enable auto-round quantization model (#6226)

Signed-off-by: Zhang, Weiwei1 <weiwei1.zhang@intel.com>
2025-09-08 13:05:35 +08:00
parent b67c277f86
commit c8295d2353
8 changed files with 528 additions and 0 deletions
--- a/python/sglang/srt/server_args.py
+++ b/python/sglang/srt/server_args.py
@@ -80,6 +80,7 @@ QUANTIZATION_CHOICES = [
    "qoq",
    "w4afp8",
    "mxfp4",
+    "auto-round",
 ]

 ATTENTION_BACKEND_CHOICES = [