#8572 not implemented (request id: 202610291622404328975248268d9d6Go8J8hpS):Traceback (most recent call last):

2408**c34d Posted at: 1 hour ago

not implemented (request id: 202610291622404328975248268d9d6Go8J8hpS):
Traceback (most recent call last):
File "videotrans\util\TestSTT.py", line 21, in run
File "videotrans\recognition\__init__.py", line 121, in run
File "videotrans\recognition\_base.py", line 99, in run
File "videotrans\recognition\_openairecognapi.py", line 39, in _exec
File "videotrans\recognition\_openairecognapi.py", line 99, in _thrid_api
File "openai\_utils\_utils.py", line 298, in wrapper
File "openai\resources\audio\transcriptions.py", line 485, in create
File "openai\_base_client.py", line 1332, in post
File "openai\_base_client.py", line 1105, in request
openai.InternalServerError: Error code: 500 - {'error': {'message': 'not implemented (request id: 202610291622404328975248268d9d6Go8J8hpS)', 'type': 'new_api_error', 'param': '', 'code': 'convert_request_failed'}}


=
system:Windows-10-10.0.19045-SP0
version:v4.14
frozen:True
language:zh_CN
root_dir:D:/Program Files (x86)/pyvideotrans4.14

1 Replies

#1 DeepSeek reply at 1 hour ago

核心原因:这不是 pyVideoTrans 本身的 bug,而是你在「OpenAI 语音识别 API」渠道里填写的中转/聚合 API 地址(返回体中 type: new_api_error、code: convert_request_failed 表明是第三方 new-api 类中转站)没有实现 /v1/audio/transcriptions 音频转录接口,服务端直接返回 500 not implemented。

解决方案(按顺序尝试):

  1. 更换 API 地址/服务商

    • 进入 菜单 → 翻译设置 → OpenAI(或对应 OpenAI 兼容渠道设置),将 API 地址改为官方地址 https://api.openai.com/v1,或换用明确支持 /v1/audio/transcriptions 的渠道。
    • 多数廉价中转站只转发 chat 类文本接口,不支持音频上传,这是该报错最常见来源。
  2. 不想折腾 API,直接用内置离线识别(推荐)

    • 语音识别渠道改为 faster-whisper(内置),模型选 large-v3-turbo(默认)或 large-v3;有 N 卡并配置好 CUDA 12.8 + cuDNN 9 时勾选 CUDA加速。
  3. 修正软件安装路径(务必处理)

    • 当前安装在 D:/Program Files (x86)/pyvideotrans4.14,属于系统受保护目录,可能引发临时文件写入、模型下载等一系列异常。请完整解压到浅层纯英文数字目录,如 D:/pyvideotrans,再双击 sp.exe。
  4. 若换官方地址后仍报 500,请先点 试听/测试(等同于本次触发的 TestSTT)确认密钥与余额有效,再重新识别。

附:本次报错栈源自 TestSTT 测试流程,说明失败发生在“识别前的接口连通性测试”阶段,视频与字幕本身无问题。

Thinking...
Post Your Reply
If you have more than 50 messages, you will no longer receive an automatic reply. Please ask a new question!
Open source and free maintenance is not easy. If this project is helpful to you, please consider making a small donation to help the project continue to maintain and update.

Dondate via ko-fi.com

Related resource