switch to learned sinsuoidal pos emb for the continuous case

denoising_diffusion_pytorch/continuous_time_gaussian_diffusion.py

@@ -125,7 +125,7 @@ def __init__(
         p2_loss_weight_k = 1
     ):
         super().__init__()
-        assert not denoise_fn.sinusoidal_cond_mlp
+        assert denoise_fn.learned_sinusoidal_cond
 
         self.denoise_fn = denoise_fn
 

denoising_diffusion_pytorch/denoising_diffusion_pytorch.py

@@ -73,20 +73,6 @@ def __init__(self, fn):
     def forward(self, x, *args, **kwargs):
         return self.fn(x, *args, **kwargs) + x
 
-class SinusoidalPosEmb(nn.Module):
-    def __init__(self, dim):
-        super().__init__()
-        self.dim = dim
-
-    def forward(self, x):
-        device = x.device
-        half_dim = self.dim // 2
-        emb = math.log(10000) / (half_dim - 1)
-        emb = torch.exp(torch.arange(half_dim, device=device) * -emb)
-        emb = x[:, None] * emb[None, :]
-        emb = torch.cat((emb.sin(), emb.cos()), dim=-1)
-        return emb
-
 def Upsample(dim):
     return nn.ConvTranspose2d(dim, dim, 4, 2, 1)
 
@@ -115,6 +101,39 @@ def forward(self, x):
         x = self.norm(x)
         return self.fn(x)
 
+# sinusoidal positional embeds
+
+class SinusoidalPosEmb(nn.Module):
+    def __init__(self, dim):
+        super().__init__()
+        self.dim = dim
+
+    def forward(self, x):
+        device = x.device
+        half_dim = self.dim // 2
+        emb = math.log(10000) / (half_dim - 1)
+        emb = torch.exp(torch.arange(half_dim, device=device) * -emb)
+        emb = x[:, None] * emb[None, :]
+        emb = torch.cat((emb.sin(), emb.cos()), dim=-1)
+        return emb
+
+class LearnedSinusoidalPosEmb(nn.Module):
+    """ following @crowsonkb 's lead with learned sinusoidal pos emb """
+    """ https://github.com/crowsonkb/v-diffusion-jax/blob/master/diffusion/models/danbooru_128.py#L8 """
+
+    def __init__(self, dim):
+        super().__init__()
+        assert (dim % 2) == 0
+        half_dim = dim // 2
+        self.weights = nn.Parameter(torch.randn(half_dim))
+
+    def forward(self, x):
+        x = rearrange(x, 'b -> b 1')
+        freqs = x * rearrange(self.weights, 'd -> 1 d') * 2 * math.pi
+        fouriered = torch.cat((freqs.sin(), freqs.cos()), dim = -1)
+        fouriered = torch.cat((x, fouriered), dim = -1)
+        return fouriered
+
 # building block modules
 
 class Block(nn.Module):
@@ -158,6 +177,7 @@ def forward(self, x, time_emb = None):
         h = self.block1(x, scale_shift = scale_shift)
 
         h = self.block2(h)
+
         return h + self.res_conv(x)
 
 class LinearAttention(nn.Module):
@@ -213,18 +233,6 @@ def forward(self, x):
 
 # model
 
-def MLP(dim_in, dim_hidden):
-    return nn.Sequential(
-        Rearrange('... -> ... 1'),
-        nn.Linear(1, dim_hidden),
-        nn.GELU(),
-        nn.LayerNorm(dim_hidden),
-        nn.Linear(dim_hidden, dim_hidden),
-        nn.GELU(),
-        nn.LayerNorm(dim_hidden),
-        nn.Linear(dim_hidden, dim_hidden)
-    )
-
 class Unet(nn.Module):
     def __init__(
         self,
@@ -235,7 +243,8 @@ def __init__(
         channels = 3,
         resnet_block_groups = 8,
         learned_variance = False,
-        sinusoidal_cond_mlp = True
+        learned_sinusoidal_cond = False,
+        learned_sinusoidal_dim = 16
     ):
         super().__init__()
 
@@ -255,17 +264,21 @@ def __init__(
 
         time_dim = dim * 4
 
-        self.sinusoidal_cond_mlp = sinusoidal_cond_mlp
+        self.learned_sinusoidal_cond = learned_sinusoidal_cond
 
-        if sinusoidal_cond_mlp:
-            self.time_mlp = nn.Sequential(
-                SinusoidalPosEmb(dim),
-                nn.Linear(dim, time_dim),
-                nn.GELU(),
-                nn.Linear(time_dim, time_dim)
-            )
+        if learned_sinusoidal_cond:
+            sinu_pos_emb = LearnedSinusoidalPosEmb(learned_sinusoidal_dim)
+            fourier_dim = learned_sinusoidal_dim + 1
         else:
-            self.time_mlp = MLP(1, time_dim)
+            sinu_pos_emb = SinusoidalPosEmb(dim)
+            fourier_dim = dim
+
+        self.time_mlp = nn.Sequential(
+            sinu_pos_emb,
+            nn.Linear(fourier_dim, time_dim),
+            nn.GELU(),
+            nn.Linear(time_dim, time_dim)
+        )
 
         # layers
 

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'denoising-diffusion-pytorch',
   packages = find_packages(),
-  version = '0.19.2',
+  version = '0.20.0',
   license='MIT',
   description = 'Denoising Diffusion Probabilistic Models - Pytorch',
   author = 'Phil Wang',