try gpu on linux #22
Merged
try gpu on linux #22
Conversation
This file contains 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
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
…da-forge-pinning 2020.12.11.09.47.37
…da-forge-pinning 2020.12.11.09.47.37
…da-forge-pinning 2020.12.11.09.47.37
…da-forge-pinning 2020.12.11.09.47.37
…da-forge-pinning 2020.12.11.09.47.37
|
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( I do have some suggestions for making it better though... For recipe:
|
isuruf
commented
Dec 12, 2020
| - cffi | ||
| - future # [py==27] | ||
| outputs: | ||
| - name: pytorch |
|
@kkraus14, @jakirkham, can you test these on gpu hardware? |
Closed
|
@isuruf what is the best way to install the packages built from this PR for offline testing? |
|
CPU or the GPU packages? GPU packages currently time out, so you need to build it locally using |
GPU indeed. Okay, thanks - will try. |
rgommers
reviewed
Dec 15, 2020
| - magma # [cuda_compiler_version != "None"] | ||
| # other requirements | ||
| - python | ||
| - numpy |
There was a problem hiding this comment.
Seeing this build warning:
WARNING:conda_build.metadata:No numpy version specified in conda_build_config.yaml.
Falling back to default numpy value of 1.11
That's probably not desired, should be 1.16?
There was a problem hiding this comment.
i think that warning is benign
…da-forge-pinning 2020.12.15.10.41.23
|
Running |
|
Ah, you probably need |
Yep that was it. It's building now, but it looks like it only uses one CPU core. That may be my default Docker config, not sure (or is there a conda-forge specific config for this?). Usually when I build PyTorch on a machine with 12 physical cores, I set |
|
You can set |
|
This is looking promising. I tested the CUDA 10.2, Python 3.8 build. The majority of the test suite runs fine, it stops on the first encountered failure at: Which is a known issue, the test itself is wrong: pytorch/pytorch#41358. So so far so good. |
|
Since it took a bit of time to figure out the local build/install/test cycle, summarizing here: |
|
A simpler way would be |
|
@hmaarrfk, merge? @chrisburr offered to build the GPU variants locally and upload. |
|
Thanks for all the hard work! |
1 task
|
@isuruf Perhaps - like for the mutex naming (cf conda-forge/conda-forge.github.io#1059) - there should be a standard way to do this? Speaking of mutexes - if I understand correctly there's now no "proc" package for pytorch to select the type of installation? Does |
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.
Checklist
0(if the version changed)conda-smithy(Use the phrase@conda-forge-admin, please rerenderin a comment in this PR for automated rerendering)