whisper.nvim

#!/bin/bash

# INSTRUCTIONS
#
# This simple script is called by Neovim to capture audio from the microphone and transcribe it with Whisper.
# In order for this to work, you need to clone the whisper.cpp repo and build the 'stream' tool
#
#   git clone https://github.com/ggerganov/whisper.cpp
#   cd whisper.cpp
#   make stream
#
# Also, make sure the current script is in your PATH env variable. You should be able to run the following command:
#
#   whisper.nvim
#
# Next, export the path to the whisper.cpp repository via the WHISPER_CPP_HOME env variable:
#
#   export WHISPER_CPP_HOME=/path/to/whisper.cpp
#
# Finally, add the following lines to your ~/.config/nvim/init.vim:
#
#   inoremap <C-G>  <C-O>:!whisper.nvim<CR><C-O>:let @a = system("cat /tmp/whisper.nvim \| tail -n 1 \| xargs -0 \| tr -d '\\n' \| sed -e 's/^[[:space:]]*//'")<CR><C-R>a
#   nnoremap <C-G>       :!whisper.nvim<CR>:let @a = system("cat /tmp/whisper.nvim \| tail -n 1 \| xargs -0 \| tr -d '\\n' \| sed -e 's/^[[:space:]]*//'")<CR>"ap
#   vnoremap <C-G> c<C-O>:!whisper.nvim<CR><C-O>:let @a = system("cat /tmp/whisper.nvim \| tail -n 1 \| xargs -0 \| tr -d '\\n' \| sed -e 's/^[[:space:]]*//'")<CR><C-R>a
#
# This allows you to press Ctrl-G in order to capture audio from the microphone and transcribe it.
# When you are done speaking - press Ctrl-C
#

# the Whisper model to use
model="base.en"

# export the path to the whisper.cpp repo in the WHISPER_CPP_HOME env variable
# https://github.com/ggerganov/whisper.cpp
cd ${WHISPER_CPP_HOME}

if [ ! -f ./stream ] ; then
    echo "whisper.nvim: the 'stream' executable was not found! WHISPER_CPP_HOME=${WHISPER_CPP_HOME}" > /tmp/whisper.nvim
    exit 1
fi

if [ ! -f ./models/ggml-${model}.bin ] ; then
    echo "whisper.nvim: the '$model' model was not found! WHISPER_CPP_HOME=${WHISPER_CPP_HOME}" > /tmp/whisper.nvim
    exit 2
fi

# fine-tune the parameters according to your machine specs
./stream -t 8 -m models/ggml-base.en.bin --step 350 --length 10000 -f /tmp/whisper.nvim 2> /dev/null

exit 0