Transformers v5 Gemma scaling adjustment

[jlarson4]

Description

• Bump pinned transformers version to 5.4.0
• Gemma models use Gemma3TextScaledWordEmbedding which scales embeddings by sqrt(d_model). Before transformers 4.54.0, this was not done by default. In newer versions, we were doubling the scaling. This removes our additional scaling and defers to HuggingFace
• Reverifying effected models – This change actually improved some of model's scores

Note

This is not technically a breaking change, but transformers is a significant dependency that I want to make sure we call out as being updated

Type of change

• Bug fix (non-breaking change which fixes an issue)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)

Checklist:

• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• I have not rewritten tests relating to key interfaces which would affect backward compatibility