ComfyUI Node
Fantastic H3 Store RefMod Encoder Frames
Add the frames H3's text encoder is shown to RefMods saved before they carried them, so the RefMod Text Encode reads them instead of decoding the RefMod. Each file is decoded once; its latent is not touched. Files that already have them and voice files are left alone. Queued by the RefMod library's Store encoder frames.
Fantastic H3 Store RefMod Encoder Frames
- vae
◄files►
Categoryconditioning/video_models
Inputs (2)
| Name | Type | Default | Description |
|---|---|---|---|
| files | STRING | RefMod files under models/refmods, one per line, e.g. characters/hero_visual. | |
| vae | VAE | MiniMax H3 video VAE. |
Outputs (0)
No outputs