--- ldm/modules/attention.py	2022-09-04 12:27:32.000000000 +0100
+++ ldm/modules/attention.py.opt	2022-09-04 12:27:03.000000000 +0100
@@ -179,25 +179,29 @@
         context = default(context, x)
         k = self.to_k(context)
         v = self.to_v(context)
+        del context, x
 
         q, k, v = map(
             lambda t: rearrange(t, 'b n (h d) -> (b h) n d', h=h), (q, k, v)
         )
 
         sim = einsum('b i d, b j d -> b i j', q, k) * self.scale
+        del q, k
 
         if exists(mask):
             mask = rearrange(mask, 'b ... -> b (...)')
             max_neg_value = -torch.finfo(sim.dtype).max
             mask = repeat(mask, 'b j -> (b h) () j', h=h)
             sim.masked_fill_(~mask, max_neg_value)
+            del mask
 
         # attention, what we cannot get enough of
-        attn = sim.softmax(dim=-1)
+        sim[4:] = sim[4:].softmax(dim=-1)
+        sim[:4] = sim[:4].softmax(dim=-1)
 
-        out = einsum('b i j, b j d -> b i d', attn, v)
-        out = rearrange(out, '(b h) n d -> b n (h d)', h=h)
-        return self.to_out(out)
+        sim = einsum('b i j, b j d -> b i d', sim, v)
+        sim = rearrange(sim, '(b h) n d -> b n (h d)', h=h)
+        return self.to_out(sim)
 
 
 class BasicTransformerBlock(nn.Module):