Add llama3-1-405b 16node recipe on A4x #53
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
suffiank
reviewed
Nov 27, 2025
| value: VERSION | ||
| - name: NCCL_ALGO | ||
| value: "Ring,Tree" | ||
| - name: NCCL_NET_GDR_LEVEL |
There was a problem hiding this comment.
Where did these (NCCL variables) get inherited from? I think these are supposed to be set a script that the host networking team ships.
Collaborator
Author
There was a problem hiding this comment.
We have these set in our internal scripts, and the training fails without these due to communication errors. Looks like its not set by the host networking script?
suffiank
approved these changes
Nov 27, 2025
| targetNodepools: null | ||
| tasSettings: | ||
| topologyRequest: | ||
| kueue.x-k8s.io/podset-preferred-topology: kubernetes.io/hostname |
There was a problem hiding this comment.
Is this right? Does it need to require SBRG and then prefer a certain number of VMs within a rack?
| @@ -0,0 +1,207 @@ | |||
| """Nemo2 pretraining recipe for Llama 3.1 405B model.""" | |||
There was a problem hiding this comment.
Nothing in this script hard codes FP8-CS or GBS 2048, I think? So curious if that needs to be in the filename.
Collaborator
There was a problem hiding this comment.
It should follow the same format like
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
Adds the optimal helm chart for llama405b on A4X(GB200) machines