[Quantization] Support Quark Dense + MoE FP8 & FP8 PTPC (#10485)

Co-authored-by: HAI <hixiao@gmail.com> Co-authored-by: kk <43161300+kkHuang-amd@users.noreply.github.com>
2025-11-13 08:16:00 -08:00
parent e7e89349c9
commit 67e9d287ee
10 changed files with 666 additions and 243 deletions
--- a/python/sglang/srt/configs/model_config.py
+++ b/python/sglang/srt/configs/model_config.py
@@ -663,6 +663,7 @@ class ModelConfig:
            "qoq",
            "w4afp8",
            "petit_nvfp4",
+            "quark",
        ]
        compatible_quantization_methods = {
            "modelopt_fp8": ["modelopt"],