Throw out the idea of conditioning on discrete codes. Oh well :(

codes/models/gpt_voice/lucidrains_dvae.py

@@ -190,8 +190,10 @@ class DiscreteVAE(nn.Module):
             arrange = 'b (h w) d -> b d h w'
             kwargs = {'h': h, 'w': w}
         image_embeds = rearrange(image_embeds, arrange, **kwargs)
-        images = self.decoder(image_embeds)
-        return images
+        images = [image_embeds]
+        for layer in self.decoder:
+            images.append(layer(images[-1]))
+        return images[-1], images[-2]
 
     def infer(self, img):
         img = self.norm(img)

codes/models/gpt_voice/unet_diffusion_vocoder_with_ref.py

@@ -11,9 +11,9 @@ from utils.util import get_mask_from_lengths
 
 
 class DiscreteSpectrogramConditioningBlock(nn.Module):
-    def __init__(self, discrete_codes, channels):
+    def __init__(self, dvae_channels, channels):
         super().__init__()
-        self.emb = nn.Embedding(discrete_codes, channels)
+        self.emb = nn.Conv1d(dvae_channels, channels, kernel_size=1)
         self.norm = normalization(channels)
         self.act = nn.SiLU()
         self.intg = nn.Sequential(nn.Conv1d(channels*2, channels*2, kernel_size=1),
@@ -30,11 +30,10 @@ class DiscreteSpectrogramConditioningBlock(nn.Module):
     :param x: bxcxS waveform latent
     :param codes: bxN discrete codes, N <= S
     """
-    def forward(self, x, codes):
-        _, c, S = x.shape
-        b, N = codes.shape
-        assert N <= S
-        emb = self.emb(codes).permute(0,2,1)
+    def forward(self, x, dvae_in):
+        b, c, S = x.shape
+        _, q, N = dvae_in.shape
+        emb = self.emb(dvae_in)
         emb = nn.functional.interpolate(emb, size=(S,), mode='nearest')
         together = torch.cat([self.act(self.norm(x)), emb], dim=1)
         together = self.intg(together)
@@ -77,7 +76,7 @@ class DiffusionVocoderWithRef(nn.Module):
             model_channels,
             in_channels=1,
             out_channels=2,  # mean and variance
-            discrete_codes=8192,
+            discrete_codes=512,
             dropout=0,
             # res           1, 2, 4, 8,16,32,64,128,256,512, 1K, 2K
             channel_mult=  (1,1.5,2, 3, 4, 6, 8, 12, 16, 24, 32, 48),
@@ -339,7 +338,8 @@ def register_unet_diffusion_vocoder_with_ref(opt_net, opt):
 # Test for ~4 second audio clip at 22050Hz
 if __name__ == '__main__':
     clip = torch.randn(2, 1, 40960)
-    spec = torch.randint(8192, (2, 40,))
+    #spec = torch.randint(8192, (2, 40,))
+    spec = torch.randn(8,512,160)
     cond = torch.randn(2, 3, 80, 173)
     ts = torch.LongTensor([555, 556])
     model = DiffusionVocoderWithRef(32, conditioning_inputs_provided=False)