mirror of
https://github.com/labring/FastGPT.git
synced 2025-07-21 11:43:56 +00:00

* New chatinput (#4995) * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * style: Add new 'xxl' size to theme spacing for improved layout options * style: Update close icon fill color to use currentColor for better theming support * style: Enhance voice input functionality and UI responsiveness; improve waveform sensitivity and amplitude * style: Conditionally render file preview based on voice input state * style: 优化移动端音频波形渲染,增强清晰度和敏感度 * style: Update comments to English to enhance code readability and consistency * style: Adjust the mobile audio waveform update frequency and optimize rendering performance * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Adjust the chat input box placeholder color and border color to enhance visual effects * fix: pg test * Test secret (#5011) * add http header auth config (#4982) * add http header auth config * optimize code * add mcp tools header auth * fix build * fix ui * fix * teamid * secret value encrypt (#5002) * perf: secret code * header auth ui (#5012) * header auth ui * fix i18n * doc * perf: type * header secret ui * reset ui * perf: check secret invalid --------- Co-authored-by: heheer <heheer@sealos.io> * feat: cq and extrat AI memory (#5013) * fix: login xss * feat: Users can download the invoice by self (#5015) * Users can download the invoice by themselves * Direct file stream implementation for transmission presentation * i18n * Chatbox-fix (#5018) * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * style: Add new 'xxl' size to theme spacing for improved layout options * style: Update close icon fill color to use currentColor for better theming support * style: Enhance voice input functionality and UI responsiveness; improve waveform sensitivity and amplitude * style: Conditionally render file preview based on voice input state * style: 优化移动端音频波形渲染,增强清晰度和敏感度 * style: Update comments to English to enhance code readability and consistency * style: Adjust the mobile audio waveform update frequency and optimize rendering performance * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Adjust the chat input box placeholder color and border color to enhance visual effects * New chatinput (#4995) * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * feat: Change border style * refactor: Improve layout and styling of ChatInput component * style: Update ChatInput component styling and mobile layout * fix: update key detection for Enter key in ChatInput component * feat: 添加 WelcomePage 组件,支持变量输入和欢迎信息展示 * style: Updated the PC voice input interface of the VoiceInput component and optimized the layout and style * style: Optimize the layout and style of the WelcomePage component, and adjust the responsive design * feat: Dynamically load the WelcomePage component and optimize the welcome information display logic * refactor: Adjust the style and behavior of the ChatInput component and delete the WelcomePage component * style: Modify the minimum height setting of the ChatInput component to simplify responsive design * style: Optimize the layout and style of PC voice input components, and enhance the processing and drawing logic of waveform data * style: Adjust ChatInput component's margin and textarea height logic for improved layout and responsiveness; refine PCVoiceInput component's positioning and display elements * style: Enhance PCVoiceInput component's time display styling with custom font properties * style: Add new 'xxl' size to theme spacing for improved layout options * style: Update close icon fill color to use currentColor for better theming support * style: Enhance voice input functionality and UI responsiveness; improve waveform sensitivity and amplitude * style: Conditionally render file preview based on voice input state * style: 优化移动端音频波形渲染,增强清晰度和敏感度 * style: Update comments to English to enhance code readability and consistency * style: Adjust the mobile audio waveform update frequency and optimize rendering performance * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Optimize the file preview rendering logic in voice input mode to enhance user experience * style: Adjust the chat input box placeholder color and border color to enhance visual effects * fix: pg test * Test secret (#5011) * add http header auth config (#4982) * add http header auth config * optimize code * add mcp tools header auth * fix build * fix ui * fix * teamid * secret value encrypt (#5002) * perf: secret code * header auth ui (#5012) * header auth ui * fix i18n * doc * perf: type * header secret ui * reset ui * perf: check secret invalid --------- Co-authored-by: heheer <heheer@sealos.io> * feat: cq and extrat AI memory (#5013) * refactor: Refactored the ChatInput component, optimized the layout of the text area and button group, and improved the user experience * refactor: Updated ChatInput component, optimized layout and style, and enhanced user experience * feat: update docs --------- Co-authored-by: archer <545436317@qq.com> Co-authored-by: heheer <heheer@sealos.io> * input ui * fix: chat input ux * Return in JSON format to handle checkres (#5019) * Users can download the invoice by themselves * Direct file stream implementation for transmission presentation * Return in JSON format to handle checkres * fix: invoice * fix: ui * doc * update package * fix: ts * fix: login checker * fix: team plan * perf: aiproxy ux --------- Co-authored-by: Theresa <63280168+sd0ric4@users.noreply.github.com> Co-authored-by: heheer <heheer@sealos.io> Co-authored-by: Zhuangzai fa <143257420+ctrlz526@users.noreply.github.com>
