From 9dd02d85ca801c99241317a8061bd025c726af93 Mon Sep 17 00:00:00 2001
From: Siyuan Li <94890248+liaoyanqing666@users.noreply.github.com>
Date: Mon, 13 Jan 2025 14:24:10 +0800
Subject: [PATCH] [Bug] Fix usage of `.transpose()` and `.view()`
 consecutively. (#11979)

---
 vllm/attention/layer.py                  | 2 +-
 vllm/model_executor/models/intern_vit.py | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/vllm/attention/layer.py b/vllm/attention/layer.py
index 55e4e14027f79..b8afd428f2cc0 100644
--- a/vllm/attention/layer.py
+++ b/vllm/attention/layer.py
@@ -230,7 +230,7 @@ def forward(
                                                  value,
                                                  scale=self.scale)
             out = out.transpose(1, 2)
-        return out.view(bsz, q_len, -1)
+        return out.reshape(bsz, q_len, -1)
 
 
 def unified_attention(
diff --git a/vllm/model_executor/models/intern_vit.py b/vllm/model_executor/models/intern_vit.py
index 7ff68bd60e8ad..8ad009d5101e4 100644
--- a/vllm/model_executor/models/intern_vit.py
+++ b/vllm/model_executor/models/intern_vit.py
@@ -271,7 +271,7 @@ def forward(self, x: torch.Tensor) -> torch.Tensor:
         v = v.transpose(1, 2)
 
         x = F.scaled_dot_product_attention(q, k, v, scale=self.scale)
-        x = x.transpose(1, 2).view(B, N, -1)
+        x = x.transpose(1, 2).reshape(B, N, -1)
 
         x = self.proj(x)
         return x