Fix missing propagation of RoPE scaling and interpolation arguments in model builders

[taehwakkwon]

What does this PR do ?

This PR ensures that seq_len_interpolation_factor and rope_scaling_factor arguments are correctly propagated from the arguments to the model builder functions.

Previously, even if the user specified --seq-len-interpolation-factor or --rope-scaling-factor, these values were not being passed to the underlying model initialization logic.

Changes:

1. GPT (gpt_builders.py): Added propagation for both rope_scaling_factor and seq_len_interpolation_factor.
2. Mamba, BERT, T5: Added propagation for seq_len_interpolation_factor.

This fix allows RoPE scaling and position embedding interpolation to work as expected when configured via command-line arguments.

Contribution process

Pre-checks

• I want this PR in a versioned release and have added the appropriate Milestone (e.g., Core 0.8)
• I have added relevant unit tests
• I have added relevant functional tests
• I have added proper typing to my code Typing guidelines
• I have added relevant documentation
• I have run the autoformatter.sh on my PR

Code review

The following process is enforced via the CODEOWNERS file for changes into megatron/core. For changes outside of megatron/core, it is up to the PR author whether or not to tag the Final Reviewer team.

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.