snicolast/ComfyUI-IndexTTS2
1
0
Fork 0
mirror of https://github.com/snicolast/ComfyUI-IndexTTS2.git synced 2026-05-01 12:11:22 +00:00
Code Issues Packages Projects Releases Wiki Activity

readme

Browse Source
This commit is contained in:
snicolast
2025-09-22 13:32:34 +12:00
parent 6ec11c4a06
commit 1af25d1b5f
1 changed files with 3 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
README.md
View File

@@ -64,7 +64,7 @@ Models (checkpoints)
**Important**: The updated code now uses local model files by default for offline usage and faster loading. **Important**: The updated code now uses local model files by default for offline usage and faster loading.
Nodes ## Nodes
- IndexTTS2 Simple - IndexTTS2 Simple
- Inputs: audio (speaker), text, emotion_control_weight (0.0-1.0), emotion_audio (optional), emotion_vector (optional) - Inputs: audio (speaker), text, emotion_control_weight (0.0-1.0), emotion_audio (optional), emotion_vector (optional)
- Outputs: AUDIO (for Preview/Save), STRING (emotion source message) - Outputs: AUDIO (for Preview/Save), STRING (emotion source message)
@@ -86,7 +86,7 @@ Nodes
- Requires: modelscope and local QwenEmotion at checkpoints/qwen0.6bemo4-merge/ - Requires: modelscope and local QwenEmotion at checkpoints/qwen0.6bemo4-merge/
- Outputs: EMOTION_VECTOR, STRING summary - Outputs: EMOTION_VECTOR, STRING summary
Examples ## Examples
- Basic: Load Audio -> IndexTTS2 Simple -> Preview/Save Audio - Basic: Load Audio -> IndexTTS2 Simple -> Preview/Save Audio
- Second audio emotion: Load Audio (speaker) + Load Audio (emotion) -> IndexTTS2 Simple -> Save - Second audio emotion: Load Audio (speaker) + Load Audio (emotion) -> IndexTTS2 Simple -> Save
- Vector emotion: IndexTTS2 Emotion Vector -> IndexTTS2 Simple -> Save - Vector emotion: IndexTTS2 Emotion Vector -> IndexTTS2 Simple -> Save
@@ -94,7 +94,7 @@ Examples
![ComfyUI-IndexTTS2 nodes](images/overview.png) ![ComfyUI-IndexTTS2 nodes](images/overview.png)
Troubleshooting ## Troubleshooting
- Tested only in Windows. DeepSpeed disabled. - Tested only in Windows. DeepSpeed disabled.
- Emotion vector sum exceeds maximum 1.5: lower one or more sliders or adjust the text-derived vector. - Emotion vector sum exceeds maximum 1.5: lower one or more sliders or adjust the text-derived vector.
- BigVGAN kernel message: custom CUDA kernel is disabled by default; falls back to PyTorch ops. - BigVGAN kernel message: custom CUDA kernel is disabled by default; falls back to PyTorch ops.