72 lines
2.1 KiB
Markdown
72 lines
2.1 KiB
Markdown
# ComfyUI_RH_mammothmoda
|
||
|
||
ComfyUI 自定义节点，用于 MammothModa2 文本生成图像。
|
||
|
||
## 功能特点
|
||
|
||
- **文本生成图像**：使用 MammothModa2 进行高质量图像合成
|
||
- **性能优化**：支持 INT8 量化实现高效推理
|
||
- **Flash Attention**：使用 flash_attention_2 加速
|
||
|
||
## 安装方法
|
||
|
||
### 通过 ComfyUI Manager 安装（推荐）
|
||
在 ComfyUI Manager 中搜索 "RunningHub Mammothmoda" 并安装。
|
||
|
||
### 手动安装
|
||
```bash
|
||
cd ComfyUI/custom_nodes
|
||
git clone https://github.com/HM-RunningHub/ComfyUI_RH_mammothmoda.git
|
||
cd ComfyUI_RH_mammothmoda
|
||
pip install -r requirements.txt
|
||
```
|
||
|
||
## 模型下载
|
||
|
||
从 [🤗 HuggingFace](https://huggingface.co/bytedance-research/MammothModa2-Preview) 下载 MammothModa2-Preview 模型并放置到：
|
||
```
|
||
ComfyUI/models/MammothModa2-Preview/
|
||
```
|
||
|
||
## 使用说明
|
||
|
||
1. **加载模型**：使用 "RunningHub Mammothmoda Loader" 节点
|
||
2. **生成图像**：连接到 "RunningHub Mammothmoda T2I Sampler" 节点
|
||
3. **配置参数**：设置提示词、尺寸（宽度/高度）、步数和引导比例
|
||
4. **运行**：执行工作流生成图像
|
||
|
||
## 节点说明
|
||
|
||
- **RunningHub Mammothmoda Loader**：加载 MammothModa2 模型
|
||
- **RunningHub Mammothmoda T2I Sampler**：从文本提示词生成图像
|
||
|
||
## 系统要求
|
||
|
||
- 支持 CUDA 的 GPU（推荐 24GB 显存）
|
||
- PyTorch（支持 CUDA）
|
||
- flash-attn 包
|
||
|
||
## 致谢
|
||
|
||
本项目基于字节跳动研究院的 [MammothModa2](https://huggingface.co/bytedance-research/MammothModa2-Preview)。感谢 MammothModa 团队的出色工作。
|
||
|
||
- **原项目**: [MammothModa2](https://github.com/bytedance/mammothmoda)
|
||
- **模型**: [MammothModa2-Preview](https://huggingface.co/bytedance-research/MammothModa2-Preview)
|
||
- **许可证**: [Apache-2.0](https://opensource.org/licenses/Apache-2.0)
|
||
|
||
## 引用
|
||
|
||
```bibtex
|
||
@misc{mammothmoda2025,
|
||
title = {MammothModa2: Jointly Optimized Autoregressive-Diffusion Models for Unified Multimodal Understanding and Generation},
|
||
author = {MammothModa Team},
|
||
year = {2025},
|
||
url = {https://github.com/bytedance/mammothmoda}
|
||
}
|
||
```
|
||
|
||
## 许可证
|
||
|
||
Apache-2.0
|
||
|