[Feature] Support for Ascend NPU backend (#3853)

Signed-off-by: Song Zhang <gepin.zs@antgroup.com> Co-authored-by: 22dimensions <waitingwind@foxmail.com>
2025-05-07 11:32:53 +08:00
parent cb69194562
commit 00c2c1f08b
7 changed files with 92 additions and 3 deletions
--- a/python/sglang/srt/server_args.py
+++ b/python/sglang/srt/server_args.py
@@ -555,7 +555,7 @@ class ServerArgs:
            "--device",
            type=str,
            default=ServerArgs.device,
-            help="The device to use ('cuda', 'xpu', 'hpu', 'cpu'). Defaults to auto-detection if not specified.",
+            help="The device to use ('cuda', 'xpu', 'hpu', 'npu', 'cpu'). Defaults to auto-detection if not specified.",
        )
        parser.add_argument(
            "--served-model-name",