Split logical names in moe module

[NuojCheng]

Description

Split following logical names:

• embed_no_exp into: embed_no_exp, embed_moe
• activation_embed into: activation_embed, activation_embed_moe
• activation_norm_length into: activation_norm_length, activation_norm_length_moe
• activation_length_no_exp into: activation_length_no_exp, activation_length_no_exp_moe
• activation_batch into activation_batch, activation_batch_moe
• activation_batch_no_exp into activation_batch_no_exp, activation_batch_no_exp_moe

Tests

1. Losses and performance match using model_name=deepseek3-test on v5p-8 VM.

• Pure FSDP: https://diff.googleplex.com/#key=GrrJyIf9fENe
• FSDP+EP: https://diff.googleplex.com/#key=G1ZyP3HnDGbs
• FSDP+TP: https://diff.googleplex.com/#key=fdiAZWeyybSP
• FSDP+CP: https://diff.googleplex.com/#key=eRmw7tx4KDcs

2. V-llm test
   Vllm test: https://diff.googleplex.com/#key=q4eBkLazREaW
   Script: https://paste.googleplex.com/6329507825975296

3. 2d-fsdp test:
   2d-fsdp: https://diff.googleplex.com/#key=sZNNYqZFXNEu

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

❌ Patch coverage is 73.91304% with 6 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/maxtext/layers/moe.py	73.91%	6 Missing ⚠️

📢 Thoughts on this report? Let us know!

[suexu1025]

I think we don't need to update for use_2d_fsdp_sharding? embed_no_exp_moe was not defined in 2dfsdp yml.

[NuojCheng]

"embed_no_exp" shares same logical rule contents with "embed_no_exp_moe", see

maxtext/src/maxtext/configs/base.yml

Lines 489 to 500 in 935ffb4

	['embed_no_exp', ['fsdp', 'fsdp_transpose', 'sequence', 'tensor_transpose', 'context']],
	['embed_no_exp', ['fsdp', 'sequence', 'tensor_transpose', 'context']],
	['embed_no_exp', ['fsdp', 'fsdp_transpose', 'sequence', 'context']],
	['embed_no_exp', ['fsdp', 'sequence', 'context']],
	['embed_moe', ['fsdp', 'fsdp_transpose', 'sequence', 'tensor_transpose', 'context', 'expert']],
	['embed_moe', ['fsdp', 'sequence', 'tensor_transpose', 'context' , 'expert']],
	['embed_moe', ['fsdp', 'fsdp_transpose', 'sequence', 'context', 'expert']],
	['embed_moe', ['fsdp', 'sequence', 'context', 'expert']],
	['embed_no_exp_moe', ['fsdp', 'fsdp_transpose', 'sequence', 'tensor_transpose', 'context']],
	['embed_no_exp_moe', ['fsdp', 'sequence', 'tensor_transpose', 'context']],
	['embed_no_exp_moe', ['fsdp', 'fsdp_transpose', 'sequence', 'context']],
	['embed_no_exp_moe', ['fsdp', 'sequence', 'context']],

. It won't affect performance.

[suexu1025]

need some tests before change for all cases.

[NuojCheng]

need some tests before change for all cases.

I added tests in the PR description. Would you like to elaborate what tests you want to see?

[NuojCheng]

need some tests before change for all cases.

I added tests in the PR description. Would you like to elaborate what tests you want to see?

I added a 2d-fsdp test in PR description.