feat: make torch index explicit to support grace-hopper/GH200/aarch64 #533
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
Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: Terry Kong <terryk@nvidia.com>
Signed-off-by: Terry Kong <terryk@nvidia.com>
parthchadha
approved these changes
Jun 24, 2025
xxman-google
pushed a commit
to xxman-google/NeMo-RL
that referenced
this pull request
Jun 25, 2025
…NVIDIA-NeMo#533) Signed-off-by: Terry Kong <terryk@nvidia.com> Signed-off-by: Xuehan Xiong <xxman@google.com>
Hello @terrykong! To support GRPO on GH, is the vllm's aarch64 wheel is the only blocker here? Just want to understand if it's only wheel availability or there is something more. We want to test this internally (nvidia) and hence want to clarify. |
therealnaveenkamal
pushed a commit
to therealnaveenkamal/RL
that referenced
this pull request
Jul 7, 2025
…NVIDIA-NeMo#533) Signed-off-by: Terry Kong <terryk@nvidia.com>
YzjiaoNvd
pushed a commit
to YzjiaoNvd/NeMo-RL
that referenced
this pull request
Jul 14, 2025
…NVIDIA-NeMo#533) Signed-off-by: Terry Kong <terryk@nvidia.com>
KiddoZhu
pushed a commit
that referenced
this pull request
Jul 28, 2025
…#533) Signed-off-by: Terry Kong <terryk@nvidia.com>
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Outdated suggestions cannot be applied.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.
Partially addresses #532
With this change, loops like SFT/DPO that only need torch (and not vllm), should work.
GRPO still needs more work since vllm needs to be manually compiled for aarch64 since they do not publish aarch64 wheels
Perf/convergence don't look to be affected by manually specifying this index, which pulls nvidia lib wheels for 12.8 instead of 12.6 (which appeared to be the default before specifying the index)
