proper python bindings #2
Description
current implementation is very naive and loads a new model on every transcription by simply using subprocess
we should reference this issue ggml-org/whisper.cpp#9 and keep a model in memory
{{ message }}
current implementation is very naive and loads a new model on every transcription by simply using subprocess
we should reference this issue ggml-org/whisper.cpp#9 and keep a model in memory