From 3979f53aa42113ec583265ed35e6b6253f9bcf9e Mon Sep 17 00:00:00 2001
From: Xiaoyu <85524621+xiaoyu-work@users.noreply.github.com>
Date: Mon, 1 Apr 2024 21:37:56 -0700
Subject: [PATCH] Update api backward compatibility (#20136)

### Description
Update api backward compatibility

### Motivation and Context
Update api backward compatibility
---
 .../python/tools/quantization/shape_inference.py     | 12 ++++++++++--
 onnxruntime/python/tools/transformers/optimizer.py   |  7 ++++++-
 2 files changed, 16 insertions(+), 3 deletions(-)

diff --git a/onnxruntime/python/tools/quantization/shape_inference.py b/onnxruntime/python/tools/quantization/shape_inference.py
index 7368304837..c07007f9d6 100644
--- a/onnxruntime/python/tools/quantization/shape_inference.py
+++ b/onnxruntime/python/tools/quantization/shape_inference.py
@@ -23,8 +23,8 @@ logger = logging.getLogger(__name__)
 
 
 def quant_pre_process(
-    input_model: Union[str, Path, onnx.ModelProto],
-    output_model_path: Union[str, Path],
+    input_model: Optional[Union[str, Path, onnx.ModelProto]] = None,
+    output_model_path: Optional[Union[str, Path]] = None,
     skip_optimization: bool = False,
     skip_onnx_shape: bool = False,
     skip_symbolic_shape: bool = False,
@@ -36,6 +36,7 @@ def quant_pre_process(
     all_tensors_to_one_file: bool = False,
     external_data_location: Optional[str] = None,
     external_data_size_threshold: int = 1024,
+    **deprecated_kwargs,
 ) -> None:
     """Shape inference and model optimization, in preparation for quantization.
 
@@ -63,6 +64,13 @@ def quant_pre_process(
         external_data_location: The file location to save the external file
         external_data_size_threshold: The size threshold for external data
     """
+
+    if input_model is None:
+        input_model = deprecated_kwargs.pop("input_model_path", None)
+    assert input_model is not None
+
+    assert output_model_path is not None, "output_model_path is required."
+
     with tempfile.TemporaryDirectory(prefix="pre.quant.") as quant_tmp_dir:
         temp_path = Path(quant_tmp_dir)
         model = None
diff --git a/onnxruntime/python/tools/transformers/optimizer.py b/onnxruntime/python/tools/transformers/optimizer.py
index e334e22d92..5f161674b6 100644
--- a/onnxruntime/python/tools/transformers/optimizer.py
+++ b/onnxruntime/python/tools/transformers/optimizer.py
@@ -68,7 +68,7 @@ MODEL_TYPES = {
 
 
 def optimize_by_onnxruntime(
-    onnx_model: Union[str, ModelProto],
+    onnx_model: Optional[Union[str, ModelProto]] = None,
     use_gpu: bool = False,
     optimized_model_path: Optional[str] = None,
     opt_level: Optional[int] = 99,
@@ -79,6 +79,7 @@ def optimize_by_onnxruntime(
     external_data_file_threshold: int = 1024,
     *,
     provider: Optional[str] = None,
+    **deprecated_kwargs,
 ) -> str:
     """
     Use onnxruntime to optimize model.
@@ -99,6 +100,10 @@ def optimize_by_onnxruntime(
     assert opt_level in [1, 2, 99]
     from torch import version as torch_version
 
+    if onnx_model is None:
+        onnx_model = deprecated_kwargs.pop("onnx_model_path", None)
+    assert onnx_model is not None
+
     if (
         use_gpu
         and provider is None