Fix aten.amax lowering issue

Summary:
There was an error when lowering amax around this line `input_tensor = self.get_tensor(input_node, node)` and the issue is that we're trying to permute the tensor inside node_visitors, op_node.meta[QCOM_AXIS_ORDER] is (0, 1), however, tensor.shape is (1, 980, 49).

Rollback Plan:

Differential Revision: D80187368

Author: cccclai

backends/qualcomm/_passes/layout_transform.py

@@ -175,6 +175,7 @@ def is_layout_agnostic(self, node: torch.fx.Node) -> bool:
             exir_ops.edge.aten.mean.dim,
             exir_ops.edge.aten.min.dim,
             exir_ops.edge.aten.sum.dim_IntList,
+            exir_ops.edge.aten.amax.default,
         }:
             # if dimemsion is not kept, we'll have no clue how to do layout transform
             if len(node.args) < 3 or not node.args[2]: