Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions README-CN.md
Original file line number Diff line number Diff line change
Expand Up @@ -208,6 +208,12 @@ Docker Compose 或 `.env` 中的 `SAG_LLM_*` 用于提供首次启动时的模

保存前,可以点击**测试**检查尚未保存的原始或可选模型连接。成功时显示绿色勾和模型名称或向量维度,失败时显示红色错误;测试不会保存草稿。可选连接保存成功后用于后续操作,已经开始的操作继续使用各自保存的配置快照。后台处理文档时,查询可以复用已初始化的兼容引擎,减少等待。运行细节详见[可选 Chatbot 连接](docs/chatbot-connections.md)。

### 在 Railway 一键部署

[![Deploy on Railway](https://railway.com/button.svg)](https://railway.com/deploy/sag)

基于本仓库 Dockerfile 一键部署 `api` 与 `web` 两个服务,SQLite 与 LanceDB 数据保存在 `/data` 持久化卷中。每次部署都会自动生成 JWT 密钥和用于加密已保存 API Key 的密钥,并启用密码登录,第一个账号注册后即关闭注册。部署完成后请立即打开 web 地址创建账号(该账号成为管理员),然后在 **设置 → 模型** 中配置模型连接,或在 `api` 服务中设置 `SAG_LLM_*` / `SAG_EMBEDDING_*` 变量。

### 导入知识

创建信源后,可以添加 Markdown、文本、PDF、Office 等支持的文档。SAG 会先将文档规范化为 Markdown,再在后台完成分块、向量化、事件抽取和实体抽取。
Expand Down
6 changes: 6 additions & 0 deletions README.md
Original file line number Diff line number Diff line change
Expand Up @@ -208,6 +208,12 @@ Select **OpenAI-compatible**, then choose the **Responses API** format for origi

Use **Test** to check an unsaved original or optional model connection before saving it. Success shows a green check with the model or embedding dimensions; failure shows a red error. Testing does not save the draft. Successful optional connection saves apply to subsequent operations, while operations already running keep their captured configuration. During background document processing, queries can reuse an initialized compatible engine to reduce waiting. See [optional chatbot connections](docs/chatbot-connections.md) for the runtime details.

### Deploy on Railway

[![Deploy on Railway](https://railway.com/button.svg)](https://railway.com/deploy/sag)

One-click deploy of the `api` and `web` services built from this repository's Dockerfiles, with SQLite and LanceDB on a persistent `/data` volume. The JWT secret and the encryption key for saved API keys are generated per deployment, and password auth is enabled with registration closed after the first account. Open the web URL and create your account right away (it becomes the admin), then add your model connections in **Settings → Models** or as `SAG_LLM_*` / `SAG_EMBEDDING_*` variables on the `api` service.

### Import knowledge

Create a source and add Markdown, text, PDF, Office, or other supported documents. SAG normalizes documents to Markdown, then runs chunking, embedding, event extraction, and entity extraction in the background.
Expand Down