[TP] Fix parameter detection issue and some invalid TP-plans #42129
+124
−93
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
Collaborator
ArthurZucker
left a comment
ArthurZucker
left a comment
There was a problem hiding this comment.
Missing a big test ! mega important!
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Cyrilvallez
changed the title
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: apertus, aria, deepseek_v2, doge, flex_olmo, gemma3, glm4v_moe, qwen3_next, t5gemma |
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Outdated suggestions cannot be applied.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.
What does this PR do?
Fixes #42016. Currently, we replace any number in a parameter name with a
*to match the tp_plan. However, params may contain numbers naturally in their name, such as Mixtral's MLPs containingw1,w2andw3param names. This means that they were replaced, and thus not detected as needing sharding. So basically they would be invisible from the tp_plan and never sharded 🥵🥵The regex now checks for
.(dots) on each side of the numbers, to only match layers or list names as we need.Also fixes a few TP-plans, that were detected wrong by the test 🙃🙃