[Feat][NVFP4] Enable NVFP4 MoE for Qwen series models (eg. Qwen3-Next) #13761 (#13761)

Co-authored-by: Kaixi Hou <kaixih@nvidia.com>
2025-11-26 17:53:45 -07:00
parent 231df4b0d4
commit 91e8dc371a
4 changed files with 77 additions and 5 deletions
@@ -24,6 +24,7 @@ suites = {
        TestFile("test_flashinfer_trtllm_gen_attn_backend.py", 300),
        TestFile("test_deepseek_v3_fp4_cutlass_moe.py", 900),
        TestFile("test_fp4_moe.py", 300),
+        TestFile("test_qwen3_fp4_trtllm_gen_moe.py", 300),
    ],
    "nightly-8-gpu-b200": [
        TestFile("test_deepseek_r1_fp8_trtllm_backend.py", 3600),