457 lines
18 KiB
Markdown
457 lines
18 KiB
Markdown
---
|
||
title: 'FastGPT 模型配置说明'
|
||
description: 'FastGPT 模型配置说明'
|
||
icon: 'api'
|
||
draft: false
|
||
toc: true
|
||
weight: 744
|
||
---
|
||
|
||
在 4.8.20 版本以前,FastGPT 模型配置在 `config.json` 文件中声明,你可以在 https://github.com/labring/FastGPT/blob/main/projects/app/data/model.json 中找到旧版的配置文件示例。
|
||
|
||
从 4.8.20 版本开始,你可以直接在 FastGPT 页面中进行模型配置,并且系统内置了大量模型,无需从 0 开始配置。下面介绍模型配置的基本流程:
|
||
|
||
## 配置模型
|
||
|
||
### 1. 对接模型提供商
|
||
|
||
#### AI Proxy
|
||
|
||
从 4.8.23 版本开始, FastGPT 支持在页面上配置模型提供商,即使用 [AI Proxy 接入教程](/docs/development/modelconfig/ai-proxy) 来进行模型聚合,从而可以对接更多模型提供商。
|
||
|
||
#### One API
|
||
|
||
也可以使用 [OneAPI 接入教程](/docs/development/modelconfig/one-api)。你需要先在各服务商申请好 API 接入 OneAPI 后,才能在 FastGPT 中使用这些模型。示例流程如下:
|
||
|
||

|
||
|
||
除了各模型官方的服务外,还有一些第三方服务商提供模型接入服务,当然你也可以用 Ollama 等来部署本地模型,最终都需要接入 OneAPI,下面是一些第三方服务商:
|
||
|
||
{{% alert icon=" " context="info" %}}
|
||
- [SiliconCloud(硅基流动)](https://cloud.siliconflow.cn/i/TR9Ym0c4): 提供开源模型调用的平台。
|
||
- [Sealos AIProxy](https://hzh.sealos.run/?uid=fnWRt09fZP&openapp=system-aiproxy): 提供国内各家模型代理,无需逐一申请 api。
|
||
{{% /alert %}}
|
||
|
||
在 OneAPI 配置好模型后,你就可以打开 FastGPT 页面,启用对应模型了。
|
||
|
||
### 2. 配置介绍
|
||
|
||
{{% alert icon="🤖 " context="success" %}}
|
||
注意:
|
||
1. 目前语音识别模型和重排模型仅会生效一个,所以配置时候,只需要配置一个即可。
|
||
2. 系统至少需要一个语言模型和一个索引模型才能正常使用。
|
||
{{% /alert %}}
|
||
|
||
#### 核心配置
|
||
|
||
- 模型 ID:接口请求时候,Body 中`model`字段的值,全局唯一。
|
||
- 自定义请求地址/Key:如果需要绕过`OneAPI`,可以设置自定义请求地址和 Token。一般情况下不需要,如果 OneAPI 不支持某些模型,可以使用该特性。
|
||
|
||
#### 模型类型
|
||
|
||
1. 语言模型 - 进行文本对话,多模态模型支持图片识别。
|
||
2. 索引模型 - 对文本块进行索引,用于相关文本检索。
|
||
3. 重排模型 - 对检索结果进行重排,用于优化检索排名。
|
||
4. 语音合成 - 将文本转换为语音。
|
||
5. 语音识别 - 将语音转换为文本。
|
||
|
||
#### 启用模型
|
||
|
||
系统内置了目前主流厂商的模型,如果你不熟悉配置,直接点击`启用`即可,需要注意的是,`模型 ID`需要和 OneAPI 中渠道的`模型`一致。
|
||
|
||
| | |
|
||
| --- | --- |
|
||
|  |  |
|
||
|
||
#### 修改模型配置
|
||
|
||
点击模型右侧的齿轮即可进行模型配置,不同类型模型的配置有区别。
|
||
|
||
| | |
|
||
| --- | --- |
|
||
|  |  |
|
||
|
||
## 新增自定义模型
|
||
|
||
如果系统内置的模型无法满足你的需求,你可以添加自定义模型。自定义模型中,如果`模型 ID`与系统内置的模型 ID 一致,则会被认为是修改系统模型。
|
||
|
||
| | |
|
||
| --- | --- |
|
||
|  |  |
|
||
|
||
#### 通过配置文件配置
|
||
|
||
如果你觉得通过页面配置模型比较麻烦,你也可以通过配置文件来配置模型。或者希望快速将一个系统的配置,复制到另一个系统,也可以通过配置文件来实现。
|
||
|
||
| | |
|
||
| --- | --- |
|
||
|  |  |
|
||
|
||
**语言模型字段说明:**
|
||
|
||
```json
|
||
{
|
||
"model": "模型 ID",
|
||
"metadata": {
|
||
"isCustom": true, // 是否为自定义模型
|
||
"isActive": true, // 是否启用
|
||
