Skip to content

windows-videotrans-v3.42

Compare
Choose a tag to compare
@jianchang512 jianchang512 released this 28 Dec 13:51
· 9 commits to main since this release

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
pip3 install --no-deps --force-reinstall "faster-whisper @ https://github.com/SYSTRAN/faster-whisper/archive/refs/heads/master.tar.gz"

ChangeLogs

  • 优化edge-tts配音渠道任务算法
  • 批量字幕配音 功能允许配音后文件保存到原字幕文件夹

Win v3.42 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/14iza4s3nRxkHaGsV5rkyrg?pwd=amb8

GitHub下载地址(仅含tiny模型): https://github.com/jianchang512/pyvideotrans/releases/download/v3.42/win-videotrans-v3.42-tiny.7z

v3.42 补丁包190MB

如果已安装过2.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/19P8MOOz34U_WeHnj4HCWqg?pwd=kgaf

GitHub地址: https://github.com/jianchang512/pyvideotrans/releases/download/v3.42/win-PatchUpdate-3.42.7z

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


api.exe 下载,api.exe是用于通过http请求调用的文件,不可独立使用,需先下载完整包,然后下载api.exe和 sp.exe放在同一目录下

api.exe下载地址:https://github.com/jianchang512/pyvideotrans/releases/download/v3.36/api.exe


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