use pytorch built-in SiLU function to save GPU memory usage

ldm/modules/diffusionmodules/model.py

@@ -5,6 +5,7 @@ import torch.nn as nn
 import numpy as np
 from einops import rearrange
 from typing import Optional, Any
+import torch.nn.functional as F
 
 from ldm.modules.attention import MemoryEfficientCrossAttention
 
@@ -40,7 +41,7 @@ def get_timestep_embedding(timesteps, embedding_dim):
 
 def nonlinearity(x):
     # swish
-    return x*torch.sigmoid(x)
+    return F.silu(x)
 
 
 def Normalize(in_channels, num_groups=32):