diff --git a/tests/test_model.py b/tests/test_model.py
index 2f7957a..a1397f2 100644
--- a/tests/test_model.py
+++ b/tests/test_model.py
@@ -75,7 +75,6 @@ def test_attention(fairscale_init):
         iattention.v_projector.weight.copy_(attention.wv.weight)
         iattention.output_projector.weight.copy_(attention.wo.weight)
 
-        iattention.cuda()
         iattention.eval()
 
         output2 = iattention(x, freqs_cis, 0)        
@@ -163,7 +162,6 @@ def test_decoder():
         transformer_block.attention.cache_k.random_()
         transformer_block.attention.cache_v.random_()
 
-        transformer_block
         transformer_block.eval()
 
         itransformer_block.attention.k_cache.sequence_cache.copy_(transformer_block.attention.cache_k.transpose(1, 2))
@@ -177,8 +175,6 @@ def test_decoder():
         itransformer_block.ffn.output_layer.weight.copy_(transformer_block.feed_forward.w2.weight)
         itransformer_block.ffn.gate_layer.weight.copy_(transformer_block.feed_forward.w3.weight)
 
-        itransformer_block
-
         x = torch.randn(2, 2048, 4096)
 
         freqs_cis = precompute_freqs_cis(4096, 2048 * 2, 500000)