Add missing output_size kwarg to repeat_interleave (#2691)

Fixes #2687

---------

Signed-off-by: Yuan Yao <yuanyao@nvidia.com>

Author: yuanyao-nv

onnxscript/function_libs/torch_lib/ops/core.py

@@ -7539,7 +7539,10 @@ def aten_repeat(self: TTensor, repeats: Sequence[TInt]) -> TTensor:
 
 @torch_op("aten::repeat_interleave.self_int", trace_only=True)
 def aten_repeat_interleave_self_int(
-    self: TensorType, repeats: int, dim: Optional[int] = None
+    self: TensorType,
+    repeats: int,
+    dim: Optional[int] = None,
+    output_size: Optional[int] = None,
 ) -> TensorType:
     """repeat_interleave.self_int(Tensor self, SymInt repeats, int? dim=None, *, SymInt? output_size=None) -> Tensor