fix processing logic of the arange function when dtype is empty. (#53800)

* fix processing logic of the arange function when dtype is empty.

* update commit version

* fix ValueError when end is None.

* add unitest for new case.

* fix tensor type.

* remove paddle.to_tensor(), add more test unit.

* remove useless line.

* fix enable_static

* add new test unit.

* fix by comment.

python/paddle/fluid/tests/unittests/test_arange.py

@@ -170,12 +170,46 @@ class TestArangeImperative(unittest.TestCase):
         end = paddle.to_tensor(np.array([5], 'float32'))
         step = paddle.to_tensor(np.array([1], 'float32'))
         x4 = paddle.arange(start, end, step, 'int64')
-        paddle.enable_static()
 
         expected_data = np.arange(0, 5, 1).astype(np.int64)
         for i in [x1, x2, x3, x4]:
             self.assertEqual((i.numpy() == expected_data).all(), True)
 
+        start_float = paddle.to_tensor(np.array([0.5], 'float32'))
+        end_float = paddle.to_tensor(np.array([1.5], 'float32'))
+        step_float = paddle.to_tensor(np.array([0.5], 'float32'))
+        # all [start, end, step] is float
+        x5 = paddle.arange(start_float, end_float, step_float)
+        x5_expected_data = np.arange(0.5, 1.5, 0.5).astype(np.float32)
+        self.assertEqual((x5.numpy() == x5_expected_data).all(), True)
+
+        # [start, end] is float , [step] is int
+        x6 = paddle.arange(start_float, end_float, 1)
+        x6_expected_data = np.arange(0.5, 1.5, 1).astype(np.float32)
+        self.assertEqual((x6.numpy() == x6_expected_data).all(), True)
+
+        # [start] is float , [end] is int
+        x7 = paddle.arange(start_float, 1)
+        x7_expected_data = np.arange(0.5, 1).astype(np.float32)
+        self.assertEqual((x7.numpy() == x7_expected_data).all(), True)
+
+        # [start] is float
+        x8 = paddle.arange(start_float)
+        x8_expected_data = np.arange(0.5).astype(np.float32)
+        self.assertEqual((x8.numpy() == x8_expected_data).all(), True)
+
+        # [start] is int
+        x9 = paddle.arange(1)
+        x9_expected_data = np.arange(1).astype(np.int64)
+        self.assertEqual((x9.numpy() == x9_expected_data).all(), True)
+
+        # [start] is float
+        x10 = paddle.arange(1.0)
+        x10_expected_data = np.arange(1).astype(np.float32)
+        self.assertEqual((x10.numpy() == x10_expected_data).all(), True)
+
+        paddle.enable_static()
+
 
 if __name__ == "__main__":
     unittest.main()

python/paddle/tensor/creation.py

@@ -1295,12 +1295,21 @@ def arange(start=0, end=None, step=1, dtype=None, name=None):
             # [3, 4, 5, 6]
 
     """
-    if dtype is None:
-        dtype = 'int64'
     if end is None:
         end = start
         start = 0
 
+    if dtype is None:
+        for val in [start, end, step]:
+            if isinstance(val, Variable) and not val.is_integer():
+                dtype = paddle.get_default_dtype()
+                break
+            elif not isinstance(val, int) and not isinstance(val, Variable):
+                dtype = paddle.get_default_dtype()
+                break
+            else:
+                dtype = 'int64'
+
     out_shape = None
     if not in_dynamic_mode() and (
         not isinstance(start, Variable)