fix: make sure not to override user set values for from_sample (#3610)

src/bentoml/_internal/io_descriptors/numpy.py

@@ -483,8 +483,10 @@ async def predict(input: NDArray[np.int16]) -> NDArray[Any]:
             raise BentoMLException(
                 f"Failed to create a 'numpy.ndarray' from given sample {sample}"
             ) from None
-        self._dtype = sample.dtype
-        self._shape = sample.shape
+        if self._dtype is None:
+            self._dtype = sample.dtype
+        if self._shape is None:
+            self._shape = sample.shape
         return sample
 
     async def from_proto(self, field: pb.NDArray | bytes) -> ext.NpNDArray:

src/bentoml/_internal/io_descriptors/pandas.py

@@ -427,8 +427,10 @@ def predict(inputs: pd.DataFrame) -> pd.DataFrame: ...
                 raise InvalidArgument(
                     f"Failed to create a 'pd.DataFrame' from sample {sample}: {e}"
                 ) from None
-        self._shape = sample.shape
-        self._columns = [str(i) for i in list(sample.columns)]
+        if self._shape is None:
+            self._shape = sample.shape
+        if self._columns is None:
+            self._columns = [str(i) for i in list(sample.columns)]
         if self._dtype is None:
             self._dtype = sample.dtypes
         return sample
@@ -933,8 +935,10 @@ def predict(inputs: pd.Series) -> pd.Series: ...
         """
         if not isinstance(sample, pd.Series):
             sample = pd.Series(sample)
-        self._dtype = sample.dtype
-        self._shape = sample.shape
+        if self._dtype is None:
+            self._dtype = sample.dtype
+        if self._shape is None:
+            self._shape = sample.shape
         return sample
 
     def input_type(self) -> LazyType[ext.PdSeries]:

tests/unit/_internal/io/test_numpy.py

@@ -117,6 +117,14 @@ def test_verify_numpy_ndarray(caplog: LogCaptureFixture):
     assert "Failed to reshape" in caplog.text
 
 
+def test_from_sample_ensure_not_override():
+    example = NumpyNdarray.from_sample(np.ones((2, 2, 3)), dtype=np.float32)
+    assert example._dtype == np.float32
+
+    example = NumpyNdarray.from_sample(np.ones((2, 2, 3)), shape=(2, 2, 3))
+    assert example._shape == (2, 2, 3)
+
+
 def generate_1d_array(dtype: pb.NDArray.DType.ValueType, length: int = 3):
     if dtype == pb.NDArray.DTYPE_BOOL:
         return [True] * length