Port hypot method_tests() to OpInfo (pytorch#55140)

Summary:
Related pytorch#54261

Pull Request resolved: pytorch#55140

Reviewed By: ngimel

Differential Revision: D27562164

Pulled By: mruberry

fbshipit-source-id: fc698ddc624d2abf5d540aac76baa5d398993f1f

torch/testing/_internal/common_methods_invocations.py

@@ -737,6 +737,14 @@ def sample_inputs_hstack_dstack_vstack(op_info, device, dtype, requires_grad, **
 
     return (SampleInput(tensors),)
 
+def sample_inputs_hypot(op_info, device, dtype, requires_grad):
+    input = make_tensor((S, S), device, dtype, requires_grad=requires_grad)
+    args = make_tensor((S, S), device, dtype, requires_grad=requires_grad)
+
+    return (
+        SampleInput(input, args=(args,)),
+    )
+
 def sample_inputs_gather(op_info, device, dtype, requires_grad, **kwargs):
     return (
         SampleInput(
@@ -3867,6 +3875,12 @@ def gradcheck_wrapper_triangular_input(op, input, *args, upper=False, **kwargs):
            skips=(
                # hstack does not correctly warn when resizing out= inputs
                SkipInfo('TestCommon', 'test_out'),),),
+    OpInfo('hypot',
+           dtypes=floating_types(),
+           dtypesIfCPU=floating_types_and(torch.bfloat16),
+           dtypesIfCUDA=floating_types_and(torch.half),
+           sample_inputs_func=sample_inputs_hypot,
+           ),
     OpInfo('vstack',
            dtypes=all_types_and_complex_and(torch.bool, torch.float16, torch.bfloat16),
            sample_inputs_func=sample_inputs_hstack_dstack_vstack,
@@ -4347,7 +4361,6 @@ def method_tests():
         ('fmod', (), (non_differentiable(uniform_scalar(1.5)),), 'scalar_tensor'),
         ('fmod', (), (non_differentiable(torch.rand(S, S, S) + 1.5),), 'scalar_tensor_broadcast_lhs'),
         ('fmod', (S, S, S), (non_differentiable(uniform_scalar(1.5)),), 'scalar_tensor_broadcast_rhs'),
-        ('hypot', (S, S), ((S, S),)),
         ('remainder', (S, S, S), (1.5,), '', (True,)),
         ('remainder', (), (1.5,), 'scalar', (True,)),
         ('remainder', (S, S, S), (non_differentiable(torch.rand(S, S, S) + 1.5),), 'tensor'),