nodes/Enemyx-net-VibeVoice-ComfyUI
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
Files
main
Enemyx-net-VibeVoice-ComfyUI/examples/Pause-Tag.json
T
Fabio Sarracino 2d29a48624 Bump version to 1.8.0 
- **New Official 8-bit Quantized Model**: VibeVoice-Large-Q8
  - Released on HuggingFace: [FabioSarracino/VibeVoice-Large-Q8](https://huggingface.co/FabioSarracino/VibeVoice-Large-Q8)
  - Model size: 11.6GB (38% reduction from 18.7GB full precision)
  - VRAM usage: ~12GB (40% reduction from ~20GB)
  - **Perfect audio quality**: Identical to full precision model - no quality degradation
  - **Selective quantization approach**: audio-critical components (diffusion head, VAE, connectors) kept at full precision
  - Optimized for 12GB VRAM GPUs (RTX 3060, 4070 Ti, etc.)
  - Solves the common 8-bit "noise problem" by carefully selecting which components to quantize
- **Added 8-bit Dynamic LLM Quantization**
  - New "8bit" option in `quantize_llm` parameter for both Single and Multiple Speaker nodes
  - Options now: "full precision" (default), "4bit", "8bit"
  - Dynamically quantizes only the LLM component for non-quantized models
  - Skips all audio-critical components (diffusion_head, acoustic/semantic connectors, tokenizers)
  - Provides good balance between quality and VRAM savings
  - Requires CUDA GPU and bitsandbytes library
  - Automatically ignored for pre-quantized models
2025-10-01 04:31:39 +02:00

1 line
7.2 KiB
JSON
Raw Permalink Blame History

This file contains invisible Unicode characters
This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
{"id":"b70cf6f7-8531-4faa-9843-9c963a4ba577","revision":0,"last_node_id":47,"last_link_id":58,"nodes":[{"id":28,"type":"LoadTextFromFileNode","pos":[-51.13530731201172,497.1748352050781],"size":[289.5152282714844,58],"flags":{},"order":0,"mode":4,"inputs":[{"localized_name":"file","name":"file","type":"COMBO","widget":{"name":"file"},"link":null}],"outputs":[{"localized_name":"text","name":"text","type":"STRING","links":null}],"properties":{"Node name for S&R":"LoadTextFromFileNode","cnr_id":"VibeVoice-ComfyUI","ver":"5a24489a7b0bf0c406d291dd51e82a085d338d44"},"widgets_values":["No text files found in any directory"],"color":"#323","bgcolor":"#535"},{"id":38,"type":"Note","pos":[775.2548828125,307.8158874511719],"size":[415,88],"flags":{},"order":1,"mode":0,"inputs":[],"outputs":[],"title":"Pause System","properties":{},"widgets_values":["[pause]: add 1 second of silence.\n[pause:{number}] add {number}ms of pause\nWARNING: the pause tag forces the text to be split into chunks. This may worsen the model’s ability to understand the context. The model’s context is represented ONLY by its own chunk."],"color":"#432","bgcolor":"#653"},{"id":15,"type":"LoadAudio","pos":[-52.503074645996094,163.9591064453125],"size":[270,136],"flags":{},"order":2,"mode":0,"inputs":[{"localized_name":"audio","name":"audio","type":"COMBO","widget":{"name":"audio"},"link":null},{"localized_name":"audioUI","name":"audioUI","type":"AUDIO_UI","widget":{"name":"audioUI"},"link":null},{"localized_name":"upload","name":"upload","type":"AUDIOUPLOAD","widget":{"name":"upload"},"link":null}],"outputs":[{"localized_name":"AUDIO","name":"AUDIO","type":"AUDIO","links":[57]}],"properties":{"cnr_id":"comfy-core","ver":"0.3.49","Node name for S&R":"LoadAudio"},"widgets_values":["Voice.mp3",null,null],"color":"#2a363b","bgcolor":"#3f5159"},{"id":21,"type":"Note","pos":[-119.67156219482422,637.6148071289062],"size":[415,88],"flags":{},"order":3,"mode":0,"inputs":[],"outputs":[],"title":"Load Text From File","properties":{},"widgets_values":["Use Load Text From File if you want to use a .txt file instead of text-area. You can load .txt files from ComfyUI/input, ComfyUI/output or ComfyUI/temp directories."],"color":"#432","bgcolor":"#653"},{"id":16,"type":"PreviewAudio","pos":[845.1698608398438,163.10276794433594],"size":[270,88],"flags":{},"order":8,"mode":0,"inputs":[{"localized_name":"audio","name":"audio","type":"AUDIO","link":58},{"localized_name":"audioUI","name":"audioUI","type":"AUDIO_UI","widget":{"name":"audioUI"},"link":null}],"outputs":[],"properties":{"cnr_id":"comfy-core","ver":"0.3.49","Node name for S&R":"PreviewAudio"},"widgets_values":[],"color":"#323","bgcolor":"#535"},{"id":40,"type":"Note","pos":[325.02294921875,636.903564453125],"size":[415,88],"flags":{},"order":4,"mode":0,"inputs":[],"outputs":[],"title":"Voice Speed Factor","properties":{},"widgets_values":["The voice speed factor influences the original source audio to attempt to achieve a slower or faster final speech. 