"provider": "OpenAI", // 模型提供商,主要用于分类展示,目前已经内置提供商包括:https://github.com/labring/FastGPT/blob/main/packages/global/core/ai/provider.ts, 可 pr 提供新的提供商,或直接填写 Other
|
||
"model": "gpt-4o-mini", // 模型ID(对应OneAPI中渠道的模型名)
|
||
"name": "gpt-4o-mini", // 模型别名
|
||
"maxContext": 125000, // 最大上下文
|
||
"maxResponse": 16000, // 最大回复
|
||
"quoteMaxToken": 120000, // 最大引用内容
|
||
"maxTemperature": 1.2, // 最大温度
|
||
"charsPointsPrice": 0, // n积分/1k token(商业版)
|
||
"censor": false, // 是否开启敏感校验(商业版)
|
||
"vision": true, // 是否支持图片输入
|
||
"datasetProcess": true, // 是否设置为文本理解模型(QA),务必保证至少有一个为true,否则知识库会报错
|
||
"usedInClassify": true, // 是否用于问题分类(务必保证至少有一个为true)
|
||
"usedInExtractFields": true, // 是否用于内容提取(务必保证至少有一个为true)
|
||
"usedInToolCall": true, // 是否用于工具调用(务必保证至少有一个为true)
|
||
"toolChoice": true, // 是否支持工具选择(分类,内容提取,工具调用会用到。)
|
||
"functionCall": false, // 是否支持函数调用(分类,内容提取,工具调用会用到。会优先使用 toolChoice,如果为false,则使用 functionCall,如果仍为 false,则使用提示词模式)
|
||
"defaultSystemChatPrompt": "", // 对话默认携带的系统提示词
|
||
"defaultConfig": {}, // 请求API时,挟带一些默认配置(比如 GLM4 的 top_p)
|
||
"fieldMap": {} // 字段映射(o1 模型需要把 max_tokens 映射为 max_completion_tokens)
|
||
}
|
||
}
|
||
```
|
||
|
||
**索引模型字段说明:**
|
||
|
||
```json
|
||
{
|
||
"model": "模型 ID",
|
||
"metadata": {
|
||
"isCustom": true, // 是否为自定义模型
|
||
"isActive": true, // 是否启用
|
||
"provider": "OpenAI", // 模型提供商
|
||
"model": "text-embedding-3-small", // 模型ID
|
||
"name": "text-embedding-3-small", // 模型别名
|
||
"charsPointsPrice": 0, // n积分/1k token
|
||
"defaultToken": 512, // 默认文本分割时候的 token
|
||
"maxToken": 3000 // 最大 token
|
||
}
|
||
}
|
||
```
|
||
|
||
**重排模型字段说明:**
|
||
|
||
```json
|
||
{
|
||
"model": "模型 ID",
|
||
"metadata": {
|
||
"isCustom": true, // 是否为自定义模型
|
||
"isActive": true, // 是否启用
|
||
"provider": "BAAI", // 模型提供商
|
||
"model": "bge-reranker-v2-m3", // 模型ID
|
||
"name": "ReRanker-Base", // 模型别名
|
||
"requestUrl": "", // 自定义请求地址
|
||
"requestAuth": "", // 自定义请求认证
|
||
"type": "rerank" // 模型类型
|
||
}
|
||
}
|
||
```
|
||
|
||
**语音合成模型字段说明:**
|
||
|
||
```json
|
||
{
|
||
"model": "模型 ID",
|
||
"metadata": {
|
||
"isActive": true, // 是否启用
|
||
"isCustom": true, // 是否为自定义模型
|
||
"type": "tts", // 模型类型
|
||
"provider": "FishAudio", // 模型提供商
|
||
"model": "fishaudio/fish-speech-1.5", // 模型ID
|
||
"name": "fish-speech-1.5", // 模型别名
|
||
"voices": [ // 音色
|
||
{
|
||
"label": "fish-alex", // 音色名称
|
||
"value": "fishaudio/fish-speech-1.5:alex", // 音色ID
|
||
},
|
||
{
|
||
"label": "fish-anna", // 音色名称
|
||
"value": "fishaudio/fish-speech-1.5:anna", // 音色ID
|
||
}
|
||
],
|
||
"charsPointsPrice": 0 // n积分/1k token
|
||
}
|
||
}
|
||
```
|
||
|
||
**语音识别模型字段说明:**
|
||
|
||
```json
|
||
{
|
||
"model": "whisper-1",
|
||
"metadata": {
|
||
"isActive": true, // 是否启用
|
||
"isCustom": true, // 是否为自定义模型
|
||
"provider": "OpenAI", // 模型提供商
|
||
"model": "whisper-1", // 模型ID
|
||
"name": "whisper-1", // 模型别名
|
||
"charsPointsPrice": 0, // n积分/1k token
|
||
"type": "stt" // 模型类型
|
||
}
|
||
}
|
||
```
|
||
|
||
## 模型测试
|
||
|
||
FastGPT 页面上提供了每类模型的简单测试,可以初步检查模型是否正常工作,会实际按模板发送一个请求。
|
||
|
||

|
||
|
||
## 特殊接入示例
|
||
|
||
### ReRank 模型接入
|
||
|
||
由于 OneAPI 不支持 Rerank 模型,所以需要单独配置。FastGPT 中,模型配置支持自定义请求地址,可以绕过 OneAPI,直接向提供商发起请求,可以利用这个特性来接入 Rerank 模型。
|
||
|
||
|
||
#### 使用硅基流动的在线模型
|
||
|
||
有免费的 `bge-reranker-v2-m3` 模型可以使用。
|
||
|
||
1. [点击注册硅基流动账号](https://cloud.siliconflow.cn/i/TR9Ym0c4)
|
||
2. 进入控制台,获取 API key: https://cloud.siliconflow.cn/account/ak
|
||
3. 打开 FastGPT 模型配置,新增一个`BAAI/bge-reranker-v2-m3`的重排模型(如果系统内置了,也可以直接变更,无需新增)。
|
||
|
||

|
||
|
||
#### 私有部署模型
|
||
|
||
[点击查看部署 ReRank 模型教程](/docs/development/custom-models/bge-rerank/)
|
||
|
||
### 接入语音识别模型
|
||
|
||
OneAPI 的语言识别接口,无法正确的识别其他模型(会始终识别成 whisper-1),所以如果想接入其他模型,可以通过自定义请求地址来实现。例如,接入硅基流动的 `FunAudioLLM/SenseVoiceSmall` 模型,可以参考如下配置:
|
||
|
||
点击模型编辑:
|
||
|
||

|
||
|
||
填写硅基流动的地址:`https://api.siliconflow.cn/v1/audio/transcriptions`,并填写硅基流动的 API Key。
|
||
|
||

|
||
|
||
## 其他配置项说明
|
||
|
||
### 自定义请求地址
|
||
|
||
如果填写了该值,则可以允许你绕过 OneAPI,直接向自定义请求地址发起请求。需要填写完整的请求地址,例如:
|
||
|
||
- LLM: {{host}}/v1/chat/completions
|
||
- Embedding: {{host}}/v1/embeddings
|
||
- STT: {{host}}/v1/audio/transcriptions
|
||
- TTS: {{host}}/v1/audio/speech
|
||
- Rerank: {{host}}/v1/rerank
|
||
|
||
自定义请求 Key,则是向自定义请求地址发起请求时候,携带请求头:Authorization: Bearer xxx 进行请求。
|
||
|
||
所有接口均遵循 OpenAI 提供的模型格式,可参考 [OpenAI API 文档](https://platform.openai.com/docs/api-reference/introduction) 进行配置。
|
||
|
||
由于 OpenAI 没有提供 ReRank 模型,遵循的是 Cohere 的格式。[点击查看接口请求示例](/docs/development/faq/#如何检查模型问题)
|
||
|
||
|
||
### 模型价格配置
|
||
|
||
商业版用户可以通过配置模型价格,来进行账号计费。系统包含两种计费模式:按总 tokens 计费和输入输出 Tokens 分开计费。
|
||
|
||
如果需要配置`输入输出 Tokens 分开计费模式`,则填写`模型输入价格`和`模型输出价格`两个值。
|
||
如果需要配置`按总 tokens 计费模式`,则填写`模型综合价格`一个值。
|
||
|
||
## 如何提交内置模型
|
||
|
||
由于模型更新非常频繁,官方不一定及时更新,如果未能找到你期望的内置模型,你可以[提交 Issue](https://github.com/labring/FastGPT/issues),提供模型的名字和对应官网。或者直接[提交 PR](https://github.com/labring/FastGPT/pulls),提供模型配置。
|
||
|
||
|
||
### 添加模型提供商
|
||
|
||
如果你需要添加模型提供商,需要修改以下代码:
|
||
|
||
1. FastGPT/packages/web/components/common/Icon/icons/model - 在此目录下,添加模型提供商的 svg 头像地址。
|
||
2. 在 FastGPT 根目录下,运行`pnpm initIcon`,将图片加载到配置文件中。
|
||
3. FastGPT/packages/global/core/ai/provider.ts - 在此文件中,追加模型提供商的配置。
|
||
|
||
### 添加模型
|
||
|
||
你可以在`FastGPT/packages/service/core/ai/config/provider`目录下,找对应模型提供商的配置文件,并追加模型配置。请自行全文检查,`model`字段,必须在所有模型中唯一。具体配置字段说明,参考[模型配置字段说明](/docs/development/modelconfig/intro/#通过配置文件配置)
|
||
|
||
## 旧版模型配置说明
|
||
|
||
配置好 OneAPI 后,需要在`config.json`文件中,手动的增加模型配置,并重启。
|
||
|
||
