Set lower validation batch size for yolov8 (#1629)

* Reducing validation batch size

* updating validator

* set minimum to 1

* style changes

---------

Co-authored-by: Alexandre Marques <alexandre@neuralmagic.com>

src/sparseml/yolov8/trainers.py

@@ -279,6 +279,15 @@ def _setup_train(self, rank, world_size):
         super()._setup_train(rank, world_size)
         # NOTE: self.resume_training() was called in ^
 
+        if rank in {0, -1}:
+            self.test_loader = self.get_dataloader(
+                self.testset,
+                batch_size=max(1, self.train_loader.batch_size // 4),
+                rank=-1,
+                mode="val",
+            )
+            self.validator = self.get_validator()
+
         if rank in {0, -1}:
             config = dict(self.args)
             if self.manager is not None: