ComfyUI Node
Eric Krea2 Apply LoRA
A ComfyUI node in Eric/Krea2 with 15 inputs and 4 outputs.
Eric Krea2 Apply LoRA
- pipeline
- pipeline
- settings
- trigger_words
- prompt
◄lora_name▾►
◄strength1.00►
◄per_stage_weightsfalse►
◄weight_s11.00►
◄weight_s21.00►
◄weight_s31.00►
◄ephemeraltrue►
◄lora_path_override►
◄enabledtrue►
◄apply_lora_presetcustom►
◄fetch_triggerstrue►
◄add_triggersoff►
◄force_refetchfalse►
◄prompt►
CategoryEric/Krea2
Inputs (15)
| Name | Type | Default | Description |
|---|---|---|---|
| pipeline | KREA2_PIPELINE | — | |
| lora_name | COMBO | LoRA from ComfyUI/models/loras/Krea2/. 'none' = pass through unchanged. Use lora_path_override for a LoRA outside that folder. | |
| strength | FLOAT | 1.00-10–50000 | LoRA weight applied to ALL stages (unless per_stage_weights is on). This is the LoRA scale, unrelated to the Krea2 cfg/g convention. |
| per_stage_weightsopt | BOOLEAN | false | OFF: use 'strength' for every stage. ON: use the weight_s1/s2/s3 fields instead (e.g. detail LoRA 0/0.7/1.0, style LoRA 1.0/0.6/0.3). |
| weight_s1opt | FLOAT | 1.00-10–50000 | Stage-1 (composition) weight. Only used when per_stage_weights is ON. |
| weight_s2opt | FLOAT | 1.00-10–50000 | Stage-2 (refine) weight. Only used when per_stage_weights is ON. |
| weight_s3opt | FLOAT | 1.00-10–50000 | Stage-3 (final detail) weight. Only used when per_stage_weights is ON. |
| ephemeralopt | BOOLEAN | true | ON (recommended): load this LoRA fresh each run and unload it after generation (no stale/orphaned weights; small reload cost). OFF: leave it resident after the run (faster reuse, but you must Unload manually to free VRAM / before switching workflows). If ANY LoRA in the stack is ephemeral, the whole stack is cleared after the run. |
| lora_path_overrideopt | STRING | Optional absolute path to a LoRA file (overrides the dropdown). | |
| enabledopt | BOOLEAN | true | Quick on/off for this node without clearing its settings. OFF = pass the pipeline through untouched (the LoRA is not queued this run). |
| apply_lora_presetopt | COMBO | custom | Load a saved Apply-LoRA recipe (lora / strength / per-stage weights / toggles) into the panel for this run. 'custom' = use the panel as-is. Use the ★ Save Preset button to add one; the list refreshes on graph reload. |
| fetch_triggersopt | BOOLEAN | true | Look up this LoRA's trigger words (safetensors metadata / Civitai, cached locally in Eric_Krea2/data/lora_triggers.json - only the first lookup per file touches the network). Fills the trigger_words output. |
| add_triggersopt | COMBO | off | Merge the found trigger words into the prompt passthrough output. off = prompt passes through unchanged (use the trigger_words output manually). Triggers already present in the prompt are not duplicated. |
| force_refetchopt | BOOLEAN | false | Bypass the local trigger cache and re-query Civitai for this LoRA (use if triggers were updated on Civitai or a lookup failed). Leave OFF normally. |
| promptopt | STRING | Optional prompt passthrough. Connect your prompt here and take the 'prompt' output onward; with add_triggers on, trigger words are merged in. Without a connection the prompt output is just the triggers (or empty). |
Outputs (4)
| Name | Type | Description |
|---|---|---|
| pipeline | KREA2_PIPELINE | — |
| settings | STRING | — |
| trigger_words | STRING | — |
| prompt | STRING | — |