由于环境变量不利于配置复杂的内容,FastGPT 采用了 ConfigMap 的形式挂载配置文件,你可以在 `projects/app/data/config.json` 看到默认的配置文件。可以参考 [docker-compose 快速部署](/docs/development/docker/) 来挂载配置文件。
|
||
|
||
**开发环境下**,你需要将示例配置文件 `config.json` 复制成 `config.local.json` 文件才会生效。
|
||
**Docker部署**,修改`config.json` 文件,需要重启容器。
|
||
|
||
下面配置文件示例中包含了系统参数和各个模型配置:
|
||
|
||
```json
|
||
{
|
||
"feConfigs": {
|
||
"lafEnv": "https://laf.dev" // laf环境。 https://laf.run (杭州阿里云) ,或者私有化的laf环境。如果使用 Laf openapi 功能,需要最新版的 laf 。
|
||
},
|
||
"systemEnv": {
|
||
"vectorMaxProcess": 15, // 向量处理线程数量
|
||
"qaMaxProcess": 15, // 问答拆分线程数量
|
||
"tokenWorkers": 50, // Token 计算线程保持数,会持续占用内存,不能设置太大。
|
||
"hnswEfSearch": 100 // 向量搜索参数,仅对 PG 和 OB 生效。越大,搜索越精确,但是速度越慢。设置为100,有99%+精度。
|
||
},
|
||
"llmModels": [
|
||
{
|
||
"provider": "OpenAI", // 模型提供商,主要用于分类展示,目前已经内置提供商包括:https://github.com/labring/FastGPT/blob/main/packages/global/core/ai/provider.ts, 可 pr 提供新的提供商,或直接填写 Other
|
||
"model": "gpt-4o-mini", // 模型名(对应OneAPI中渠道的模型名)
|
||
"name": "gpt-4o-mini", // 模型别名
|
||
"maxContext": 125000, // 最大上下文
|
||
"maxResponse": 16000, // 最大回复
|
||
"quoteMaxToken": 120000, // 最大引用内容
|
||
"maxTemperature": 1.2, // 最大温度
|
||
"charsPointsPrice": 0, // n积分/1k token(商业版)
|
||
"censor": false, // 是否开启敏感校验(商业版)
|
||
"vision": true, // 是否支持图片输入
|
||
"datasetProcess": true, // 是否设置为文本理解模型(QA),务必保证至少有一个为true,否则知识库会报错
|
||
"usedInClassify": true, // 是否用于问题分类(务必保证至少有一个为true)
|
||
"usedInExtractFields": true, // 是否用于内容提取(务必保证至少有一个为true)
|
||
"usedInToolCall": true, // 是否用于工具调用(务必保证至少有一个为true)
|
||
"toolChoice": true, // 是否支持工具选择(分类,内容提取,工具调用会用到。)
|
||
"functionCall": false, // 是否支持函数调用(分类,内容提取,工具调用会用到。会优先使用 toolChoice,如果为false,则使用 functionCall,如果仍为 false,则使用提示词模式)
|
||
"defaultSystemChatPrompt": "", // 对话默认携带的系统提示词
|
||
"defaultConfig": {}, // 请求API时,挟带一些默认配置(比如 GLM4 的 top_p)
|
||
"fieldMap": {} // 字段映射(o1 模型需要把 max_tokens 映射为 max_completion_tokens)
|
||
},
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "gpt-4o",
|
||
"name": "gpt-4o",
|
||
"maxContext": 125000,
|
||
"maxResponse": 4000,
|
||
"quoteMaxToken": 120000,
|
||
"maxTemperature": 1.2,
|
||
"charsPointsPrice": 0,
|
||
"censor": false,
|
||
"vision": true,
|
||
"datasetProcess": true,
|
||
"usedInClassify": true,
|
||
"usedInExtractFields": true,
|
||
"usedInToolCall": true,
|
||
"toolChoice": true,
|
||
"functionCall": false,
|
||
"defaultSystemChatPrompt": "",
|
||
"defaultConfig": {},
|
||
"fieldMap": {}
|
||
},
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "o1-mini",
|
||
"name": "o1-mini",
|
||
"maxContext": 125000,
|
||
"maxResponse": 65000,
|
||
"quoteMaxToken": 120000,
|
||
"maxTemperature": 1.2,
|
||
"charsPointsPrice": 0,
|
||
"censor": false,
|
||
"vision": false,
|
||
"datasetProcess": true,
