I am trying to use the the DeepSpeech model 0.4.1 and I try to install ds_ctcdecoder. But I am getting HTTP error. I look into where ds_ctcdecoder is being downloaded from, and it seems to end up supporting 0.4.1 ?
I am using Ubuntu 16.04 and I use the following command to install
pip3 install $ (python3 util/taskcluster.py --decoder)
lissyx
((slow to reply) [NOT PROVIDING SUPPORT])
2
The artifact expired, please rebuild using make -C native_client/ctcdecoder
Thanks, that seems to work. And I also notice that
Copying ds_ctcdecoder.egg-info to temp_build/temp_build/ds_ctcdecoder-0.4.1-py2.7.egg-info
Does this mean that this only builds for Python2. Is there a way I can build for Python3?
lissyx
((slow to reply) [NOT PROVIDING SUPPORT])
4
This depends on the Python version you run the build with. So if you setup a virtualenv with a different python version, you should be able to produce for different versions of Python