1.0 is the normal speed. It is recommended not to exceed values ​​between 0.95 and 1.05. The effect is best when you provide a sample audio of at least 20 seconds."],"color":"#432","bgcolor":"#653"},{"id":45,"type":"VibeVoiceSingleSpeakerNode","pos":[327.48126220703125,164.61436462402344],"size":[400,420],"flags":{},"order":7,"mode":0,"inputs":[{"localized_name":"voice_to_clone","name":"voice_to_clone","shape":7,"type":"AUDIO","link":57},{"localized_name":"lora","name":"lora","shape":7,"type":"LORA_CONFIG","link":null},{"localized_name":"text","name":"text","type":"STRING","widget":{"name":"text"},"link":null},{"localized_name":"model","name":"model","type":"COMBO","widget":{"name":"model"},"link":null},{"localized_name":"attention_type","name":"attention_type","type":"COMBO","widget":{"name":"attention_type"},"link":null},{"localized_name":"quantize_llm","name":"quantize_llm","type":"COMBO","widget":{"name":"quantize_llm"},"link":null},{"localized_name":"free_memory_after_generate","name":"free_memory_after_generate","type":"BOOLEAN","widget":{"name":"free_memory_after_generate"},"link":null},{"localized_name":"diffusion_steps","name":"diffusion_steps","type":"INT","widget":{"name":"diffusion_steps"},"link":null},{"localized_name":"seed","name":"seed","type":"INT","widget":{"name":"seed"},"link":null},{"localized_name":"cfg_scale","name":"cfg_scale","type":"FLOAT","widget":{"name":"cfg_scale"},"link":null},{"localized_name":"use_sampling","name":"use_sampling","type":"BOOLEAN","widget":{"name":"use_sampling"},"link":null},{"localized_name":"temperature","name":"temperature","shape":7,"type":"FLOAT","widget":{"name":"temperature"},"link":null},{"localized_name":"top_p","name":"top_p","shape":7,"type":"FLOAT","widget":{"name":"top_p"},"link":null},{"localized_name":"max_words_per_chunk","name":"max_words_per_chunk","shape":7,"type":"INT","widget":{"name":"max_words_per_chunk"},"link":null},{"localized_name":"voice_speed_factor","name":"voice_speed_factor","shape":7,"type":"FLOAT","widget":{"name":"voice_speed_factor"},"link":null}],"outputs":[{"localized_name":"audio","name":"audio","type":"AUDIO","links":[58]}],"properties":{"Node name for S&R":"VibeVoiceSingleSpeakerNode"},"widgets_values":["Hello, this is a test of the VibeVoice text-to-speech system. [pause] Do you like my voice? [pause:500] What's your name?","VibeVoice-1.5B","auto","full precision",true,20,42,"fixed",1.3,false,0.95,0.95,250,1],"color":"#223","bgcolor":"#335"},{"id":46,"type":"Note","pos":[-576.477294921875,222.4726104736328],"size":[408.66363525390625,236.39089965820312],"flags":{},"order":5,"mode":0,"inputs":[],"outputs":[],"title":"1) Download Models","properties":{},"widgets_values":["You have to manually download the models you would like to use and put them into: ComfyUI/models/vibevoice/\n\nMake a directory for each model and put all the files inside them.\n\nVibeVoice-1.5B model (~ 5.4 GB):\nhttps://huggingface.co/microsoft/VibeVoice-1.5B/tree/main\n\nVibeVoice-Large model (~ 18.7 GB):\nhttps://huggingface.co/aoi-ot/VibeVoice-Large/tree/main\n\nVibeVoice-Large-Q-8bit model (~ 11.6 GB):\nhttps://huggingface.co/FabioSarracino/VibeVoice-Large-Q8/tree/main\n\nVibeVoice-Large-Q-4bit model (~ 6.6 GB):\nhttps://huggingface.co/DevParker/VibeVoice7b-low-vram/tree/main/4bit"],"color":"#432","bgcolor":"#653"},{"id":47,"type":"Note","pos":[-576.177978515625,522.212646484375],"size":[407.2561950683594,155.19009399414062],"flags":{},"order":6,"mode":0,"inputs":[],"outputs":[],"title":"2) Download Tokenizer","properties":{},"widgets_values":["You have to manually download the Qwen2.5 Tokenizer files and put them into: ComfyUI/models/vibevoice/tokenizer/\n\nhttps://huggingface.co/Qwen/Qwen2.5-1.5B/tree/main\n\nRequired files: tokenizer_config.json, vocab.json, merges.txt, tokenizer.json (~11MB)\n\nPut the files directly inside tokenizer directory without make another directory inside."],"color":"#432","bgcolor":"#653"}],"links":[[57,15,0,45,0,"AUDIO"],[58,45,0,16,0,"AUDIO"]],"groups":[{"id":2,"title":"Instructions before use:","bounding":[-599.3793334960938,124.93412017822266,453.3775939941406,595.2697143554688],"color":"#3f789e","font_size":24,"flags":{}}],"config":{},"extra":{"ds":{"scale":0.8264462809917354,"offset":[815.9689977237014,-22.084207406969263]}},"version":0.4}
Reference in New Issue View Git Blame Copy Permalink