---
title: HeartMuLa-oss_ComfyUI
canonical_url: "https://www.modelscope.cn/models/AIGCCrazyMa/HeartMuLa-oss_ComfyUI"
md_url: "https://www.modelscope.cn/models/AIGCCrazyMa/HeartMuLa-oss_ComfyUI.md"
repository: AIGCCrazyMa/HeartMuLa-oss_ComfyUI
chinese_name: HeartMuLa-oss_ComfyUI
last_updated: 2026-01-22
parameters: 5.6B
tensor_type:
  - F32
library_name:
  - safetensors
  - pytorch
frameworks:
  - Pytorch
downloads: 398
stars: 0
---

# HeartMuLa-oss_ComfyUI

> HeartMuLa-oss_ComfyUI - AIGCCrazyMa 在 ModelScope 开源的模型。HeartMuLa 音乐生成系统的合并权重文件，专为 ComfyUI 优化

- **Repository**: AIGCCrazyMa/HeartMuLa-oss_ComfyUI
- **Parameters**: 5.6B
- **Downloads**: 398
- **Stars**: 0
- **Last updated**: 2026-01-22

Source: https://www.modelscope.cn/models/AIGCCrazyMa/HeartMuLa-oss_ComfyUI

---

# HeartMuLa 合并模型说明 / Merged HeartMuLa Models

[中文](#中文指南) | [English](#english-guide)

<a id="中文指南"></a>

## 📦 模型来源

本目录包含 HeartMuLa 音乐生成系统的合并权重文件，专为 ComfyUI 优化。

这些模型源自官方 [HeartMuLa](https://github.com/HeartMuLa/heartlib) 开源项目：

*   **HeartMuLa-oss-3B**: 音乐语言模型，可根据歌词和标签生成音乐。
*   **HeartCodec-oss**: 12.5 Hz 的高保真音乐编解码器。

原始权重获取自 [HuggingFace](https://huggingface.co/HeartMuLa/HeartMuLa-oss-3B) / [ModelScope](https://modelscope.cn/models/HeartMuLa/HeartMuLa-oss-3B)。

## 🚀 在 ComfyUI 中使用

这些合并后的模型是专门为 **ComfyUI-HeartMuLa** 自定义节点准备的。

### 1. 安装插件

请从以下仓库安装 ComfyUI 自定义节点：

👉 **[https://github.com/crazyma99/ComfyUI-HeartMuLa](https://github.com/crazyma99/ComfyUI-HeartMuLa)**

请按照该仓库 README 中的说明进行环境安装。

### 2. 模型放置

请将本目录下的 `.safetensors` 文件放入您的 ComfyUI Checkpoints 文件夹中：

*   将 `HeartMuLa-oss-3B-merged.safetensors` 复制到 `ComfyUI/models/checkpoints/`
*   将 `HeartCodec-oss-merged.safetensors` 复制到 `ComfyUI/models/checkpoints/`

### 3. 使用流程

1.  启动 ComfyUI。
2.  使用 **HeartMuLa Loader** 节点加载这些模型文件。
3.  连接到 **HeartMuLa Generator** 节点开始创作音乐！

## ⚖️ 许可协议

原始 HeartMuLa 模型和代码遵循 **Apache 2.0** 协议。
详情请参阅 [LICENSE](./LICENSE) 文件。

---

<a id="english-guide"></a>

## 📦 Model Source

This directory contains the merged weights for the HeartMuLa music generation system, optimized for use with ComfyUI.

The models included here are derived from the official [HeartMuLa](https://github.com/HeartMuLa/heartlib) open-source project:

*   **HeartMuLa-oss-3B**: The music language model that generates music conditioned on lyrics and tags.
*   **HeartCodec-oss**: The 12.5 Hz music codec with high reconstruction fidelity.

Original weights were obtained from [HuggingFace](https://huggingface.co/HeartMuLa/HeartMuLa-oss-3B) / [ModelScope](https://modelscope.cn/models/HeartMuLa/HeartMuLa-oss-3B).

## 🚀 Usage in ComfyUI

These merged models are specifically prepared for the **ComfyUI-HeartMuLa** custom node.

### 1. Installation

Please install the ComfyUI custom node from the following repository:

👉 **[https://github.com/crazyma99/ComfyUI-HeartMuLa](https://github.com/crazyma99/ComfyUI-HeartMuLa)**

Follow the installation instructions in the repository's README to set up the environment.

### 2. Model Placement

Place the `.safetensors` files from this directory into your ComfyUI checkpoints folder:

*   Copy `HeartMuLa-oss-3B-merged.safetensors` to `ComfyUI/models/checkpoints/`
*   Copy `HeartCodec-oss-merged.safetensors` to `ComfyUI/models/checkpoints/`

### 3. Workflow

1.  Launch ComfyUI.
2.  Use the **HeartMuLa Loader** node to load these checkpoints.
3.  Connect to **HeartMuLa Generator** to start creating music!

## ⚖️ License

The original HeartMuLa models and code are licensed under **Apache 2.0**.
See the [LICENSE](./LICENSE) file for details.