diff --git a/vall_e/data.py b/vall_e/data.py
index b3d67b8..440a709 100755
--- a/vall_e/data.py
+++ b/vall_e/data.py
@@ -675,6 +675,10 @@ class Dataset(_Dataset):
 		# this just makes it be happy
 		if len(self.dataset) == 0:
 			self.dataset = cfg.dataset.training
+
+		# hard error because I kept getting tricked by this myself
+		if self.sampler_order == "duration" and self.sampler_type != "path":
+			raise Exception(f'Requesting sample_type={self.sampler_type} with sample_order={self.sampler_order}, yet combination will not give expected results.')
 		
 		# dict of paths keyed by speaker names
 		self.paths_by_spkr_name = _load_paths(self.dataset, self.dataset_type)
diff --git a/vall_e/engines/deepspeed.py b/vall_e/engines/deepspeed.py
index 4abe44f..04809f2 100755
--- a/vall_e/engines/deepspeed.py
+++ b/vall_e/engines/deepspeed.py
@@ -64,6 +64,7 @@ class Engine(DeepSpeedEngine):
 
 		self.max_nan_losses = 8
 		self.current_batch_size = 0
+		self.skip_on_nan = True
 
 	def freeze(self, freeze_all=True):
 		# freeze non-LoRA params if requested
@@ -142,14 +143,16 @@ class Engine(DeepSpeedEngine):
 		losses = self.gather_attribute("loss")
 		loss = torch.stack([*losses.values()]).sum()
 
+		stats = {}
+		stats |= {k: v.item() for k, v in losses.items()}
+		stats |= self.gather_attribute("scalar")
+
 		if torch.isnan(loss).any():
 			self.max_nan_losses = self.max_nan_losses - 1
 			if self.max_nan_losses < 0:
 				raise RuntimeError("Too many NaN losses detected.")
-
-		stats = {}
-		stats |= {k: v.item() for k, v in losses.items()}
-		stats |= self.gather_attribute("scalar")
+			
+			return stats
 
 		self.backward(loss)
 		self.step()