-
|
Hi. Now I have a question. Every time I run whisper, this loads the model to the GPU memory, I want to know if it's possible to keep the model always in memory so it will be faster. Best regards. Max |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments
-
|
Hi, you can specify multiple audio files in the command line like |
Beta Was this translation helpful? Give feedback.
-
|
Thank you so much @jongwook for your answer, now I understand better why some other whisper projects can keep the models loaded! |
Beta Was this translation helpful? Give feedback.
Hi, you can specify multiple audio files in the command line like
whisper *.mp3to load the model once and transcribe all files. For more control, you'll need to use the Python interface for this because the GPU memory is released once the Python process exits. Please also see #516.