[Fix] Fix SegTTAModel with no attribute '_gt_sem_seg' error (open-mmlab#3152)

## Motivation

When using the - tta command for multi-scale prediction, and the test
set is not annotated, although format_only has been set true in
test_evaluator, but SegTTAModel class still threw error 'AttributeError:
'SegDataSample' object has no attribute '_gt_sem_seg''.

## Modification

The reason is SegTTAModel didn't determine if there were annotations in
the dataset, so I added the code to make the judgment and let the
program run normally on my computer.

Author: ZiAn-Su

mmseg/models/segmentors/seg_tta.py

@@ -6,7 +6,6 @@
 from mmengine.structures import PixelData
 
 from mmseg.registry import MODELS
-from mmseg.structures import SegDataSample
 from mmseg.utils import SampleList
 
 
@@ -39,11 +38,10 @@ def merge_preds(self, data_samples_list: List[SampleList]) -> SampleList:
                             ).to(logits).squeeze(1)
             else:
                 seg_pred = logits.argmax(dim=0)
-            data_sample = SegDataSample(
-                **{
-                    'pred_sem_seg': PixelData(data=seg_pred),
-                    'gt_sem_seg': data_samples[0].gt_sem_seg
-                })
+            data_sample.set_data({'pred_sem_seg': PixelData(data=seg_pred)})
+            if hasattr(data_samples[0], 'gt_sem_seg'):
+                data_sample.set_data(
+                    {'gt_sem_seg': data_samples[0].gt_sem_seg})
             data_sample.set_metainfo({'img_path': data_samples[0].img_path})
             predictions.append(data_sample)
         return predictions