diffuserfix

audioldm/variational_autoencoder/autoencoder.py

@@ -40,7 +40,7 @@ class AutoencoderKL(nn.Module):
         self.quant_conv = torch.nn.Conv2d(2 * ddconfig["z_channels"], 2 * embed_dim, 1)
         self.post_quant_conv = torch.nn.Conv2d(embed_dim, ddconfig["z_channels"], 1)
 
-        self.vocoder = BigVGANVocoder(device=device)
+        self.vocoder = BigVGANVocoder(device="cpu")
         #self.vocoder = get_vocoder(None, "cpu")
         self.embed_dim = embed_dim