[lidanserj]

Hey team! Found and fixed dead link in src/transformers/models/mamba/modeling_mamba.py

https://github.com/NVIDIA/Megatron-LM/blob/main/megatron/model/gpt_model.py - old
https://github.com/NVIDIA/Megatron-LM/blob/main/megatron/legacy/model/gpt_model.py - new

[github-actions]

[For maintainers] Suggested jobs to run (before merge)

run-slow: mamba

[Rocketknight1]

Yup, thank you for the fix!

[Rocketknight1]

@lidanserj some models inherit from mamba via modular - you might need to change those too or run make fix-copies locally to get the tests to pass!

[lidanserj]

@Rocketknight1 should I wait for your suggestions, or can I get started?

[Rocketknight1]

@lidanserj no suggestions - but you do need to run make fix-copies or something like that to get the CI to pass. Once it's green, you can ping me and I'll merge this!

[ArthurZucker]

Thanks for the PR, and sorry it didn't make it in! This branch is based on v4, and v5 is very different, so rather than rebasing we'd love to invite you to open a new PR against main using our latest tools and good practices:

• Transformers tenets
• mlinter and the modeling rules it enforces (make typing)
• Modular transformers
• Contributing guide

Closing this one in the meantime, thanks again!