ComfyUI-Translate

ComfyUI-Translate

[中文] | [English]

这是一个 ComfyUI 的本地翻译节点，支持 Google TranslateGemma-4B 模型。 It allows you to use the TranslateGemma-4b-it model directly within ComfyUI for text translation and image-text extraction.

✨ Features (功能)

• 🚀 Local Inference: Runs locally, no API key required. (本地运行，无需 API Key)
• 🖼️ Multimodal: Supports text-to-text and image-to-text translation. (支持文本翻译及图像文字提取翻译)
• ⚡ Auto Caching: Loads model once, fast inference for subsequent runs. (自动缓存模型，拒绝重复加载)
• 🛠️ Smart UI: Dropdown menu for common languages + Manual override support. (常用语言下拉菜单 + 支持手动输入代码)

📦 Installation (安装)

1. Clone the repository (克隆代码)

Go to your ComfyUI custom_nodes folder and run:

cd ComfyUI/custom_nodes/
git clone https://github.com/wanaigc/ComfyUI-Translate

2. Install Dependencies (安装依赖)

Important! You must install the required Python packages:

cd ComfyUI-Translate
pip install -r requirements.txt

(Requires transformers>=4.48.0, accelerate, sentencepiece)

📥 Model Download (下载模型)

Please download the model from HuggingFace or ModelScope (Recommended for CN users).

• HuggingFace: google/translategemma-4b-it
• ModelScope: google/translategemma-4b-it

Directory Structure (目录结构必须如下):

ComfyUI/
  models/
    Translate/
      translategemma-4b-it/
        ├── config.json
        ├── model.safetensors
        ├── tokenizer.json
        └── ... (other model files)

🛠️ Usage (使用说明)

1. Restart ComfyUI.
2. Double click on the canvas and search for: "Translate (Gemma 4B)".
3. Connect your text or image input.
4. Select source/target languages and run!

---

Developed by WanAIGC Team.