[AMD][Quantization] Add int4fp8_moe online quantization on ROCm (#7392)

Co-authored-by: Dehua Tang <dehtang@amd.com>
Co-authored-by: HAI <hixiao@gmail.com>
Co-authored-by: YC Tseng <yctseng@amd.com>

test/srt/run_suite.py

@@ -99,6 +99,8 @@ suite_amd = {
         # TestFile("test_no_overlap_scheduler.py", 234), # Disabled temporarily and track in #7703
         # TestFile("test_vision_chunked_prefill.py", 175), # Disabled temporarily and track in #7701
         # TestFile("test_wave_attention_backend.py", 150), # Disabled temporarily, see https://github.com/sgl-project/sglang/issues/11127
+        # The time estimation for `test_int4fp8_moe.py` assumes `mistralai/Mixtral-8x7B-Instruct-v0.1` is already cached (running on 1xMI300X).
+        TestFile("test_int4fp8_moe.py", 313),
     ],
     "per-commit-4-gpu-amd": [
         TestFile("test_pp_single_node.py", 150),