diff --git a/test/sparsity/test_fast_sparse_training.py b/test/sparsity/test_fast_sparse_training.py
index 2bd0d1878c..a0886dd898 100644
--- a/test/sparsity/test_fast_sparse_training.py
+++ b/test/sparsity/test_fast_sparse_training.py
@@ -14,7 +14,7 @@
 )
 from torchao.utils import TORCH_VERSION_AFTER_2_4, is_fbcode
 
-class TestModel(nn.Module):
+class ToyModel(nn.Module):
     def __init__(self):
         super().__init__()
         self.linear1 = nn.Linear(128, 256, bias=False)
@@ -36,7 +36,7 @@ def test_runtime_weight_sparsification(self):
         from torch.sparse import SparseSemiStructuredTensorCUSPARSELT
         input = torch.rand((128, 128)).half().cuda()
         grad = torch.rand((128, 128)).half().cuda()
-        model = TestModel().half().cuda()
+        model = ToyModel().half().cuda()
         model_c = copy.deepcopy(model)
 
         for name, mod in model.named_modules():
@@ -77,7 +77,7 @@ def test_runtime_weight_sparsification_compile(self):
         from torch.sparse import SparseSemiStructuredTensorCUSPARSELT
         input = torch.rand((128, 128)).half().cuda()
         grad = torch.rand((128, 128)).half().cuda()
-        model = TestModel().half().cuda()
+        model = ToyModel().half().cuda()
         model_c = copy.deepcopy(model)
 
         for name, mod in model.named_modules():