I am getting the following error while running ./setup.sh. ERROR: Could not find a version that satisfies the requirement tensorflow<1.14,>=1.13 (from mycroft-precise==0.3.0) (from versions: none)

[mgrove6]

ERROR: Could not find a version that satisfies the requirement tensorflow<1.14,>=1.13 (from mycroft-precise==0.3.0) (from versions: none) ERROR: No matching distribution found for tensorflow<1.14,>=1.13 (from mycroft-precise==0.3.0)

[el-tocino]

try pip3 install tensorflow==1.11 Should show you the versions available if not, check if there's a 1.13 on your platform.

[docholiday]

I get the same error as mgrove6. The command by el-tocino gives me the same error, only for version 11.11 now:

ERROR: Could not find a version that satisfies the requirement tensorflow==11.11 (from versions: none) ERROR: No matching distribution found for tensorflow==11.11

[stratus-ss]

I have the same problem on Ubuntu 20.04

ERROR: Could not find a version that satisfies the requirement tensorflow<1.14,>=1.13 (from mycroft-precise==0.3.0) (from versions: 2.2.0rc1, 2.2.0rc2, 2.2.0rc3)
ERROR: No matching distribution found for tensorflow<1.14,>=1.13 (from mycroft-precise==0.3.0)

and a manual attempt

pip3 install tensorflow==11.11
ERROR: Could not find a version that satisfies the requirement tensorflow==11.11 (from versions: 2.2.0rc1, 2.2.0rc2, 2.2.0rc3)
ERROR: No matching distribution found for tensorflow==11.11

[janvda]

I am getting same problem when starting from docker application from python:3 in my dockerfile.

to fix this I have used following docker application from tensorflow/tensorflow:1.13.0rc2-py3

So my complete dockerfile looks like:

from tensorflow/tensorflow:1.13.0rc2-py3

WORKDIR /usr/src/app

RUN apt-get update && apt-get install -y  git sudo

RUN git clone https://github.com/mycroftai/mycroft-precise
WORKDIR /usr/src/app/mycroft-precise
RUN ./setup.sh

# replace below command by actual command you want to run 
CMD [ "sleep", "10000" ]

[forslund]

As far as I been able to determine this is an issue with python 3.8+ support. The tensorflow version used (needed?) isn't available if python > 3.7 causing this issue.

If you're running docker, try a python:3,7 version of the container and see if that helps.

There has been a lot of work to update the TF version in PR #141, and I think that should work with newer pythons.

[0vermind]

This issue is still there. Ubuntu 20.04.1 Linux T14 5.4.0-42-generic #46-Ubuntu x86_64 x86_64 x86_64 GNU/Linux Python 3.8.5 pip 20.2.3 mycroft-core from git repo trying to install precise from git as per documentation. Are the docs lagging behind the development?

[forslund]

No as stated above the problem is python versions...The current combo of packages aren't available for python 3.8. Try the PR for updating tensorflow which should work.

[utaccn]

Hi, I was getting the same error, so I have tried to download the PR 141, that @forslund was suggesting. The downloading worked fine, but now, when I run the training with precise-train, I got the following error:

Traceback (most recent call last): File "/home/utakkn/mycroft-precise/.venv/bin/precise-train", line 33, in sys.exit(load_entry_point('mycroft-precise', 'console_scripts', 'precise-train')()) File "/home/utakkn/mycroft-precise/.venv/bin/precise-train", line 25, in importlib_load_entry_point return next(matches).load() File "/usr/lib/python3.8/importlib/metadata.py", line 77, in load module = import_module(match.group('module')) File "/usr/lib/python3.8/importlib/init.py", line 127, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "", line 1014, in _gcd_import File "", line 991, in _find_and_load File "", line 975, in _find_and_load_unlocked File "", line 671, in _load_unlocked File "", line 783, in exec_module File "", line 219, in _call_with_frames_removed File "/home/utakkn/mycroft-precise/precise/scripts/train.py", line 16, in from tensorflow.keras.callbacks import LambdaCallback File "/home/utakkn/mycroft-precise/.venv/lib/python3.8/site-packages/tensorflow/init.py", line 41, in from tensorflow.python.tools import module_util as _module_util File "/home/utakkn/mycroft-precise/.venv/lib/python3.8/site-packages/tensorflow/python/init.py", line 50, in from tensorflow.python import pywrap_tensorflow ImportError: cannot import name 'pywrap_tensorflow' from partially initialized module 'tensorflow.python' (most likely due to a circular import) (/home/utakkn/mycroft-precise/.venv/lib/python3.8/site-packages/tensorflow/python/init.py)

Do you think it is a problem with tensorflow and python3.8?

[MatthewScholefield]

The internet suggests you can solve your problem by running pip install tensorflow --upgrade --force-reinstall inside the venv (after running source .venv/bin/activate). For reference, I just searched "from tensorflow.python import pywrap_tensorflow ImportError: cannot import name" on Google.

[utaccn]

Thanks for the fast reply. I have done that but still seems to not be working. I get the following error;

tensorflow 2.3.1 requires tensorboard<3,>=2.3.0, but you'll have tensorboard 2.2.0 which is incompatible. tensorflow-gpu 2.2.0 requires tensorflow-estimator<2.3.0,>=2.2.0, but you'll have tensorflow-estimator 2.3.0 which is incompatible. tensorflow-cpu 2.2.0 requires tensorflow-estimator<2.3.0,>=2.2.0, but you'll have tensorflow-estimator 2.3.0 which is incompatible.

I then tried to install different versions of tensorflow without fixing my problem. The version 2.2.0 and 2.2.0rc2 seems to be working, but they raise me an error saying;

Illegal instruction (core dumped)

Which seems that my CPU can not compute this thing. I'm really stuck here now.

[modern-online]

I had the same problem (and I believe this thread provided a theoretical solution without much detail for me, a noob).

Anyway, my solution (Ubuntu20.04) was to install Python3.6 (3.7 also supports TF 1.x apparently): $ add-apt-repository ppa:deadsnakes/ppa $ apt-get update $ apt-get install python3.6

Install dev tools for the fresh python version (otherwise building pyaudio will fail): $ apt-get install python3.6-dev

After cloning the 'Precise' repo, edit setup.sh script: line51: python3 -m venv "$VENV" --without pip; change to python3.6 -m venv "$VENV" --without pip;

This will initialize virtual environment with the correct python version and allow installing older tensorflow. Hopefully they add TF2 support soon!

[wjz27]

I met same issue using setup.sh to build.

But directly using the last few lines inside setup.sh could run pass.

    pip install -e runner/
    pip install -e .

Could see the requirements are actually met:

Requirement already satisfied: tensorflow-estimator<1.14.0rc0,>=1.13.0 in /usr/local/lib/python3.7/site-packages (from tensorflow<1.14,>=1.13->mycroft-precise==0.3.0) (1.13.0)