diff --git a/configs/mmseg/segmentation_openvino_static-1024x2048.py b/configs/mmseg/segmentation_openvino_static-1024x2048.py
new file mode 100644
index 0000000000..472e923e61
--- /dev/null
+++ b/configs/mmseg/segmentation_openvino_static-1024x2048.py
@@ -0,0 +1,4 @@
+_base_ = ['./segmentation_static.py', '../_base_/backends/openvino.py']
+onnx_config = dict(input_shape=[2048, 1024])
+backend_config = dict(
+    model_inputs=[dict(opt_shapes=dict(input=[1, 3, 1024, 2048]))])
diff --git a/configs/mmseg/segmentation_openvino_static-512x512.py b/configs/mmseg/segmentation_openvino_static-512x512.py
index 29d7582758..ef974335e4 100644
--- a/configs/mmseg/segmentation_openvino_static-512x512.py
+++ b/configs/mmseg/segmentation_openvino_static-512x512.py
@@ -1,4 +1,4 @@
 _base_ = ['./segmentation_static.py', '../_base_/backends/openvino.py']
-
+onnx_config = dict(input_shape=[512, 512])
 backend_config = dict(
     model_inputs=[dict(opt_shapes=dict(input=[1, 3, 512, 512]))])
diff --git a/mmdeploy/apis/pytorch2torchscript.py b/mmdeploy/apis/pytorch2torchscript.py
index c984892360..8b54ce4ce8 100644
--- a/mmdeploy/apis/pytorch2torchscript.py
+++ b/mmdeploy/apis/pytorch2torchscript.py
@@ -42,8 +42,8 @@ def torch2torchscript_impl(model: torch.nn.Module,
             ir=IR.TORCHSCRIPT), torch.no_grad(), torch.jit.optimized_execution(
                 True):
         # for exporting models with weight that depends on inputs
-        patched_model(
-            *inputs) if isinstance(inputs, Sequence) else patched_model(inputs)
+        patched_model(*inputs) if isinstance(inputs, Sequence) \
+            else patched_model(inputs)
         ts_model = torch.jit.trace(patched_model, inputs)
 
     # perform optimize, note that optimizing models may trigger errors when
diff --git a/mmdeploy/codebase/mmpose/deploy/pose_detection.py b/mmdeploy/codebase/mmpose/deploy/pose_detection.py
index 52bb1632bc..4dee279a8c 100644
--- a/mmdeploy/codebase/mmpose/deploy/pose_detection.py
+++ b/mmdeploy/codebase/mmpose/deploy/pose_detection.py
@@ -198,6 +198,7 @@ def evaluate_outputs(model_cfg: mmcv.Config,
                          out: Optional[str] = None,
                          metric_options: Optional[dict] = None,
                          format_only: bool = False,
+                         log_file: Optional[str] = None,
                          **kwargs):
         """Perform post-processing to predictions of model.
 
@@ -215,10 +216,15 @@ def evaluate_outputs(model_cfg: mmcv.Config,
                 evaluation. It is useful when you want to format the result
                 to a specific format and submit it to the test server. Defaults
                 to `False`.
+            log_file (str | None): The file to write the evaluation results.
+                Defaults to `None` and the results will only print on stdout.
         """
+        from mmcv.utils import get_logger
+        logger = get_logger('test', log_file=log_file, log_level=logging.INFO)
+
         res_folder = '.'
         if out:
-            logging.info(f'\nwriting results to {out}')
+            logger.info(f'\nwriting results to {out}')
             mmcv.dump(outputs, out)
             res_folder, _ = os.path.split(out)
         os.makedirs(res_folder, exist_ok=True)
@@ -229,7 +235,7 @@ def evaluate_outputs(model_cfg: mmcv.Config,
 
         results = dataset.evaluate(outputs, res_folder, **eval_config)
         for k, v in sorted(results.items()):
-            print(f'{k}: {v}')
+            logger.info(f'{k}: {v:.4f}')
 
     def get_model_name(self) -> str:
         """Get the model name.