[2/N]Support DeepSeek-R1 w4a8 low latency deepep (#8464)

Co-authored-by: Hank Han <hanhan7630@outlook.com> Co-authored-by: Shangchuan Huang <2510421000@qq.com>
2025-10-25 08:41:16 +08:00
parent e51046beaa
commit 13bf565d60
8 changed files with 531 additions and 9 deletions
--- a/test/srt/run_suite.py
+++ b/test/srt/run_suite.py
@@ -172,7 +172,7 @@ suites = {
        TestFile("test_disaggregation_hybrid_attention.py", 200),
    ],
    "per-commit-8-gpu-h20": [
-        TestFile("quant/test_w4a8_deepseek_v3.py", 371),
+        TestFile("quant/test_w4a8_deepseek_v3.py", 520),
        TestFile("test_disaggregation_different_tp.py", 600),
        TestFile("test_disaggregation_pp.py", 140),
    ],