scatter_add_decomposition (#2740)

Author: apbose

py/torch_tensorrt/dynamo/lowering/_decompositions.py

@@ -243,6 +243,44 @@ def empty_strided_decomposition(*args, **kwargs) -> torch.Tensor:
     )
 
 
+@register_torch_trt_decomposition(
+    torch.ops.aten.scatter_add.default, registry=TORCH_TRT_DECOMPOSITIONS
+)
+def scatter_add_decomposition(
+    input_tensor: torch.Tensor,
+    dim: int,
+    index: torch.Tensor,
+    src_tensor: torch.Tensor,
+) -> torch.Tensor:
+    scatter_add_tensor = input_tensor
+    src_shape = list(src_tensor.shape)
+    src_dim = src_shape[dim]
+    for i in range(0, src_dim):
+        to_scatter_tensor = torch.zeros_like(input_tensor)
+
+        # index and src slice
+        src_slice = torch.select(src_tensor, dim, i)
+        index_slice = torch.select(index, dim, i)
+
+        # unsqueeze src and index in dim
+        src_slice = torch.unsqueeze(src_slice, dim)
+        index_slice = torch.unsqueeze(index_slice, dim)
+
+        # moving tensor to default device
+        device = to_torch_device(default_device())
+        scatter_add_tensor = scatter_add_tensor.to(device)
+        to_scatter_tensor = to_scatter_tensor.to(device)
+        index_slice = index_slice.to(device)
+        src_slice = src_slice.to(device)
+
+        scatter_add_tensor = torch.add(
+            scatter_add_tensor,
+            torch.scatter(to_scatter_tensor, dim, index_slice, src_slice),
+        )
+
+    return scatter_add_tensor
+
+
 def get_decompositions(
     enable_experimental_decompositions: bool = False,
 ) -> Dict[OpOverload, Callable[[Any], Any]]:

tests/py/dynamo/lowering/test_decompositions.py

@@ -962,6 +962,119 @@ def forward(self, input):
             f"The optimized model results shape and torch model results shape should be equal in empty_stride",
         )
 
+    @parameterized.expand(
+        [
+            (
+                "scatter_add_zero_dim_indexOne_constant",
+                0,
+                torch.tensor([[0, 1, 2, 0]]).cuda(),
+                torch.tensor([[1, 2, 3, 4]], dtype=torch.int32).cuda(),
+                {torch.ops.aten.add.Tensor},
+            ),
+            (
+                "scatter_add_zero_dim_indexTwo_constant",
+                0,
+                torch.tensor([[0, 1, 2, 0], [1, 2, 1, 1]]).cuda(),
+                torch.tensor([[1, 2, 3, 4], [5, 6, 7, 8]], dtype=torch.int32).cuda(),
+                {torch.ops.aten.add.Tensor, torch.ops.aten.scatter.src},
+            ),
+            (
+                "scatter_add_one_dim_indexOne_constant",
+                1,
+                torch.tensor([[0, 1, 2, 0]]).cuda(),
+                torch.tensor([[1, 2, 3, 1]], dtype=torch.int32).cuda(),
+                {
+                    torch.ops.aten.add.Tensor,
+                    torch.ops.aten.scatter.src,
+                    torch.ops.aten.full_like.default,
+                },
+            ),
+            (
+                "scatter_add_one_dim_indexTwo_constant",
+                1,
+                torch.tensor([[0, 1, 2, 0], [1, 2, 1, 1]]).cuda(),
+                torch.tensor([[1, 2, 3, 1], [5, 6, 5, 5]], dtype=torch.int32).cuda(),
+                {
+                    torch.ops.aten.add.Tensor,
+                    torch.ops.aten.scatter.src,
+                    torch.ops.aten.full_like.default,
+                },
+            ),
+            (
+                "scatter_add_one_dim_indexTwo_constant",
+                1,
+                torch.tensor([[0, 1, 2, 0], [1, 2, 1, 1], [3, 2, 1, 2]]).cuda(),
+                torch.tensor(
+                    [[1, 2, 3, 1], [5, 6, 5, 5], [2, 4, 3, 2]], dtype=torch.int32
+                ).cuda(),
+                {
+                    torch.ops.aten.add.Tensor,
+                    torch.ops.aten.scatter.src,
+                    torch.ops.aten.full_like.default,
+                },
+            ),
+        ]
+    )
+    def test_scatter_add(self, _, dim, index, src, expected_ops_param):
+        class TestModule(torch.nn.Module):
+            def __init__(self):
+                super().__init__()
+
+            def forward(self, input):
+                return torch.ops.aten.scatter_add.default(input, dim, index, src)
+
+        # Operations expected to be included in the traced graph after decompositions
+        expected_ops = expected_ops_param
+        unexpected_ops = {torch.ops.aten.scatter_add.default}
+
+        input = torch.zeros(3, 5, dtype=torch.int32).cuda()
+        inputs = [input]
+
+        fx_graph = torch.fx.symbolic_trace(TestModule())
+        unexpected_ops_seen, expected_ops_unseen = lower_graph_testing(
+            fx_graph,
+            inputs,
+            expected_ops=expected_ops,
+            unexpected_ops=unexpected_ops,
+            min_block_size=2,
+        )
+
+        self.assertEquals(
+            len(expected_ops_unseen),
+            0,
+            f"The following expected ops were not encountered: {expected_ops_unseen}",
+        )
+
+        self.assertEquals(
+            len(unexpected_ops_seen),
+            0,
+            f"The following expected ops were not encountered: {unexpected_ops_seen}",
+        )
+
+        torch._dynamo.reset()
+
+        # Validate that the results between Torch and Torch-TRT are similar
+        optimized_model = torch_tensorrt.compile(
+            fx_graph,
+            "torch_compile",
+            inputs,
+            min_block_size=1,
+            truncate_double=True,
+            pass_through_build_failures=True,
+        )
+        optimized_model_results = optimized_model(*inputs).detach().cpu()
+        torch_model_results = fx_graph(*inputs).detach().cpu()
+
+        max_diff = float(
+            torch.max(torch.abs(optimized_model_results - torch_model_results))
+        )
+        self.assertAlmostEqual(
+            max_diff,
+            0,
+            DECIMALS_OF_AGREEMENT,
+            f"Scatter_add TRT outputs don't match with the original model.",
+        )
+
 
 if __name__ == "__main__":
     run_tests()