Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

语音转文本长时间无响应 #2022

Closed
6 of 7 tasks
Jie2GG opened this issue Jul 11, 2024 · 8 comments
Closed
6 of 7 tasks

语音转文本长时间无响应 #2022

Jie2GG opened this issue Jul 11, 2024 · 8 comments
Labels
bug Something isn't working

Comments

@Jie2GG
Copy link

Jie2GG commented Jul 11, 2024 •
edited
Loading

例行检查

  • 我已确认目前没有类似 issue
  • 我已完整查看过项目 README，以及项目文档
  • 我使用了自己的 key，并确认我的 key 是可正常使用的
  • 我理解并愿意跟进此 issue，协助测试和提供反馈
  • 我理解并认可上述内容，并理解项目维护者精力有限，不遵循规则的 issue 可能会被无视或直接关闭

你的版本

  • 公有云版本
  • 私有部署版本, 具体版本号: v4.8.5

问题描述, 日志截图

使用语音输入时, 未转换出任何文本. OneApi中提示 当前分组 default 下对于模型 whisper-1 无可用渠道! 无渠道问题在 #1139 中已出现过

已排查并确认:

  1. 语音输入接口正常, 使用 Xinference 部署 whisper-large-v3 模型, 并且在 dify 中可以正常将语音转换为文本
  2. FastGPT容器已重启, 并且确认配置文件中已修改语音模型为 whisper-large-v3 (这在日志中有体现)

OneApi日志:

[SYS] 2024/07/11 - 08:21:50 | batch update finished 
[ERR] 2024/07/11 - 08:21:58 | 2024071108215857794614272133362 | 当前分组 default 下对于模型 whisper-1 无可用渠道 
[GIN] 2024/07/11 - 08:21:58 | 2024071108215857794614272133362 | 503 |    2.066591ms |      172.21.0.7 |    POST /v1/audio/transcriptions
[SYS] 2024/07/11 - 08:22:00 | batch update started

FastGPT日志:

▲ Next.js 14.2.3
  - Local:        http://79509f5ea5b3:3000
  - Network:      http://172.21.0.7:3000

