addressing review comments

pytorch · Aug 16, 2024 · 8295117 · 8295117
1 parent a5ebd76
commit 8295117
Showing 1 changed file with 26 additions and 24 deletions.
diff --git a/py/torch_tensorrt/dynamo/conversion/impl/select.py b/py/torch_tensorrt/dynamo/conversion/impl/select.py
@@ -156,48 +156,50 @@ def index(
         _LOGGER.debug(f"The shape of transpose tensor is {transpose_tensor_shape}")
 
         mult_d0 = 1
+        dim_tensor_shape_mult_d0 = 1
         for i in range(adv_indx_count):
+            if transpose_tensor_shape[i] == DYNAMIC_DIM:
+                dim_tensor_shape_mult_d0 = get_shape(
+                    ctx,
+                    target,
+                    source_ir,
+                    name + f"_transpose_tensor_shape_mult_d0_{i}",
+                    transpose_tensor,
+                    i,
+                )
+            else:
+                dim_tensor_shape_mult_d0 = transpose_tensor_shape[i]
             mult_d0 = convert_binary_elementwise(
                 ctx,
                 target,
                 source_ir,
                 name + f"_shape_{i}",
                 trt.ElementWiseOperation.PROD,
                 mult_d0,
-                (
-                    get_shape(
-                        ctx,
-                        target,
-                        source_ir,
-                        name + f"_transpose_tensor_shape_mult_d0_{i}",
-                        transpose_tensor,
-                        i,
-                    )
-                    if transpose_tensor_shape[i] == DYNAMIC_DIM
-                    else transpose_tensor_shape[i]
-                ),
+                dim_tensor_shape_mult_d0,
             )
         mult_d1 = 1
+        dim_tensor_shape_mult_d1 = 1
         for i in range(adv_indx_count, rank):
+            if transpose_tensor_shape[i] == DYNAMIC_DIM:
+                dim_tensor_shape_mult_d1 = get_shape(
+                    ctx,
+                    target,
+                    source_ir,
+                    name + f"_transpose_tensor_shape_mult_d0_{i}",
+                    transpose_tensor,
+                    i,
+                )
+            else:
+                dim_tensor_shape_mult_d1 = transpose_tensor_shape[i]
             mult_d1 = convert_binary_elementwise(
                 ctx,
                 target,
                 source_ir,
                 name + f"_shape_{i}",
                 trt.ElementWiseOperation.PROD,
                 mult_d1,
-                (
-                    get_shape(
-                        ctx,
-                        target,
-                        source_ir,
-                        name + f"_transpose_tensor_shape_mult_d1_{i}",
-                        transpose_tensor,
-                        i,
-                    )
-                    if transpose_tensor_shape[i] == DYNAMIC_DIM
-                    else transpose_tensor_shape[i]
-                ),
+                dim_tensor_shape_mult_d1,
             )
 
         concat_tensor_layer = ctx.net.add_concatenation(