Last active
January 25, 2024 21:19
-
-
Save 2shortplanks/0432463c8197b9974f00ddcc2f060930 to your computer and use it in GitHub Desktop.
script to transcripe anything with whisper
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/bin/sh | |
# this script relies on a bunch of stuff | |
# | |
# 1. ffmpeg has been installed by brew | |
# 2. The instructions from https://blog.castopod.org/install-whisper-cpp-on-your-mac-in-5mn-and-transcribe-all-your-podcasts-for-free/ | |
# have been followed to install whisper.cpp and the models. These are: | |
# cd ~/co | |
# git clone https://github.com/ggerganov/whisper.cpp.git | |
# cd whisper.cpp | |
# make | |
# cd models | |
# ./download-ggml-model.sh small | |
# Create a temporary file | |
tempfile="$(mktemp).wav" | |
# convert the file into a wav file that whisper.cpp can read and put it there | |
ffmpeg -i "$1" -ar 16000 $tempfile 2>/dev/null | |
# use whisper to transcribe it | |
~/co/whisper.cpp/main \ | |
-nt \ | |
-m ~/co/whisper.cpp/models/ggml-small.bin \ | |
-f $tempfile \ | |
2>/dev/null \ | |
| perl -e '$_ = join "", <>; s/^\s+//ms; s/\s+$//ms; print' | |
# Delete the temp file manually at the end of the script | |
rm -f $tempfile |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment