sammykumar-ComfyUI-SwissArmyKnife/web/docs/SuperDualLoraLoader.md

SuperDualLoraLoader

Load and manage separate high-noise and low-noise LoRA stacks in a single node, including trigger extraction and template bundles.

Inputs

• high_noise_lora, low_noise_lora – WANVIDLORA stacks from WanVideo selectors.
• Optional clip tensor plus lora_bundle JSON coming from the UI template editor.

Outputs

• Updated high_noise_lora and low_noise_lora models.
• CLIP – forwarded/modified CLIP instance.
• TRIGGER_WORDS – comma-separated string of trigger phrases from enabled LoRAs.

Usage Tips

1. Configure bundles in the UI (templates panel) and send them to this node.
2. Feed the resulting WANVIDLORA tensors into Wan's sampler chain.

Additional Resources

• Full documentation