add PixelShuffleUpsample thanks to @MalumaDev and @marunine for running the experiment and verifyng absence of checkboard artifacts

dalle2_pytorch/dalle2_pytorch.py

@@ -1223,16 +1223,35 @@ class DiffusionPrior(nn.Module):
 
 # decoder
 
-def ConvTransposeUpsample(dim, dim_out = None):
+class PixelShuffleUpsample(nn.Module):
-    dim_out = default(dim_out, dim)
+    """
-    return nn.ConvTranspose2d(dim, dim_out, 4, 2, 1)
+    code shared by @MalumaDev at DALLE2-pytorch for addressing checkboard artifacts
+    https://arxiv.org/ftp/arxiv/papers/1707/1707.02937.pdf
+    """
+    def __init__(self, dim, dim_out = None):
+        super().__init__()
+        dim_out = default(dim_out, dim)
+        conv = nn.Conv2d(dim, dim_out * 4, 1)
 
-def NearestUpsample(dim, dim_out = None):
+        self.net = nn.Sequential(
-    dim_out = default(dim_out, dim)
+            conv,
-    return nn.Sequential(
+            nn.SiLU(),
-        nn.Upsample(scale_factor = 2, mode = 'nearest'),
+            nn.PixelShuffle(2)
-        nn.Conv2d(dim, dim_out, 3, padding = 1)
+        )
-    )
+
+        self.init_conv_(conv)
+
+    def init_conv_(self, conv):
+        o, i, h, w = conv.weight.shape
+        conv_weight = torch.empty(o // 4, i, h, w)
+        nn.init.kaiming_uniform_(conv_weight)
+        conv_weight = repeat(conv_weight, 'o ... -> (o 4) ...')
+
+        conv.weight.data.copy_(conv_weight)
+        nn.init.zeros_(conv.bias.data)
+
+    def forward(self, x):
+        return self.net(x)
 
 def Downsample(dim, *, dim_out = None):
     dim_out = default(dim_out, dim)
@@ -1496,7 +1515,7 @@ class Unet(nn.Module):
         cross_embed_downsample_kernel_sizes = (2, 4),
         memory_efficient = False,
         scale_skip_connection = False,
-        nearest_upsample = False,
+        pixel_shuffle_upsample = True,
         final_conv_kernel_size = 1,
         **kwargs
     ):
@@ -1610,7 +1629,7 @@ class Unet(nn.Module):
 
         # upsample klass
 
-        upsample_klass = ConvTransposeUpsample if not nearest_upsample else NearestUpsample
+        upsample_klass = ConvTransposeUpsample if not pixel_shuffle_upsample else PixelShuffleUpsample
 
         # give memory efficient unet an initial resnet block
 

dalle2_pytorch/version.py

@@ -1 +1 @@
-__version__ = '0.20.1'
+__version__ = '0.21.0'