update cuda version in local build #9736
Conversation
|
rerun tests |
Codecov Report
@@ Coverage Diff @@
## branch-22.02 #9736 +/- ##
=============================================
Coverage 10.49% 10.49%
=============================================
Files 119 119
Lines 20305 20305
=============================================
Hits 2130 2130
Misses 18175 18175 Continue to review full report at Codecov.
|
| @@ -18,12 +18,12 @@ Build and test your local repository using a base gpuCI Docker image | |||
| where: | |||
| -H Show this help text | |||
| -r Path to repository (defaults to working directory) | |||
| -i Use Docker image (default is gpuci/rapidsai:${NIGHTLY_VERSION}-cuda10.1-devel-ubuntu16.04-py3.7) | |||
| -i Use Docker image (default is gpuci/rapidsai:${NIGHTLY_VERSION}-cuda11.5-devel-ubuntu20.04-py3.8) | |||
There was a problem hiding this comment.
What is the difference between developing from gpuci/rapidsai and rapidsai/rapidsai-dev? It sounds like both are development containers, but I am not familiar with the differences or how to locate that information as a user/developer.
In general, we have a lot of similarly-named containers and I don't know how to identify the purpose of each. The DockerHub page for the gpuci/rapidsai container (https://hub.docker.com/r/gpuci/rapidsai) is slim on docs, but I don't know of a better source for this kind of information. The rapidsai/rapidsai-dev container has more documentation, which I found very valuable. Should this be raised as a separate issue?
There was a problem hiding this comment.
gpuci images are development images for continuous integration(CI).
gpuci docker images are supposed to be same environment as CI. If you want to run CI tests and build locally, these images could be used. These images are available during development cycle of each version of rapidsai.
rapidsai/rapidsai images are not supposed be used for development of rapidsai library itself. It is for development using rapidsai libraries. It contains rapidsai libraries installed. It's available after the each rapidsai release. Not during the development.
rapidsai dockers have 3 flavors: Base, Runtime, Devel. Devel is supposed to be for development using rapidsai libraries. Devel images are not upto date (last updated 2 years ago)
The RAPIDS images are based on nvidia/cuda, and are intended to be drop-in replacements for the corresponding CUDA images in order to make it easy to add RAPIDS libraries while maintaining support for existing CUDA applications.
There was a problem hiding this comment.
Not sure of the difference between gpuci/rapidsai and rapidsai/rapidsai-dev. It seems like they moved all devel containers from rapidsai to rapidsai-dev.
|
@gpucibot merge |
update cuda, ubuntu, python versions in local build using gpuci docker image.