diff --git a/setup.py b/setup.py
index 5011590..efd518c 100644
--- a/setup.py
+++ b/setup.py
@@ -3,7 +3,7 @@
 setup(
   name = 'soundstorm-pytorch',
   packages = find_packages(exclude=[]),
-  version = '0.4.6',
+  version = '0.4.7',
   license='MIT',
   description = 'SoundStorm - Efficient Parallel Audio Generation from Google Deepmind, in Pytorch',
   author = 'Phil Wang',
diff --git a/soundstorm_pytorch/soundstorm.py b/soundstorm_pytorch/soundstorm.py
index 303d79c..733286c 100644
--- a/soundstorm_pytorch/soundstorm.py
+++ b/soundstorm_pytorch/soundstorm.py
@@ -1119,7 +1119,7 @@ def forward(
         rand_times = torch.empty(b, device = device).uniform_(0, 1)
         rand_probs = self.schedule_fn(rand_times)
 
-        mask = get_mask_subset_prob(mask, rand_probs)
+        mask = get_mask_subset_prob(mask, rand_probs, min_mask = 1)
 
         # random quantizer position, in groups