Solution: Choose a LoRA made for the connected model’s architecture profile, or connect the matching Klein model instead.

Nunchaku FLUX.2 Klein LoRA Loader Related Nodes

Go back to the extension to check out more related nodes.
ComfyUI-Nunchaku-Klein
RunComfy
Copyright 2025 RunComfy. All Rights Reserved.

RunComfy is the premier ComfyUI platform, offering ComfyUI online environment and services, along with ComfyUI workflows featuring stunning visuals. RunComfy also provides AI Models, enabling artists to harness the latest AI tools to create incredible art.

Nunchaku FLUX.2 Klein LoRA Loader