Re-enable passing BytesIO as path in `.to_onnx()` (#20172)

src/lightning/pytorch/core/module.py

@@ -17,6 +17,7 @@ import logging
 import numbers
 import weakref
 from contextlib import contextmanager
+from io import BytesIO
 from pathlib import Path
 from typing import (
     IO,
@@ -1364,7 +1365,7 @@ class LightningModule(
             )
 
     @torch.no_grad()
-    def to_onnx(self, file_path: Union[str, Path], input_sample: Optional[Any] = None, **kwargs: Any) -> None:
+    def to_onnx(self, file_path: Union[str, Path, BytesIO], input_sample: Optional[Any] = None, **kwargs: Any) -> None:
         """Saves the model in ONNX format.
 
         Args:
@@ -1403,7 +1404,8 @@ class LightningModule(
         input_sample = self._on_before_batch_transfer(input_sample)
         input_sample = self._apply_batch_transfer_handler(input_sample)
 
-        torch.onnx.export(self, input_sample, str(file_path), **kwargs)
+        file_path = str(file_path) if isinstance(file_path, Path) else file_path
+        torch.onnx.export(self, input_sample, file_path, **kwargs)
         self.train(mode)
 
     @torch.no_grad()

tests/tests_pytorch/models/test_onnx.py

@@ -13,6 +13,7 @@
 # limitations under the License.
 import operator
 import os
+from io import BytesIO
 from pathlib import Path
 from unittest.mock import patch
 
@@ -45,6 +46,10 @@ def test_model_saves_with_input_sample(tmp_path):
     assert os.path.isfile(file_path)
     assert os.path.getsize(file_path) > 4e2
 
+    file_path = BytesIO()
+    model.to_onnx(file_path=file_path, input_sample=input_sample)
+    assert len(file_path.getvalue()) > 4e2
+
 
 @pytest.mark.parametrize(
     "accelerator", [pytest.param("mps", marks=RunIf(mps=True)), pytest.param("gpu", marks=RunIf(min_cuda_gpus=True))]