-
Notifications
You must be signed in to change notification settings - Fork 7
/
voice-requirements.txt
25 lines (22 loc) · 826 Bytes
/
voice-requirements.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
# Audio related
openwakeword
pyaudiowpatch
pyaudio
webrtcvad
# Install whisper from github:
# pip install --upgrade --no-deps --force-reinstall git+https://github.com/openai/whisper.git
# openai-whisper
more-itertools
sound
TTS
# Required for AWS Polly
boto3
pydub
SpeechRecognition
# torch
# To install torch for GPU (CUDA 12.1), use the following command:
# pip3 install --force-reinstall --pre torch torchvision torchaudio --index-url https://download.pytorch.org/whl/nightly/cu121
# ^^ The force reinstall is needed because the version of torch that gets installed as a dependency in other places is not correct
# NOTE: The pip install of ffmpeg-python is broken, so you need to install ffmpeg manually
# After you install ffmpeg (download from the website), you need to add the bin folder to your path
# ffmpeg-python