 ✓ Starting...
 ✓ Ready in 229ms
mongo start connect
mongo connected
System Version: 4.8.5
{
  feConfigs: {
    lafEnv: 'https://laf.dev',
    show_emptyChat: true,
    show_git: true,
    docUrl: 'https://doc.fastgpt.in',
    openAPIDocUrl: 'https://doc.fastgpt.in/docs/development/openapi',
    systemTitle: 'FastGPT',
    concatMd: '项目开源地址: [FastGPT GitHub](https://github.com/labring/FastGPT)\n' +
      '交流群: ![](https://oss.laf.run/htr4n1-images/fastgpt-qr-code.jpg)',
    limit: { exportDatasetLimitMinutes: 0, websiteSyncLimitMinuted: 0 },
    scripts: [],
    favicon: '/favicon.ico',
    uploadFileMaxSize: 500,
    isPlus: false
  },
  systemEnv: {
    openapiPrefix: 'fastgpt',
    vectorMaxProcess: 15,
    qaMaxProcess: 15,
    pgHNSWEfSearch: 100
  },
  subPlans: undefined,
  llmModels: [
    {
      model: 'Qwen1.5-14B-Chat',
      name: '通义千问1.5-14B-Chat',
      maxContext: 32768,
      avatar: '/imgs/model/openai.svg',
      maxResponse: 8000,
      quoteMaxToken: 20000,
      maxTemperature: 1,
      charsPointsPrice: 0,
      censor: false,
      vision: false,
      datasetProcess: true,
      usedInClassify: true,
      usedInExtractFields: true,
      usedInToolCall: true,
      usedInQueryExtension: true,
      toolChoice: false,
      functionCall: false,
      customCQPrompt: '',
      customExtractPrompt: '',
      defaultSystemChatPrompt: '',
      defaultConfig: {}
    },
    {
      model: 'qwen2-57b-a14b-instruct',
      name: ' 通义千问2-57B-A14B-Instruct',
      maxContext: 131072,
      avatar: '/imgs/model/openai.svg',
      maxResponse: 6144,
      quoteMaxToken: 131072,
      maxTemperature: 1,
      charsPointsPrice: 0,
      censor: false,
      vision: false,
      datasetProcess: true,
      usedInClassify: true,
      usedInExtractFields: true,
      usedInToolCall: true,
      usedInQueryExtension: true,
      toolChoice: false,
      functionCall: false,
      customCQPrompt: '',
      customExtractPrompt: '',
      defaultSystemChatPrompt: '',
      defaultConfig: {}
    },
    {
      model: 'qwen2-72b-instruct',
      name: ' 通义千问2-72B-Instruct',
      maxContext: 131072,
      avatar: '/imgs/model/openai.svg',
      maxResponse: 6144,
      quoteMaxToken: 131072,
      maxTemperature: 1,
      charsPointsPrice: 0,
      censor: false,
      vision: false,
      datasetProcess: true,
      usedInClassify: true,
      usedInExtractFields: true,
      usedInToolCall: true,
      usedInQueryExtension: true,
      toolChoice: false,
      functionCall: false,
      customCQPrompt: '',
      customExtractPrompt: '',
      defaultSystemChatPrompt: '',
      defaultConfig: {}
    }
  ],
  vectorModels: [
    {
      model: 'bge-m3',
      name: 'bge-m3',
      avatar: '/imgs/model/openai.svg',
      charsPointsPrice: 0,
      defaultToken: 512,
      maxToken: 8192,
      weight: 100,
      dbConfig: {},
      queryConfig: {}
    }
  ],
  reRankModels: [
    {
      model: 'bge-reranker-v2-m3',
      name: 'bge-reranker-v2-m3',
      charsPointsPrice: 0,
      requestUrl: 'http://10.206.143.73:9997/v1/rerank',
      requestAuth: 'EMPTY'
    }
  ],
  audioSpeechModels: [
    {
      model: 'tts-1',
      name: 'OpenAI TTS1',
      charsPointsPrice: 0,
      voices: [Array]
    }
  ],
  whisperModel: {
    model: 'whisper-large-v3',
    name: 'whisper-large-v3',
    charsPointsPrice: 0
  }
}
{
  communityPlugins: [
    {
      author: 'FastGPT Team',
      version: '481',
      templateType: 'tools',
      name: '文本加工',
      avatar: '/imgs/workflow/textEditor.svg',
      intro: '可对固定或传入的文本进行加工后输出，非字符串类型数据最终会转成字符串类型。',
      showStatus: false,
      isTool: false,
      weight: 100,
      nodes: [Array],
      edges: [Array],
      id: 'community-textEditor',
      source: 'community'
    },
    {
      author: 'FastGPT Team',
      version: '481',
      templateType: 'tools',
      name: '获取当前时间',
      avatar: '/imgs/workflow/getCurrentTime.svg',
      intro: '获取用户当前时区的时间。',
      showStatus: false,
      isTool: true,
      weight: 10,
      nodes: [Array],
      edges: [Array],
      id: 'community-getCurrentTime',
      source: 'community'
    },
    {
      author: 'FastGPT Team',
      version: '481',
      templateType: 'other',
      name: '自定义反馈',
      avatar: '/imgs/workflow/customFeedback.svg',
      intro: '该模块被触发时，会给当前的对话记录增加一条反馈。可用于自动记录对话效果等。',
      showStatus: false,
      isTool: false,
      weight: 0,
      nodes: [Array],
      edges: [Array],
      id: 'community-customFeedback',
      source: 'community'
    }
  ]
}
pg connected
default team exist new ObjectId("66440c5cc0e5bb500b3b60ee")
root user init: { username: 'root', password: '**********' }
init pg successful
[Info] 2024-07-11 08:21:45 [QA Queue] Done 
[Info] 2024-07-11 08:21:45 [Vector Queue] Done 
Warning: data for page "/app/list" is 133 kB which exceeds the threshold of 128 kB, this amount of data can reduce performance.
See more info here: https://nextjs.org/docs/messages/large-page-data
[Info] 2024-07-11 08:22:00 [QA Queue] Done 
[Info] 2024-07-11 08:22:00 [Vector Queue] Done 
[Info] 2024-07-11 08:23:00 [QA Queue] Done 
[Info] 2024-07-11 08:23:00 [Vector Queue] Done 
[Info] 2024-07-11 08:24:00 [QA Queue] Done 
[Info] 2024-07-11 08:24:00 [Vector Queue] Done 
[Info] 2024-07-11 08:25:00 [QA Queue] Done 
[Info] 2024-07-11 08:25:00 [Vector Queue] Done

