diff --git a/docker/rocm.Dockerfile b/docker/rocm.Dockerfile
index d196bf173..de05e97e4 100644
--- a/docker/rocm.Dockerfile
+++ b/docker/rocm.Dockerfile
@@ -31,7 +31,7 @@ ENV BUILD_TRITON="0"
 ENV BUILD_LLVM="0"
 ENV BUILD_AITER_ALL="1"
 ENV BUILD_MOONCAKE="1"
-ENV AITER_COMMIT="v0.1.7.post1"
+ENV AITER_COMMIT="v0.1.7.post5"
 ENV NO_DEPS_FLAG=""
 
 # ===============================
@@ -42,7 +42,7 @@ ENV BUILD_TRITON="0"
 ENV BUILD_LLVM="0"
 ENV BUILD_AITER_ALL="0"
 ENV BUILD_MOONCAKE="1"
-ENV AITER_COMMIT="v0.1.7.post2"
+ENV AITER_COMMIT="v0.1.7.post5"
 ENV NO_DEPS_FLAG=""
 # ===============================
 # Chosen arch and args
diff --git a/python/sglang/srt/models/deepseek_v2.py b/python/sglang/srt/models/deepseek_v2.py
index dc30b4f0a..9fdfbcb6a 100644
--- a/python/sglang/srt/models/deepseek_v2.py
+++ b/python/sglang/srt/models/deepseek_v2.py
@@ -1839,7 +1839,7 @@ class DeepseekV2AttentionMLA(nn.Module):
                 current_stream.wait_stream(self.alt_stream)
             else:
                 if _use_aiter_gfx95 and self.q_b_proj.weight.dtype == torch.uint8:
-                    q, k_nope, *_ = fused_rms_mxfp4_quant(
+                    q, _, k_nope, *_ = fused_rms_mxfp4_quant(
                         q,
                         self.q_a_layernorm.weight,
                         self.q_a_layernorm.variance_epsilon,