Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Building on apple silicon #3

Open
peichins opened this issue Dec 4, 2023 · 0 comments
Open

Building on apple silicon #3

peichins opened this issue Dec 4, 2023 · 0 comments

Comments

@peichins
Copy link
Member

peichins commented Dec 4, 2023

Currently building the docker image on mac M1 is failing during the poetry install.

  1. Simple docker build -t cr01 .
  • This stops outputting the poetry install stage, due to [output clipped, log limit 2MiB reached]. I believe it's still working in the background, it just doesn't show the output. This continued for 40 mins or so, until it stopped with executor failed running [/bin/sh -c poetry install --no-interaction --no-ansi -vvv]: exit code: 137, which is apparently the container stopping due to high memory usage.
  1. Tried export DOCKER_BUILDKIT=0 then docker build --progress plain -t cr03 . > output01.txt
  • This was recommended to fix the "output clipped" problem, but probably won't fix the problem of the build never finishing.
  • appears that the output stopped being added to that file after about 40 mins with no indication that it had died, but after waiting a further 20 mins with no update to the file, decided that had also died
  1. Tried commenting the poetry install from the Dockerfile, then installing poetry in an interactive container.
  • this ran for ages then finally Resolving dependencies... (4089.7s)Killed, meaning it tried to resolve dependencies for over an hour then died, with no more error message than that. Tensorflow is not installed after this.
  1. Also tried just poetry install into a conda venv on the m1 without docker, unsuccessfully
  • tensorflow-io seems to be the only one that fails, can't find compatible version, but then anything tensorflow related is not installed. Tried going down the rabbithole of instructions for getting tensorflow working on M1 but it seems very horrible. Didn't find anything specific to the situation of using conda to install specific versions of python then poetry to install dependencies into the conda environment.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant