Skip to content

move pynvml install to setup.py - #5840

Merged
loadams merged 1 commit into
deepspeedai:masterfrom
ROCm:make_pynvml_optional
Aug 15, 2024
Merged

move pynvml install to setup.py#5840
loadams merged 1 commit into
deepspeedai:masterfrom
ROCm:make_pynvml_optional

Conversation

@Rohan138

@Rohan138 Rohan138 commented Aug 5, 2024

Copy link
Copy Markdown
Contributor

Only install pynvml on nvidia gpus; not all accelerators

@Rohan138
Rohan138 requested a review from loadams as a code owner August 5, 2024 23:49
@Rohan138

Rohan138 commented Aug 6, 2024

Copy link
Copy Markdown
Contributor Author

@microsoft-github-policy-service agree company="AMD"

@loadams

loadams commented Aug 14, 2024

Copy link
Copy Markdown
Collaborator

Hi @Rohan138 - we're not able to re-sync master with your branch, could you please do that, that way the tests will re-run with the new changes we have merged? Thanks!

@loadams
loadams enabled auto-merge August 15, 2024 15:45
@loadams
loadams added this pull request to the merge queue Aug 15, 2024
Merged via the queue into deepspeedai:master with commit 30428d0 Aug 15, 2024
@Rohan138
Rohan138 deleted the make_pynvml_optional branch August 15, 2024 20:55
weiji14 added a commit to regro-cf-autotick-bot/deepspeed-feedstock that referenced this pull request Aug 16, 2024
weiji14 added a commit to conda-forge/deepspeed-feedstock that referenced this pull request Aug 16, 2024
* updated v0.14.5

* Remove ninja as runtime dependency

Xref #1

* Require nvidia-ml-py only on CUDA builds

Xref deepspeedai/DeepSpeed#5840

---------

Co-authored-by: Wei Ji <[email protected]>
mauryaavinash95 pushed a commit to DataStates/DeepSpeed that referenced this pull request Mar 20, 2025
Only install pynvml on nvidia gpus; not all accelerators
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

Successfully merging this pull request may close these issues.

2 participants