take a gamble on cosine sim attention

denoising_diffusion_pytorch/denoising_diffusion_pytorch.py

@@ -58,6 +58,9 @@ def convert_image_to(img_type, image):
         return image.convert(img_type)
     return image
 
+def l2norm(t):
+    return F.normalize(t, dim = -1)
+
 # normalization functions
 
 def normalize_to_neg_one_to_one(img):
@@ -215,9 +218,9 @@ class LinearAttention(nn.Module):
         return self.to_out(out)
 
 class Attention(nn.Module):
-    def __init__(self, dim, heads = 4, dim_head = 32):
+    def __init__(self, dim, heads = 4, dim_head = 32, scale = 16):
         super().__init__()
-        self.scale = dim_head ** -0.5
+        self.scale = scale
         self.heads = heads
         hidden_dim = dim_head * heads
         self.to_qkv = nn.Conv2d(dim, hidden_dim * 3, 1, bias = False)
@@ -227,10 +230,10 @@ class Attention(nn.Module):
         b, c, h, w = x.shape
         qkv = self.to_qkv(x).chunk(3, dim = 1)
         q, k, v = map(lambda t: rearrange(t, 'b (h c) x y -> b h c (x y)', h = self.heads), qkv)
-        q = q * self.scale
 
-        sim = einsum('b h d i, b h d j -> b h i j', q, k)
-        sim = sim - sim.amax(dim = -1, keepdim = True).detach()
+        q, k = map(l2norm, (q, k))
+
+        sim = einsum('b h d i, b h d j -> b h i j', q, k) * self.scale
         attn = sim.softmax(dim = -1)
 
         out = einsum('b h i j, b h d j -> b h i d', attn, v)

setup.py

@@ -3,7 +3,7 @@ from setuptools import setup, find_packages
 setup(
   name = 'denoising-diffusion-pytorch',
   packages = find_packages(),
-  version = '0.25.3',
+  version = '0.26.0',
   license='MIT',
   description = 'Denoising Diffusion Probabilistic Models - Pytorch',
   author = 'Phil Wang',