diff --git a/README.md b/README.md index 2c39afe..59d452b 100644 --- a/README.md +++ b/README.md @@ -6,7 +6,7 @@ [English](README_en.md) | 中文 -MTran + 一个超低资源消耗超快的离线翻译服务器，无需显卡。单个请求平均响应时间 50 毫秒。支持全世界主要语言的翻译。 @@ -14,10 +14,6 @@ > 注意本模型专注于速度和多种设备私有部署，所以翻译质量肯定是不如大模型翻译的效果。需要高质量的翻译建议使用在线大模型 API。 -## Demo - -> 即将上线 - ## 同类项目效果(CPU,英译中) @@ -32,117 +28,11 @@ > 表中数据仅供参考，非严格测试，非量化版本对比。 -## 更新日志 - -[更新日志](./CHANGELOG.md) - -## 桌面端 - -即将发布桌面端软件，敬请期待。 - -## 服务器部署 - -> 对普通用户来说有难度，建议使用桌面端。 - -### 1.1 环境要求 - -- Docker -- Docker Compose（可选） - -### 1.2 镜像版本 - -- 默认无内置模型翻译时自动下载：`xxnuo/mtranserver:latest` -- 内置中英互译模型：`xxnuo/mtranserver:latest-zh` -- 内置日英互译模型：`xxnuo/mtranserver:latest-ja` - -### 1.3 Docker 部署 - -复制下面的命令，在终端执行。 - -```bash -docker run -d --name mtranserver -p 8989:8989 -e API_TOKEN=your_token xxnuo/mtranserver:latest -``` - -### 1.4 Docker Compose 部署 - -服务器准备一个存放配置的文件夹，打开终端执行以下命令 - -```bash -mkdir mtranserver -cd mtranserver -touch compose.yml -``` - -用编辑器打开 `compose.yml` 文件，写入以下内容。 - -> 1. 修改下面的 `your_token` 为你自己设置的一个密码，使用英文大小写和数字。自己内网可以不设置，如果是`云服务器`强烈建议设置一个密码，保护服务以免被`扫到、攻击、滥用`。 -> -> 2. 如果需要更改端口，修改 `ports` 的值，比如修改为 `9999:8989` 表示将服务端口映射到本机 9999 端口。 -> -> 3. 建议使用 `API_TOKEN` 设置密码，旧版本的 `CORE_API_TOKEN` 也还可以生效。 - -```yaml -services: - mtranserver: - image: xxnuo/mtranserver:latest - container_name: mtranserver - restart: unless-stopped - ports: - - "8989:8989" - environment: - - API_TOKEN=your_token -``` - -先启动测试，确保 8989 端口没被占用。 - -```bash -docker compose up -``` - -正常输出示例： - -```bash -HTTP Service URL: http://0.0.0.0:8989 -Swagger UI: http://0.0.0.0:8989/docs -``` - -然后按 `Ctrl+C` 停止服务运行，然后正式启动服务器 - -```bash -docker compose up -d -``` - -这时候服务器就在后台运行了。 - -## 准备模型 - -⚠️ 注意：第一次请求翻译 API 时会在后台自动下载模型，无需手动下载。 - -模型自动下载功能需要连接网络（中国大陆不需要代理），**后续翻译及其他功能均无需联网完全离线**。 - -**所以第一次翻译不是秒回，要等待一会儿！** - -可在 Docker 日志处观察进度。下载速度取决于网络速度，一般在 10s 内能完成一个语言模型的下载。如果下载超时/失败，请检查容器是否能正常联网。 - -如果属于内网机器无法访问互联网可按照下文指导手动下载模型。 - -### API 使用 - -下面表格内的 `localhost` 可以替换为你的服务器地址或 Docker 容器名。 - -下面表格内的 `8989` 端口可以替换为你在 `compose.yml` 文件中设置的端口值。 - -如果未设置环境变量 `API_TOKEN`、`CORE_API_TOKEN` 或者设置为空，翻译插件使用`无密码`的 API。 - -如果设置了 `API_TOKEN`、`CORE_API_TOKEN`，翻译插件使用`有密码`的 API。 - -下面表格中的 `your_token` 替换为你在环境变量中设置的 `API_TOKEN` 值。 - #### 翻译插件接口 > 注： > -> - [沉浸式翻译](https://immersivetranslate.com/zh-Hans/docs/services/custom/) 在`设置`页面，开发者模式中启用`Beta`特性，即可在`翻译服务`中看到`自定义 API 设置`([官方图文教程](https://immersivetranslate.com/zh-Hans/docs/services/custom/))。然后将`自定义 API 设置`的`每秒最大请求数`拉高以充分发挥服务器性能准备体验飞一般的感觉。我设置的是`每秒最大请求数`为`5000`，`每次请求最大段落数`为`10`。你可以根据自己服务器配置设置。 +> - [沉浸式翻译](https://immersivetranslate.com/zh-Hans/docs/services/custom/) 在`设置`页面，开发者模式中启用`Beta`特性，即可在`翻译服务`中看到`自定义 API 设置`([官方图文教程](https://immersivetranslate.com/zh-Hans/docs/services/custom/))。然后将`自定义 API 设置`的`每秒最大请求数`拉高以充分发挥服务器性能准备体验飞一般的感觉。我设置的是`每秒最大请求数`为`5000`，`每次请求最大段落数`为`1`。你可以根据自己服务器配置设置。 > > - [简约翻译](https://github.com/fishjar/kiss-translator) 在`设置`页面，接口设置中滚动到下面，即可看到自定义接口 `Custom`。同理，设置`最大请求并发数量`、`每次请求间隔时间`以充分发挥服务器性能。我设置的是`最大请求并发数量`为`100`，`每次请求间隔时间`为`1`。你可以根据自己服务器配置设置。 > @@ -157,49 +47,10 @@ docker compose up -d **普通用户参照表格内容设置好插件使用的接口地址就可以使用了。