Unable to install DS_ctcdecoder

DeepSpeech
#1

I am unable to install ds_ctcdecoder in my system using the command
pip3 install $(python3 util/taskcluster.py --target ./ds_ctc)

i am getting the following statement as output…
Defaulting to user installation because normal site-packages is not writeable
ERROR: Invalid requirement: ‘…’

and when i run just partial command
python3 util/taskcluster.py --target ./ds_ctcdecode

i get the following output

Downloading https://community-tc.services.mozilla.com/api/index/v1/task/project.deepspeech.deepspeech.native_client.v0.7.1.cpu/artifacts/public/native_client.tar.xz …
Downloading: 100%

libdeepspeech.so
LICENSE
deepspeech
deepspeech.h
README.mozilla

Please help me how to proceed further and also how to get the .whl file to install ds_ctcdecoder

#2

None of your commands are expected to install this package. Please read the documentation and provide useful context. There’s nothing we can do here, you don’t share any errors, you don’t share your system’s informations.

#4

the below command give the following error.
pip3 install $(python3 util/taskcluster.py --decoder)
usage: taskcluster.py [-h] [–target TARGET] [–arch ARCH]
[–artifact ARTIFACT] [–source SOURCE]
[–branch BRANCH]
taskcluster.py: error: unrecognized arguments: --decoder
Defaulting to user installation because normal site-packages is not writeable
ERROR: You must give at least one requirement to install (see “pip help install”)

#5

what version are you on?

Le ven. 5 juin 2020 à 11:57, rahul.yarramreddy@aiedge.tech via Mozilla Discourse notifications@discourse.mozilla.org a écrit :

#7

I am using deepspeech v0.7.1

#8

Please follow the documentation, ds_ctcdecoder will be installed when you install the deepspeech_training package …

#9

I have followed the documentation step by step but when i tried to run
the following command

python3 DeepSpeech.py --train_files ./train.csv --dev_files ./dev.csv --test_files ./test.csv

i get the below error.

Training package is not installed. See training documentation.
Traceback (most recent call last):
  File "DeepSpeech.py", line 7, in <module>
    from deepspeech_training import train as ds_train
  File "/mnt/d/../deepspeech_training/train.py", line 29, in <module>
    from ds_ctcdecoder import ctc_beam_search_decoder, Scorer
ImportError: cannot import name 'ctc_beam_search_decoder'
#10

Have you ? Obviously, something went wrong. As you may not know, I’m not able to read in your mind, nor on your screen. So if you don’t share us install steps that you are supposed to have followed, and their outcome, how can you expect we can help you?

#11

I have followed the steps provided here.
https://deepspeech.readthedocs.io/en/v0.7.1/TRAINING.html

#12

That’s not helping here: the error you have proves something during those steps has not been done properly. Or that your context is different from our assumptions.

Since you refuse to share any informations, I can’t help you anymore.

#13

To clarify, you have to share the list of commands and the complete log (console output) of every command you ran, from cloning the repository, creating the virtualenv and following the steps on the documentation, if you want us to have any chance to help you.

#15

Hi reuben, lissyx.

When I re-installed my ubuntu Subsystem and ran the steps again … ds_ctcdecoder is installed properly. Thanks for the help.

1 Like