From d06e52a36933ca66436088ebfe3a17f8ec9611c4 Mon Sep 17 00:00:00 2001
From: one_east <wanyidong@huawei.com>
Date: Fri, 7 Nov 2025 11:03:29 +0800
Subject: [PATCH] bugfix: np.concatenat error

---
 tests/st/python/test_multi_modal_inputs.py | 81 ++++++++++++++++++++++
 vllm_mindspore/multimodal/inputs.py        |  2 +-
 2 files changed, 82 insertions(+), 1 deletion(-)
 create mode 100644 tests/st/python/test_multi_modal_inputs.py

diff --git a/tests/st/python/test_multi_modal_inputs.py b/tests/st/python/test_multi_modal_inputs.py
new file mode 100644
index 000000000..b0ac24ee8
--- /dev/null
+++ b/tests/st/python/test_multi_modal_inputs.py
@@ -0,0 +1,81 @@
+# SPDX-License-Identifier: Apache-2.0
+
+# Copyright 2025 Huawei Technologies Co., Ltd.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Test multi-modal inputs function."""
+
+# isort: skip_file
+
+import vllm_mindspore
+
+from dataclasses import dataclass
+from typing import Union
+import mindspore
+import torch
+import pytest
+import numpy as np
+from vllm.multimodal.inputs import MultiModalFlatField, NestedTensors
+
+
+@dataclass(frozen=True)
+class MockMultiModalFlatField(MultiModalFlatField):
+    slices: Union[list[slice], list[list[slice]]]
+    dim: int = 0
+
+
+@pytest.mark.level0
+@pytest.mark.platform_arm_ascend910b_training
+@pytest.mark.env_onecard
+def test_reduce_data_with_torch_tensors():
+    batch = [torch.randn(2, 3, 4), torch.randn(2, 3, 4), torch.randn(2, 3, 4)]
+    field = MockMultiModalFlatField(slices=[slice(None)], dim=1)
+    result = field._reduce_data(batch)
+    assert isinstance(result, mindspore.Tensor)
+    assert result.shape == (2, 9, 4)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_arm_ascend910b_training
+@pytest.mark.env_onecard
+def test_reduce_data_with_single_tensor():
+    batch = [torch.randn(2, 3, 4)]
+    field = MockMultiModalFlatField(slices=[slice(None)], dim=2)
+    result = field._reduce_data(batch)
+    assert isinstance(result, mindspore.Tensor)
+    assert result.shape == (2, 3, 4)
+
+
+@pytest.mark.level0
+@pytest.mark.platform_arm_ascend910b_training
+@pytest.mark.env_onecard
+def test_reduce_data_with_nested_tensors():
+    first_tensor = torch.randn(3, 5)
+    second_tensor = torch.randn(3, 5)
+    batch = (first_tensor, second_tensor)
+    field = MockMultiModalFlatField(slices=[slice(None)], dim=0)
+    result = field._reduce_data(batch)
+    assert len(result) == 6
+    assert result[0].size(0) == 5
+
+
+@pytest.mark.level0
+@pytest.mark.platform_arm_ascend910b_training
+@pytest.mark.env_onecard
+def test_reduce_data_with_mismatched_dim():
+    first_tensor = torch.randn(3, 5)
+    second_tensor = torch.randn(3, 5)
+    batch = (first_tensor, second_tensor)
+    field = MockMultiModalFlatField(slices=[slice(None)], dim=1)
+    with pytest.raises(AssertionError):
+        field._reduce_data(batch)
diff --git a/vllm_mindspore/multimodal/inputs.py b/vllm_mindspore/multimodal/inputs.py
index 7b7436ab6..bf3509516 100644
--- a/vllm_mindspore/multimodal/inputs.py
+++ b/vllm_mindspore/multimodal/inputs.py
@@ -157,7 +157,7 @@ def flat_reduce_data(self, batch: list[NestedTensors]) -> NestedTensors:
 
         if all(_expect_same_shape(elem) == first_shape for elem in batch):
             return mindspore.from_numpy(
-                np.concatenat([b.numpy() for b in batch], axis=self.dim))
+                np.concatenate([b.numpy() for b in batch], axis=self.dim))
 
     assert self.dim == 0, "dim == 0 is required for nested list"
     return [e for elem in batch for e in elem]
-- 
Gitee