Support Llama4 fp8 inference (#5194)

Co-authored-by: laixinn <xielx@shanghaitech.edu.cn> Co-authored-by: sleepcoo <sleepcoo@gmail.com> Co-authored-by: zhyncs <me@zhyncs.com>
2025-04-09 20:14:34 +08:00
parent 86a876d883
commit 4065248214
14 changed files with 537 additions and 106 deletions
--- a/test/srt/run_suite.py
+++ b/test/srt/run_suite.py
@@ -76,6 +76,7 @@ suites = {
        TestFile("test_create_kvindices.py", 2),
        TestFile("test_hicache.py", 60),
        TestFile("test_hicache_mla.py", 90),
+        TestFile("test_triton_moe_channel_fp8_kernel.py", 25),
    ],
    "per-commit-2-gpu": [
        TestFile("models/lora/test_lora_tp.py", 300),