search

tensorflow / tfjs

A WebGL accelerated JavaScript library for training and deploying ML models.
https://js.tensorflow.org
Apache License 2.0
18.51k stars 1.93k forks source link

Converter python package installs tensorflow-cpu package on top of already installed tensorflow package #3953

Closed Ark-kun closed 4 years ago

Ark-kun commented 4 years ago

Please make sure that this is a build/installation issue. As per our GitHub Policy, we only address code/doc bugs, performance issues, feature requests and build/installation issues on GitHub. tag:build_template

System information

Describe the problem

Converter python package installs tensorflow-cpu package on top of already installed tensorflow package

Provide the exact sequence of commands / steps that you executed before running into the problem

$ docker run tensorflow/tensorflow:2.3.0 pip3 install tensorflowjs
...
Collecting tensorflow-cpu<3,>=2.1.0
  Downloading tensorflow_cpu-2.3.0-cp36-cp36m-manylinux2010_x86_64.whl (137.3 MB)

The official Tensorflow documentation does not seem to mention the tensorflow-cpu package. It mentions the tensorflow package instead. Maybe tensorlfowjs can depend on it instead.

rthadur commented 4 years ago

@Ark-kun tensorflow-cpu and tensorflow are both same, I believe the output has been printed in that way. cc @pyu10055 can you please correct me if i am wrong ? Thank you

pyu10055 commented 4 years ago

@Ark-kun we recommend installing tensorflowjs pip in a new virtualenv, please refer the converter readme file for details.

Ark-kun commented 4 years ago

@Ark-kun tensorflow-cpu and tensorflow are both same, I believe the output has been printed in that way.

Yes, they seem to be the same. My problem is that in my case TFJS installation downloads and installs a big dependency which is unnecessary since it's already installed. It can be a bit of a problem for us since we dynamically install TFJS every time and want to minimize the amount of dynamically installed packages. I'd understand if it was the tensorflow vs tensorflow-gpu which is a problem with TF, not TFJS. But tensorflow-cpu vs. tensorflow seems to me more TFJS converter setup related.

@Ark-kun we recommend installing tensorflowjs pip in a new virtualenv, please refer the converter readme file for details.

I'm running TensorflowJS inside the official Tensorflow Docker containers.

I do not want to take time from the actual TFJS development. Just thought that maybe there is some explanation for the choice.

google-ml-butler[bot] commented 4 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in 7 dyas if no further activity occurs. Thank you.

google-ml-butler[bot] commented 4 years ago

Closing as stale. Please @mention us if this needs more attention.

google-ml-butler[bot] commented 4 years ago

Are you satisfied with the resolution of your issue? Yes No