24 lines
582 B
Markdown
24 lines
582 B
Markdown
# Comfyui audio nodes
|
|
A collection of ComfyUI nodes related to audio.
|
|
|
|
Current nodes:
|
|
* Bark
|
|
* Load models
|
|
* Encode prompt
|
|
* Generate
|
|
1. Semantic
|
|
2. Coarse
|
|
3. Fine
|
|
* Load/save speaker npz, put speaker .npz files in `models/bark/speakers`
|
|
* Encodec
|
|
* Load models
|
|
* Encode audio
|
|
* Decode audio
|
|
|
|
# TODO
|
|
* Bark
|
|
* Hubert quantizer for creating npz from audio https://github.com/gitmylo/bark-voice-cloning-HuBERT-quantizer
|
|
* Coarse-ify codebooks for creating custom npz files
|
|
* Direct way to support long-form generations
|
|
* Other
|
|
* HuBert model loader |