- Add .github/workflows/publish_action.yml for Comfy registry - Update README: ComfyUI Manager install (recommended), Beta badge + warning - Minor docs tweaks
38 lines
1.1 KiB
Markdown
38 lines
1.1 KiB
Markdown
# IllustriousSmartPromptCache
|
|
|
|
A node for caching and reusing prompt conditioning in Illustrious workflows.
|
|
|
|
## What does it do?
|
|
|
|
- Stores prompt-to-conditioning mappings for fast reuse.
|
|
- Reduces recomputation when prompts are unchanged.
|
|
- Supports smart invalidation when upstream nodes change.
|
|
|
|
## Inputs
|
|
|
|
- **prompt**: The text prompt to cache.
|
|
- **conditioning**: The computed conditioning to store.
|
|
- **mode**: Cache mode (auto, force, manual).
|
|
- **key**: (Optional) Custom cache key.
|
|
- **log**: Enable for debug info.
|
|
|
|
## Outputs
|
|
|
|
- Cached conditioning for use in samplers or other nodes.
|
|
|
|
## Artist Tips
|
|
|
|
- Use in large workflows to speed up repeated prompt runs.
|
|
- Auto mode is best for most cases; force for guaranteed refresh.
|
|
- Manual mode lets you control cache keys for advanced setups.
|
|
- Combine with Smart Scene Generator for complex prompt pipelines.
|
|
|
|
## Example Workflow
|
|
|
|
1. Connect your prompt and conditioning.
|
|
2. Choose cache mode.
|
|
3. Use the cached output in your sampler or prompt node.
|
|
|
|
---
|
|
For more on prompt caching, see the Smart Scene and workflow optimization guides.
|