docker pull failed due to a package unzip error

#1
by jietian1 - opened

tried with command:docker run -it -p 7860:7860 --platform=linux/amd64 registry.hf.space/k2-fsa-automatic-speech-recognition:latest python app.py
but got error, tried both under windows and macos

image.png

could you write a dockerfile to build an image on your own? We have never tried to use a docker image provided by huggingface.
We can help you if you have any issues about writing your own dockerfile.

ok. I am trying. I supposed the docker file and command is provided by dev team.
thanks for replying.

by the way, please ask in our wechat group or qq group to get a faster reply.

csukuangfj changed discussion status to closed

sorry I am new here, is there any instructions for setup the environment?

my step:

conda create -n asr python=10
pip install -r requirement.txt
python app.py

currently I am using Mac M1, the requirement.txt reports an error that torch you defined in it is not supported on my platform.

I am continue to manually install torch now.

sorry I am new here, is there any instructions for setup the environment?

If you only want to use the pre-trained models for speech recognition and if you are not going to train your own model, then
please follow
https://k2-fsa.github.io/sherpa/onnx/index.html
to setup your environment. You don't need to install k2 or pytorch to use sherpa-onnx.

Basically, if you only want use Python, then all you need is to follow
https://k2-fsa.github.io/sherpa/onnx/python/install.html#method-3-for-developers
to install sherpa-onnx

csukuangfj changed discussion status to open
csukuangfj changed discussion status to closed

Sign up or log in to comment