** -### 保持更新 - -```bash -cd mtranserver -docker pull xxnuo/mtranserver:latest -docker compose up -d -``` - -### 开发者接口 - -> Base URL: `http://localhost:8989` -> -> 在 v3.0.0 及以上版本中，API 文档可通过访问 `http://localhost:8989/docs` 查看完整 Swagger 文档 - -| 名称 | URL | 请求格式 | 返回格式 | 认证头 | -| ------------------ | ------------------------ | -------------------------------------------------------------------------------------- | ------------------------------------------------------------------ | ------------------------- | -| 服务版本 | `/version` | 无 | `{"version": "v3.0.0"}` | 无 | -| 支持的语言列表 | `/languages` | 无 | `{"languages":["en","zh","ja","ko"...]}` | Authorization: your_token | -| 普通翻译接口 | `/translate` | `{"from": "en", "to": "zh", "text": "Hello, world!"}` | `{"result": "你好，世界！"}` | Authorization: your_token | -| 批量翻译接口 | `/translate/batch` | `{"from": "en", "to": "zh", "texts": ["Hello, world!", "Hello, world!"]}` | `{"results": ["你好，世界！", "你好，世界！"]}` | Authorization: your_token | -| 健康检查 | `/health` | 无 | `{"status": "ok"}` | 无 | -| 心跳检查 | `/__heartbeat__` | 无 | `Ready` | 无 | -| 负载均衡心跳检查 | `/__lbheartbeat__` | 无 | `Ready` | 无 | -| 谷歌翻译兼容接口 1 | `/language/translate/v2` | `{"q": "The Great Pyramid of Giza", "source": "en", "target": "zh", "format": "text"}` | `{"data": {"translations": [{"translatedText": "吉萨大金字塔"}]}}` | Authorization: your_token | - -> 开发者高级设置请参考 [CONFIG.md](./CONFIG.md) - -## 贡献者 - - - - - - - - -
Lv Meiyang
Lv Meiyang		Leo
Leo
- ## Star History [![Star History Chart](https://api.star-history.com/svg?repos=xxnuo/MTranServer&type=Timeline)](https://www.star-history.com/#xxnuo/MTranServer&Timeline) - ## Thanks [Bergamot Project](https://browser.mt/) for awesome idea of local translation. diff --git a/README_en.md b/README_en.md index a33f3be..d25ac12 100644 --- a/README_en.md +++ b/README_en.md @@ -1,6 +1,6 @@ # MTranServer -MTran + [中文](README.md) | English @@ -12,10 +12,6 @@ Note: This model focuses on speed and private deployment on various devices, so For high-quality translation, consider using online large language model APIs. -## Demo - -> Coming soon - ## Comparison with Similar Projects (CPU, English to Chinese) @@ -32,98 +28,6 @@ For high-quality translation, consider using online large language model APIs. > > Table data is for reference only, not strict testing, non-quantized version comparison. -## Update Log - -[Update Log](./CHANGELOG.md) - -## Desktop Client - -Desktop client software coming soon, stay tuned. - -## Server Deployment - -> This may be challenging for regular users, consider using the desktop client when available. - -### 1.1 Requirements - -- Docker -- Docker Compose (optional) - -### 1.2 Image Versions - -- Default version (download models automatically when no built-in models are available): `xxnuo/mtranserver:latest` -- Built-in Chinese <-> English translation models: `xxnuo/mtranserver:latest-zh` -- Built-in Japanese <-> English translation models: `xxnuo/mtranserver:latest-ja` - -### 1.3 Docker Deployment - -Copy the command below and execute it in your terminal. - -```bash -docker run -d --name mtranserver -p 8989:8989 -e API_TOKEN=your_token xxnuo/mtranserver:latest -``` - -### 1.4 Docker Compose Deployment - -Prepare a folder for configuration files on your server and run the following commands in terminal: - -```bash -mkdir mtranserver -cd mtranserver -touch compose.yml -``` - -Open `compose.yml` with an editor and add the following content: - -> 1. Change `your_token` below to your own password using English letters and numbers. For internal network use, setting a password is optional, but for `cloud servers`, it is strongly recommended to set a password to protect against `scanning, attacks, and abuse`. -> -> 2. To change the port, modify the `ports` value. For example, change to `9999:8989` to map the service port to local port 9999. - -```yaml -services: - mtranserver: - image: xxnuo/mtranserver:latest - container_name: mtranserver - restart: unless-stopped - ports: - - "8989:8989" - environment: - - CORE_API_TOKEN=your_token -``` - -First, test the service to ensure the port isn't occupied: - -```bash -docker compose up -``` - -Example normal output: - -```bash -HTTP Service URL: http://0.0.0.0:8989 -Swagger UI: http://0.0.0.0:8989/docs -``` - -Then press `Ctrl+C` to stop the service, and start it officially: - -```bash -docker compose up -d -``` - -The server will now run in the background. - -## Preparing Models - -⚠️ Note: Models will be automatically downloaded in the background when you first request the translation API, no manual download needed. - -The automatic model download feature requires internet connection (no proxy needed in mainland China), **all subsequent translations and other functions work completely offline without internet**. - -**So the first translation won't be instant, you'll need to wait a moment!** - -You can monitor the progress in the Docker logs. Download speed depends on your network speed, typically completing a language model download within 10 seconds. If the download times out or fails, check if your container has normal internet access. - -If your machine is on an internal network without internet access, you can follow the instructions below to manually download models. - ### API Usage In the following tables, `localhost` can be replaced with your server address or Docker container name. @@ -140,7 +44,7 @@ Replace `your_token` in the following tables with your `API_TOKEN` or `CORE_API_ > Note: > -> - [Immersive Translation](https://immersivetranslate.com/docs/services/custom/) - Enable `Beta` features in developer mode in `Settings` to see `Custom API Settings` under `Translation Services` ([official tutorial with images](https://immersivetranslate.com/docs/services/custom/)). Then increase the `Maximum Requests per Second` in `Custom API Settings` to fully utilize server performance. I set `Maximum Requests per Second` to `5000` and `Maximum Paragraphs per Request` to `10`. You can adjust based on your server hardware. +> - [Immersive Translation](https://immersivetranslate.com/docs/services/custom/) - Enable `Beta` features in developer mode in `Settings` to see `Custom API Settings` under `Translation Services` ([official tutorial with images](https://immersivetranslate.com/docs/services/custom/)). Then increase the `Maximum Requests per Second` in `Custom API Settings` to fully utilize server performance. I set `Maximum Requests per Second` to `5000` and `Maximum Paragraphs per Request` to `1`. You can adjust based on your server hardware. > > - [Kiss Translator](https://github.com/fishjar/kiss-translator) - Scroll down in `Settings` page to find the custom interface `Custom`. Similarly, set `Maximum Concurrent Requests` and `Request Interval Time` to fully utilize server performance. I set `Maximum Concurrent Requests` to `100` and `Request Interval Time` to `1`. You can adjust based on your server configuration. > @@ -155,50 +59,18 @@ Replace `your_token` in the following tables with your `API_TOKEN` or `CORE_API_ **Regular users can start using the service after setting up the plugin interface address according to the table above.** -### Keep Updating - -```bash -cd mtranserver -docker pull xxnuo/mtranserver:latest -docker compose up -d -``` - -### Developer APIs - -> Base URL: `http://localhost:8989` -> -> In v3.0.0, complete API documentation can be viewed at `http://localhost:8989/docs` using Swagger UI - -| Name | URL | Request Format | Response Format | Auth Header | -| --------------------------------------- | ------------------------ | -------------------------------------------------------------------------------------- | ------------------------------------------------------------------ | ------------------------- | -| Service Version | `/version` | None | `{"version": "v3.0.0"}` | None | -| Supported Languages | `/languages` | None | `{"languages":["en","zh","ja","ko"...]}` | Authorization: your_token | -| Standard Translation | `/translate` | `{"from": "en", "to": "zh", "text": "Hello, world!"}` | `{"result": "你好，世界！"}` | Authorization: your_token | -| Batch Translation | `/translate/batch` | `{"from": "en", "to": "zh", "texts": ["Hello, world!", "Hello, world!"]}` | `{"results": ["你好，世界！", "你好，世界！"]}` | Authorization: your_token | -| Health Check | `/health` | None | `{"status": "ok"}` | None | -| Heartbeat Check | `/__heartbeat__` | None | `Ready` | None | -| Load Balancer Heartbeat | `/__lbheartbeat__` | None | `Ready` | None | -| Google Translate Compatible Interface 1 | `/language/translate/v2` | `{"q": "The Great Pyramid of Giza", "source": "en", "target": "zh", "format": "text"}` | `{"data": {"translations": [{"translatedText": "吉萨大金字塔"}]}}` | Authorization: your_token | - -> Developer advanced settings please refer to [CONFIG.md](./CONFIG.md) - ## Support Me [Buy me a coffee ☕️](https://www.creem.io/payment/prod_3QOnrHlGyrtTaKHsOw9Vs1) [Mainland China 💗 Like](./DONATE.md) -## Contributors - - - - - - - - -
Lv Meiyang
Lv Meiyang		Leo
Leo
- ## Star History [![Star History Chart](https://api.star-history.com/svg?repos=xxnuo/MTranServer&type=Timeline)](https://star-history.com/#xxnuo/MTranServer&Timeline) + +## Thanks + +[Bergamot Project](https://browser.mt/) for awesome idea of local translation. + +[Mozilla](https://github.com/mozilla) for the [models](https://github.com/mozilla/firefox-translations-models).