Starstruk Logo MarkStarstruk

v3.69

jianchang512/pyvideotransv3.69May 9, 2025by jianchang512

AI Summary

Implemented LLM-based rephrasing logic and improved compatibility with various F5-TTS interfaces.

Key Highlights

  • Rephrasing logic switched to use LLM large models.
  • F5-TTS interface compatibility added for spark-tts, index-tts, and dia-1.6b.
  • Fixed ElevenLabs.io TTS functionality.

New Features

  • LLM-based rephrasing
  • F5-TTS compatibility improvements

Full Release Notes

> 预打包版仅适用于Windows10/11，MacOS和Linux请源码部署
> MacOS/Linux升级：重新拉取源码覆盖，然后执行
>  `pip3 install --upgrade openai-whisper elevenlabs` 
> `pip install --upgrade gradio_client`
> `pip3 install --no-deps  --force-reinstall "faster-whisper @ https://github.com/SYSTRAN/faster-whisper/archive/refs/heads/master.tar.gz"`

## Change

- Feat: 重新断句改为使用LLM大模型进行 [使用和注意事项](https://pyvideotrans.com/rephrase-llm)
- Fix: F5-TTS接口兼容spark-tts/index-tts/dia-1.6b  [仅兼容官方webui接口](https://pvt9.com/f5tts)


## Win v3.69 完整包下载

> 如果未安装过旧版本，请在此下载完整版，如需cuda加速，需有英伟达显卡并且安装cuda12.x及cudnn9

百度网盘下载地址(含tiny/medium模型):   https://pan.baidu.com/s/1xxvju4FmpmYO9l2bAly3tg?pwd=3qzk

GitHub地址: https://github.com/jianchang512/pyvideotrans/releases/download/v3.69/win-videotrans-v3.69-0512-tiny.7z

 

## v3.69 补丁包190MB

> 如果已安装过3.x版本，可下载补丁包后解压在sp.exe所在目录，覆盖已有sp.exe和文件夹

百度网盘下载地址: https://pan.baidu.com/s/1lgoilZA5cSjOkmzNkiBnCA?pwd=v4p3 

GitHub地址: https://github.com/jianchang512/pyvideotrans/releases/download/v3.69/win-PatchUpdate-3.69-0512.7z


> 若补丁覆盖后打开提示`需下载完整包`，请升级显卡驱动、升级cuda到12.x、升级cudnn到cudnn9，并重新下载完整包


----

### 所有模型下载地址

为避免压缩包体积过大，预打包版只内置最小模型 tiny，识别效果不佳，效果更好的模型请点击下载 ，建议至少使用medium模型，推荐large-v2

https://github.com/jianchang512/stt/releases/tag/0.0