[FIXBUG] Qwen3VL hallucinations without Contiguous on Torch.SDPA (vllm-project#27744)

Signed-off-by: JartX <sagformas@epdcenter.es>
Co-authored-by: Lukas Geiger <lukas.geiger94@gmail.com>

Author: 2 people

vllm/model_executor/models/qwen2_5_vl.py

@@ -428,6 +428,14 @@ def forward(
             )
         elif self.attn_backend == _Backend.TORCH_SDPA:
             # Execute attention entry by entry for speed & less VRAM.
+            from vllm.platforms import current_platform
+
+            # Never remove the next contiguous logic
+            # Without it, hallucinations occur with the backend
+            if current_platform.is_rocm():
+                q = q.contiguous()
+                k = k.contiguous()
+                v = v.contiguous()
             outputs = []
             for i in range(1, len(cu_seqlens)):
                 start_idx = cu_seqlens[i - 1]