From c10eb811fac677ee2d7c0c38a44d084fcc613cdc Mon Sep 17 00:00:00 2001
From: VED <146507396+ved1beta@users.noreply.github.com>
Date: Mon, 18 Aug 2025 18:14:37 +0530
Subject: [PATCH] data_parallel_size in in VllmserveCliArgs (#3074)

* data_parallel_size in in VllmserveCliArgs

* moved to 43
---
 src/axolotl/cli/args.py | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/src/axolotl/cli/args.py b/src/axolotl/cli/args.py
index 31d854d41..9bb544aff 100644
--- a/src/axolotl/cli/args.py
+++ b/src/axolotl/cli/args.py
@@ -40,6 +40,12 @@ class VllmServeCliArgs:
         default=None,
         metadata={"help": "Number of tensor parallel workers to use."},
     )
+    data_parallel_size: Optional[int] = field(
+        default=None,
+        metadata={
+            "help": "Number of data parallel workers to use for vLLM serving. This controls how many model replicas are used for parallel inference."
+        },
+    )
     host: Optional[str] = field(
         default=None,  # nosec B104
         metadata={"help": "Host address to run the server on."},