复现步骤

预期结果
无法转换文本

相关截图
在 Dify 项目中使用语音模型成功的相关截图:
image
image
image
@Jie2GG Jie2GG added the bug Something isn't working label Jul 11, 2024
@sipeter
Copy link

sipeter commented Jul 23, 2024

我也是这个问题，提示当前分组 default 下对于模型 whisper-1 无可用渠道。本地xinference部署的whisper-large-v3 模型，dify中可以测试语音输入正常。config.json中whisperModel部分只保留whisper-large-v3 模型部分。

下图是oneapi设置和config.json设置，是我哪里设置的问题吗？

image
image

@c121914yu
Copy link
Collaborator

由于 oneapi 坑，目前 model 必须叫 whisper-1，这个 pr 修复了 songquanpeng/one-api#1690

@sipeter
Copy link

sipeter commented Jul 24, 2024

由于 oneapi 坑，目前 model 必须叫 whisper-1，这个 pr 修复了 songquanpeng/one-api#1690

我也试了好多次，感觉是model 必须叫 whisper-1才行，但我看你在 #1139 中提到在xinference中部署成功，那是通过模型重定向改成whisper-1的吗？这种方法我也试了，也没成功，还是没有通过oneapi接入fastgpt的呢？

@c121914yu
Copy link
Collaborator

由于 oneapi 坑，目前 model 必须叫 whisper-1，这个 pr 修复了 songquanpeng/one-api#1690

我也试了好多次，感觉是model 必须叫 whisper-1才行，但我看你在 #1139 中提到在xinference中部署成功，那是通过模型重定向改成whisper-1的吗？这种方法我也试了，也没成功，还是没有通过oneapi接入fastgpt的呢？

没有，因为 xf 中也是部署 whisper-1

@sipeter
Copy link

sipeter commented Jul 24, 2024

由于 oneapi 坑，目前 model 必须叫 whisper-1，这个 pr 修复了 songquanpeng/one-api#1690

我也试了好多次，感觉是model 必须叫 whisper-1才行，但我看你在 #1139 中提到在xinference中部署成功，那是通过模型重定向改成whisper-1的吗？这种方法我也试了，也没成功，还是没有通过oneapi接入fastgpt的呢？

没有，因为 xf 中也是部署 whisper-1

哦，明白了，多谢！

@SDAIer
Copy link

SDAIer commented Aug 19, 2024

@sipeter xf中模型名称改为whisper-1可以了吗

@fsea
Copy link

fsea commented Aug 21, 2024

这个大坑让我排查了2天，是真的要在xf中把uid设置为whisper-1

@sipeter
Copy link

sipeter commented Aug 22, 2024

@sipeter xf中模型名称改为whisper-1可以了吗
我没在xf中测试，后来直接本地fastapi起的服务，但还是需要将模型名字设置为whisper-1，看后来说新版改了这个bug，还没测试新版本。

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
bug Something isn't working
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
5 participants
@fsea @sipeter @Jie2GG @c121914yu @SDAIer