|
||
"usedInClassify": true,
|
||
"usedInExtractFields": true,
|
||
"usedInToolCall": true,
|
||
"toolChoice": false,
|
||
"functionCall": false,
|
||
"defaultSystemChatPrompt": "",
|
||
"defaultConfig": {
|
||
"temperature": 1,
|
||
"max_tokens": null,
|
||
"stream": false
|
||
}
|
||
},
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "o1-preview",
|
||
"name": "o1-preview",
|
||
"maxContext": 125000,
|
||
"maxResponse": 32000,
|
||
"quoteMaxToken": 120000,
|
||
"maxTemperature": 1.2,
|
||
"charsPointsPrice": 0,
|
||
"censor": false,
|
||
"vision": false,
|
||
"datasetProcess": true,
|
||
"usedInClassify": true,
|
||
"usedInExtractFields": true,
|
||
"usedInToolCall": true,
|
||
"toolChoice": false,
|
||
"functionCall": false,
|
||
"defaultSystemChatPrompt": "",
|
||
"defaultConfig": {
|
||
"temperature": 1,
|
||
"max_tokens": null,
|
||
"stream": false
|
||
}
|
||
}
|
||
],
|
||
"vectorModels": [
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "text-embedding-3-small",
|
||
"name": "text-embedding-3-small",
|
||
"charsPointsPrice": 0,
|
||
"defaultToken": 512,
|
||
"maxToken": 3000,
|
||
"weight": 100
|
||
},
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "text-embedding-3-large",
|
||
"name": "text-embedding-3-large",
|
||
"charsPointsPrice": 0,
|
||
"defaultToken": 512,
|
||
"maxToken": 3000,
|
||
"weight": 100,
|
||
"defaultConfig": {
|
||
"dimensions": 1024
|
||
}
|
||
},
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "text-embedding-ada-002", // 模型名(与OneAPI对应)
|
||
"name": "Embedding-2", // 模型展示名
|
||
"charsPointsPrice": 0, // n积分/1k token
|
||
"defaultToken": 700, // 默认文本分割时候的 token
|
||
"maxToken": 3000, // 最大 token
|
||
"weight": 100, // 优先训练权重
|
||
"defaultConfig": {}, // 自定义额外参数。例如,如果希望使用 embedding3-large 的话,可以传入 dimensions:1024,来返回1024维度的向量。(目前必须小于1536维度)
|
||
"dbConfig": {}, // 存储时的额外参数(非对称向量模型时候需要用到)
|
||
"queryConfig": {} // 参训时的额外参数
|
||
}
|
||
],
|
||
"reRankModels": [],
|
||
"audioSpeechModels": [
|
||
{
|
||
"provider": "OpenAI",
|
||
"model": "tts-1",
|
||
"name": "OpenAI TTS1",
|
||
"charsPointsPrice": 0,
|
||
"voices": [
|
||
{ "label": "Alloy", "value": "alloy", "bufferId": "openai-Alloy" },
|
||
{ "label": "Echo", "value": "echo", "bufferId": "openai-Echo" },
|
||
{ "label": "Fable", "value": "fable", "bufferId": "openai-Fable" },
|
||
{ "label": "Onyx", "value": "onyx", "bufferId": "openai-Onyx" },
|
||
{ "label": "Nova", "value": "nova", "bufferId": "openai-Nova" },
|
||
{ "label": "Shimmer", "value": "shimmer", "bufferId": "openai-Shimmer" }
|
||
]
|
||
}
|
||
],
|
||
"whisperModel": {
|
||
"provider": "OpenAI",
|
||
"model": "whisper-1",
|
||
"name": "Whisper1",
|
||
"charsPointsPrice": 0
|
||
}
|
||
}
|
||
```
|