Instructions to use t8star/Ming3-TTS-Comfy with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use t8star/Ming3-TTS-Comfy with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf t8star/Ming3-TTS-Comfy:Q8_0 # Run inference directly in the terminal: llama cli -hf t8star/Ming3-TTS-Comfy:Q8_0
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf t8star/Ming3-TTS-Comfy:Q8_0 # Run inference directly in the terminal: llama cli -hf t8star/Ming3-TTS-Comfy:Q8_0
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf t8star/Ming3-TTS-Comfy:Q8_0 # Run inference directly in the terminal: ./build/bin/llama cli -hf t8star/Ming3-TTS-Comfy:Q8_0
Use Docker
docker model run hf.co/t8star/Ming3-TTS-Comfy:Q8_0
- Llama new
- LM Studio
- Jan
- Ollama
How to use t8star/Ming3-TTS-Comfy with Ollama:
ollama run hf.co/t8star/Ming3-TTS-Comfy:Q8_0
- Unsloth Desktop
- Docker Model Runner
How to use t8star/Ming3-TTS-Comfy with Docker Model Runner:
docker model run hf.co/t8star/Ming3-TTS-Comfy:Q8_0
- Lemonade
How to use t8star/Ming3-TTS-Comfy with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull t8star/Ming3-TTS-Comfy:Q8_0
Run and chat with the model
lemonade run user.Ming3-TTS-Comfy-Q8_0
List all available models
lemonade list
- Atomic Chat
Ming 3.0 TTS · ComfyUI / T8star
简体中文 | English
ComfyUI原生节点与本地工作台配套的模型与原生组件仓库。Ming文件保留上游INT8 ConvRot格式,未重新量化;识别使用Confucius4-R2T2,不使用Whisper。
文件
| 目录 | 内容 |
|---|---|
Ming-Comfy-INT8/checkpoints/ |
Ming INT8 ConvRot主权重,约9.07GB |
Ming-3.0-tts-tiny/ |
官方配置、tokenizer与CAMPPlus辅助文件 |
Confucius4-R2T2/models/ |
Q8识别模型、projector与FireRedVAD |
components/ |
Windows原生CUDA、Confucius与FFmpeg组件 |
licenses/ |
各组件原始许可与来源 |
manifest.json记录来源版本、大小与SHA256;SHA256SUMS便于独立校验。本仓库不包含完整原始BF16分片,也不包含Python运行时。
使用
安装GitHub节点后,使用ComfyUI自己的Python执行:
python install.py
python scripts/fetch_models.py
前者下载并校验原生组件,后者下载模型并配置路径;可通过 --destination 指定模型目录。已有模型可在节点配置中复用,不必重复下载。模型不通过远程推理服务运行。
当前预编译组件支持Windows x64、Python3.10/3.12、CUDA12.8。Ming内核覆盖sm80/86/89/90/100/120;Confucius CUDA库为sm120,其他GPU使用原生CPU路径。详细工作流、功能和硬件验证范围见GitHub。
许可与来源
Ming官方模型 · Comfy-Org INT8权重 · Confucius项目
不同组件分别适用原始许可,见 LICENSES.md。Confucius模型许可与Ming MIT许可分别保留。
社区与资源
- Downloads last month
- 16
8-bit
Model tree for t8star/Ming3-TTS-Comfy
Base model
inclusionAI/Ming-3.0-tts-tiny