Closed
Conversation
feat: qwen3 next fix: qwen3next gated conv weight split tp feat: qwen3next mcore to hf remove useless log
Owner
|
Thanks for your support, but this project is being deprecated, I sincerely recommend you to take a look at the Megatron-Bridge which is the official version with long term maintenance. |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR provides the bridge for qwen3-next. It needs to be used in conjunction with the Mcore PR(NVIDIA/Megatron-LM#1958) to ensure the correctness of delta_net out_norm.
I have already aligned the forward logits between the hf model and mcore model.