Gradio WebUI for Faster Whisper model
This efficient script utilizes the Faster-Whisper model (available at https://github.com/guillaumekln/faster-whisper) to generate highly accurate word-level ASS subtitles. By leveraging this model, the script achieves a significant speed-up and lower VRAM usage compared to original Whisper model. Before running the script, be sure to follow the installation instructions for Faster-Whisper.
You can run this app by running gradio_app.py.
Before first run, make sure to install all the dependencies by running pip install -r requirements.txt.
python3 transcribe.py *.flac -f -l ja -t zh*.flac: Indicates that all files in the current directory with the file extension ".flac" will be translated.-f: Indicates that the instruction will be forcefully executed, overwriting the output target file if it already exists.-l ja: Indicates that the input language is Japanese.-t zh: Indicates that the input language will be translated into Chinese.
利用Gradio编写的Faster-Whisper模型WebUI
这个高效的脚本利用 Faster-Whisper 模型（可在 https://github.com/guillaumekln/faster-whisper 获取）生成高度准确的单词级别 ASS 字幕。通过这个模型，该脚本相比原版Whisper模型实现了更高的效率和更低的显存占用。在运行脚本之前，请确保按照 Faster-Whisper 的安装说明进行安装。
你可以通过运行gradio_app.py来启动WebUI。
在首次运行前，请先确保安装所有依赖，通过pip install -r requirements.txt来安装。
python3 transcribe.py *.flac -f -l ja -t zh*.flac：表示将当前目录下所有扩展名为 .flac 的文件都进行识别和翻译。-f：表示强制执行，如果输出目标文件已经存在，则覆盖它。-l ja：表示输入语言为日语。-t zh：表示将输入语言翻译成中文。