Ant-Brain / EfficientWord-Net

OneShot Learning-based hotword detection.
https://ant-brain.github.io/EfficientWord-Net/
Apache License 2.0
215 stars 34 forks source link

Problem with Dependencies #Docker Support #6

Open Balro76 opened 2 years ago

Balro76 commented 2 years ago

Hello I left a comment on Reddit saying I would give it a go, and you said if I had a problem to log it here, so here I am, with a problem 😊

I seem to get stuck with pip3 install librosa I get this error Failed building wheel for llvmlite Running setup.py clean for llvmlite Failed to build llvmlite I can push on and get EfficientWord installed and working, if I say Alexa it says Yup I hear ya

The problem is then when I try to create my own wake word I run this command … python3 -m eff_word_net.generate_reference pi@raspberrypi:~ $ python3 -m eff_word_net.generate_reference Paste Path of folder Containing audio files:/home/pi/wakewords Paste Path of location to save *_ref.json :/home/pi/wakewords Enter Wakeword Name :bender Traceback (most recent call last): File "/usr/lib/python3.7/runpy.py", line 193, in _run_module_as_main "main", mod_spec) File "/usr/lib/python3.7/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/pi/.local/lib/python3.7/site-packages/eff_word_net/generate_reference.py", line 80, in input("Enter Wakeword Name :") File "/home/pi/.local/lib/python3.7/site-packages/eff_word_net/generate_reference.py", line 47, in generate_referencefile x, = librosa.load(audio_file,sr=16000) AttributeError: module 'librosa' has no attribute 'load'

My Problem is with librosa, I am not able to install it. I tried everything I could google but it will never install

How did you get around this problem ?

TheSeriousProgrammer commented 2 years ago

I see that you are using a raspberry pi

Consider generating the wakeword refernce file in x86_64 machine , cause librosa binaries are available for those machines making installations easier , then copy the wakeword reference file to your pi to start using it

generating wakeword reference file alone requires librosa(i.e to read the audio files) , currently working on a an update to remove librosa and use "wave" library in python

Sorry for the late reply though

preachwebsite commented 2 years ago

sudo+ssh://pi@192.168.0.41:22/usr/bin/python -u /home/pi/test.py Traceback (most recent call last): File "/home/pi/test.py", line 2, in from eff_word_net.streams import SimpleMicStream File "/home/pi/eff_word_net/streams.py", line 1, in import pyaudio ModuleNotFoundError: No module named 'pyaudio'

How to solve this problem? And I can't install librosa either

aman-17 commented 2 years ago

You need to install a few pre-requisites which are mentioned in the read.me, pyaudio is one of them :)