From c330b687a116be3738120f430ddbb096ca9543a3 Mon Sep 17 00:00:00 2001
From: Xinyuan Tong <115166877+JustinTong0323@users.noreply.github.com>
Date: Sat, 14 Mar 2026 04:48:28 +0000
Subject: [PATCH] [Bugfix] Fix GLM-4.6V vision regression in glm4v_moe and
 glm_ocr (#20463)

---
 python/sglang/srt/models/glm4v_moe.py | 3 +++
 python/sglang/srt/models/glm_ocr.py   | 3 +++
 2 files changed, 6 insertions(+)

diff --git a/python/sglang/srt/models/glm4v_moe.py b/python/sglang/srt/models/glm4v_moe.py
index 324de18b4..0e77d5b3b 100644
--- a/python/sglang/srt/models/glm4v_moe.py
+++ b/python/sglang/srt/models/glm4v_moe.py
@@ -281,5 +281,8 @@ class Glm4vMoeForConditionalGeneration(Glm4vForConditionalGeneration):
                     else:
                         logger.warning(f"Parameter {name} not found in params_dict")
 
+        if not is_nextn:
+            self.visual.patch_embed.copy_conv3d_weight_to_linear()
+
 
 EntryClass = [Glm4vMoeForConditionalGeneration]
diff --git a/python/sglang/srt/models/glm_ocr.py b/python/sglang/srt/models/glm_ocr.py
index c2f4adc4c..ce1d300ab 100644
--- a/python/sglang/srt/models/glm_ocr.py
+++ b/python/sglang/srt/models/glm_ocr.py
@@ -431,5 +431,8 @@ class GlmOcrForConditionalGeneration(Glm4vForConditionalGeneration):
                     )
                 weight_loader(param, loaded_weight)
 
+        if not is_nextn:
+            self.visual.patch_embed.copy_conv3d_weight_to_linear()
+
 
 EntryClass = [GlmOcrForConditionalGeneration]