adapt w2 quant layer format for Minimax2.5

[shadowxz109]

Motivation

adapt w2 quant weight format for Minimax2.5

Modifications

add '.0.w2.weight' format in get_moe_scheme
prefix change mlp to block_sparse_moe

Accuracy Tests

Accuracy: 0.955
Invalid: 0.000
Latency: 38.473 s
Output throughput: 520.775 token/s

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.
• Follow the SGLang code style guidance.

Review Process

1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments or contact authorized users to do so.
   • /tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci
4. After green CI and required approvals, ask Merge Oncalls to merge.

[ascend-robot]

CLA Signature Guide

@shadowxz109 , thanks for your pull request.

The following commit(s) are not associated with a signed Contributor License Agreement (CLA).

Commit	Reason
53ac4615 adapt w2 quant weight format foe...	the email used in the commit is not linked to a signed CLA! please verify that it matches the email you used when signing the CLA.

To sign CLA, click here.

To check if your email is configured correctly, refer to the FAQs.

Once you've signed the CLA or updating your email, please comment /check-cla to revalidate CLA status.

[shadowxz109]

/check-cla

[ascend-robot]

CLA Signature Pass

shadowxz109, thanks for your pull request. All authors of the commits have signed the CLA. 👍

[iforgetmyname]

Suggested change

	prefix_in_quant_config_down = prefix + ".0.down_proj.weight"
	prefix_in_quant_config_w2 = prefix + ".0.w2.weight"
	prefix_in_quant_config_search_list = [".0.down_proj.weight", ".0.w2.weight"]
	quant_scheme_entry = None
	for prefix_in_quant_config in prefix_in_quant_config_search_list:
	if (quant_scheme_entry := self.quant_description.get(prefix_in_quant_config, None):
	break