#13646: Add unit test for channel bcast using repeat

tests/ttnn/unit_tests/operations/eltwise/test_mul_channel.py

@@ -0,0 +1,27 @@
+# SPDX-FileCopyrightText: © 2023 Tenstorrent Inc.
+
+# SPDX-License-Identifier: Apache-2.0
+
+import pytest
+
+import torch
+
+import ttnn
+
+from tests.ttnn.utils_for_testing import assert_with_pcc
+from torch.nn import functional as F
+
+
+@pytest.mark.parametrize("h", [32])
+@pytest.mark.parametrize("w", [64])
+def test_mul_channel_bcast_repeat(device, h, w):
+    torch_input_tensor_a = torch.rand((16, 16, h, w), dtype=torch.bfloat16)
+    torch_input_tensor_b = torch.rand((16, 1, h, w), dtype=torch.bfloat16)
+    torch_output_tensor = torch.mul(torch_input_tensor_a, torch_input_tensor_b)
+
+    input_tensor_a = ttnn.from_torch(torch_input_tensor_a, layout=ttnn.TILE_LAYOUT, device=device)
+    input_tensor_b = ttnn.from_torch(torch_input_tensor_b, layout=ttnn.TILE_LAYOUT, device=device)
+    output = ttnn.mul(input_tensor_a, input_tensor_b)
+    output = ttnn.to_torch(output)
+
+    assert_with_pcc(torch_output_tensor, output, 0.9999)

ttnn/cpp/ttnn/operations/eltwise/binary/binary.cpp

@@ -113,13 +113,11 @@ auto preprocess_inputs(
 
         // repeats second if it is smaller
         if (first_shape.rank() == 4 and second_shape.rank() == 4 and first_shape[0] > second_shape[0]) {
-            tt::log_warning(tt::LogOp, "Using repeat op to broadcast batch dim");
             Shape repeats(std::array<uint32_t, 4>{first_shape[0], 1, 1, 1});
             second = ttnn::repeat(second, repeats);
         }
         // repeats second if it is smaller
         if (first_shape.rank() == 4 and second_shape.rank() == 4 and first_shape[1] > second_shape[1]) {
-            tt::log_warning(tt::LogOp, "Using repeat op to broadcast channel dim");
             Shape repeats(std::array<uint32_t, 4>{1, first_shape[1], 1, 1});
             second = ttnn::repeat(second, repeats);
         }