Tensorflow and Horovod tests #106
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…, threads, and uses the parameter builtin method of ReFrame
…ince it also applies to TensorFlow-only tests
…t should only launch a single task
…in their ReFrame config if they need to specify additional flags in order to request a certain number of GPUs per node. This may depend on the allocation manager and its default configuration
…iable performance numbers due to less context switching with the Horovod thread. Additionally, for the Horovod test, only require the Horovod module - TensorFlow will be loaded as a dep automatically anyway.
|
This PR is now outdated and is superseded by #122 where I recreated the TensorFlow and Horovod tests with ReFrame's library of tests pattern |
trz42
added a commit
to trz42/software-layer
that referenced
this pull request
May 10, 2023
…-Rust/1.52.1-by-Rust/1.60.0 Nessi.no 2023.04 replace Rust/1.52.1 by Rust/1.60.0
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This ReFrame test contains TensorFlow and Horovod tests for both CPU and GPU. In EESSI, we don't have GPU-based modules yet, nor do we have Horovod, but the tests have been checked using a local module environment. Once we have GPUs/Horovod in EESSI, these test can be used right away. Until that time, tags can be used to select a subset:
-n TensorFlow2Nativeto only select the TensorFlow tests (no Horovod)-t cputo only select the cpu based tests(note that you can combine both arguments)
In order to run the test, run e.g. (in the
tests/reframedirectory):where your
settings.pyat least specifies a partition that hasnum_cpusdefinedfossas a valid programming environmentenvironmentsdictionaryIf your only valid programming environment in your
settings.pyisfoss, you may ommit--prgenv=fossfrom the run command