jina-ai / dalle-flow

🌊 A Human-in-the-Loop workflow for creating HD images from text
grpcs://dalle-flow.dev.jina.ai
2.83k stars 209 forks source link

Executor Failed Running (Exit Code 100) #29

Closed mattslight closed 2 years ago

mattslight commented 2 years ago

I get the following error when running docker build -t jinaai/dalle-flow .

I tried rm -rf on dalle-flow. I'm on a Silicon Mac (arm64) and suspect the packages are only available for amd64.

Anybody else had any luck building the docker container on Mac silicon?

#10 103.8 WARNING: apt does not have a stable CLI interface. Use with caution in scripts.
#10 103.8 
#10 103.8 Reading package lists...
#10 104.3 Building dependency tree...
#10 104.4 Reading state information...
#10 104.5 The following NEW packages will be installed:
#10 104.5   libcudnn8:amd64
#10 106.7 0 upgraded, 1 newly installed, 0 to remove and 27 not upgraded.
#10 106.7 Need to get 0 B/421 MB of archives.
#10 106.7 After this operation, 1253 MB of additional disk space will be used.
#10 106.7 Get:1 /dalle/libcudnn8_8.4.0.27-1+cuda11.6_amd64.deb libcudnn8 amd64 8.4.0.27-1+cuda11.6 [421 MB]
#10 106.7 debconf: delaying package configuration, since apt-utils is not installed
#10 106.8 dpkg: error processing archive /dalle/libcudnn8_8.4.0.27-1+cuda11.6_amd64.deb (--unpack):
#10 106.8  package architecture (amd64) does not match system (arm64)
#10 106.8 Errors were encountered while processing:
#10 106.8  /dalle/libcudnn8_8.4.0.27-1+cuda11.6_amd64.deb
#10 106.8 E: Sub-process /usr/bin/dpkg returned an error code (1)
------
executor failed running [/bin/sh -c wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/libcudnn8_8.4.0.27-1+cuda11.6_amd64.deb     && apt install ./libcudnn8_8.4.0.27-1+cuda11.6_amd64.deb]: exit code: 100
JoanFM commented 2 years ago

Hey @mattslight ,

This seems to be something related with your specific docker, maybe this comment https://github.com/docker-library/mongo/issues/276#issuecomment-644770237 could help you?

mattslight commented 2 years ago

Thanks - I managed to deploy direct to Azure, I don't think building the docker on an M1 Mac was a good idea of mine 😂