admin管理员组文章数量:1401959
I wanted to generate world level highlighting for my videos using WhisperX as is shown here where the subtitle shows up and words get highlighted as they are spoken.
I have installed the dependencies in as given in the GitHub page . However when I run the command on Command Prompt in Windows
C:\Users\profe>whisperx "C:mypathtofile\test audio 2.mp3" --model medium --align_model WAV2VEC2_ASR_LARGE_LV60K_960H --highlight_words True --verbose True
the Command Prompt stops at Performing alignment... as is shown here
the files that are generated as output are:
The JSON file has the world level timestamps. Things I tired doing:
- changed the alignment model
- change the model
- added --device CUDA & --verbose True None of it seemed to let it proceed beyond the "Performing Alignment..." part. Any leads on this will be greatly appreciated. Thank you
本文标签:
版权声明:本文标题:openai whisper - Unable to perfrom alignment for word-level timestamp in WhisperX run locally on NVIDIA GForce RTX 3050 - Stack 内容由网友自发贡献,该文观点仅代表作者本人, 转载请联系作者并注明出处:http://www.betaflare.com/web/1744327641a2600805.html, 本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容,一经查实,本站将立刻删除。
发表评论