Fix autocast in SSIM metric

CHANGELOG.md

@@ -39,6 +39,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 - Fixed type checking on the `maximize` parameter at the initialization of `MetricTracker` ([#1428](https://github.com/Lightning-AI/metrics/issues/1428))
 
+- Fixed mixed precision autocast for `SSIM` metric ([#1454](https://github.com/Lightning-AI/metrics/pull/1454))
 
 - Fixed wrongly reset method in `MultioutputWrapper` ([#1460](https://github.com/Lightning-AI/metrics/issues/1460))
 

src/torchmetrics/functional/image/ssim.py

@@ -33,10 +33,7 @@ def _ssim_check_inputs(preds: Tensor, target: Tensor) -> Tuple[Tensor, Tensor]:
     """
 
     if preds.dtype != target.dtype:
-        raise TypeError(
-            "Expected `preds` and `target` to have the same data type."
-            f" Got preds: {preds.dtype} and target: {target.dtype}."
-        )
+        target = target.to(preds.dtype)
     _check_same_shape(preds, target)
     if len(preds.shape) not in (4, 5):
         raise ValueError(

tests/unittests/image/test_ssim.py

@@ -231,11 +231,8 @@ def test_ssim_half_gpu(self, preds, target, sigma):
     ],
 )
 def test_ssim_invalid_inputs(pred, target, kernel, sigma):
-    pred_t = torch.rand(pred, dtype=torch.float32)
-    target_t = torch.rand(target, dtype=torch.float64)
-    with pytest.raises(TypeError):
-        structural_similarity_index_measure(pred_t, target_t)
-
+    """Test that an value errors are raised if input sizes are different, kernel length and sigma does not match
+    size or invalid values are provided."""
     pred = torch.rand(pred)
     target = torch.rand(target)
     with pytest.raises(ValueError):