Mistral Large 3 NVFP4 support

[dcampora]

Support Mistral Large 3 NVFP4.

Depends on #14466.

• GSM8K test results:

SGLANG_ENABLE_JIT_DEEPGEMM=0 \
python3 -m sglang.launch_server \
--model mistralai/Mistral-Large-3-675B-Instruct-2512-NVFP4 \
--kv-cache-dtype fp8_e4m3 \
--tensor-parallel-size 8 \
--disable-radix-cache \
--stream-interval 20 \
--mem-fraction-static 0.9 \
--attention-backend trtllm_mla \
--model-loader-extra-config '{"enable_multithread_load": true}' \
--max-running-requests 1024 \
--cuda-graph-max-bs 1024 \
--chat-template mistral

lm_eval \
--model local-chat-completions \
--model_args model=mistralai/Mistral-Large-3-675B-Instruct-2512-NVFP4,\
base_url=http://127.0.0.1:30000/v1/chat/completions,\
num_concurrent=128,timeout=999999,max_gen_toks=8192 \
--tasks gsm8k \
--batch_size 128 \
--apply_chat_template \
--num_fewshot 8

|Tasks|Version|     Filter     |n-shot|  Metric   |   |Value |   |Stderr|
|-----|------:|----------------|-----:|-----------|---|-----:|---|-----:|
|gsm8k|      3|flexible-extract|     8|exact_match|↑  |0.9249|±  |0.0073|
|     |       |strict-match    |     8|exact_match|↑  |0.7104|±  |0.0125|

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.
• Work with maintainers to merge your PR. See the PR Merge Process

[gemini-code-assist]

Warning

You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again!

[JustinTong0323]

/tag-and-rerun-ci

[elvischenv]

This is a mm op, why put under attention layer?

[elvischenv]

Have we tested the w4a16 code path? If not, better to do it in another PR. We may need it on Hopper or previous arch, and we don't have w4a16 moe support for now.

[elvischenv]

Same comment, can do in follow-up PR.

[elvischenv]

w4a4 supports both flashinfer and cutlass, right? I think we should do something similar to the below method, check the capability.

[elvischenv]

This seems to be only used by w4a16.

[elvischenv]

Add a comment to clarify this method is nvfp4 specific.