diff --git a/torchtune/modules/transformer.py b/torchtune/modules/transformer.py
index 1420907750..24c2c180e4 100644
--- a/torchtune/modules/transformer.py
+++ b/torchtune/modules/transformer.py
@@ -11,7 +11,6 @@
 from torch import nn
 from torchtune.modules import MultiHeadAttention
 from torchtune.modules.attention_utils import _MaskType
-from torchtune.utils.logging import deprecated
 
 
 class TransformerSelfAttentionLayer(nn.Module):
@@ -515,11 +514,6 @@ def forward(
         return output
 
 
-@deprecated(
-    msg="Please use torchtune.modules.TransformerDecoder instead. \
-If you need an example, see torchtune.models.qwen2._component_builders.py \
-and how to implement torch.modules.TiedLinear for the output projection."
-)
 class TiedEmbeddingTransformerDecoder(nn.Module):
     """
     Transformer Decoder with tied embedding weight. A key difference between