Skip to content

Open-Sora: Democratizing Efficient Video Production for All

Python

debug

shenchenhui0aec391d805f
diff --git a/opensora/models/vae/vae_3d.py b/opensora/models/vae/vae_3d.py
index 8bdce80..6c9c4f3 100644
--- a/opensora/models/vae/vae_3d.py
+++ b/opensora/models/vae/vae_3d.py
@@ -41,9 +41,9 @@ class ResBlock(nn.Module):
self.use_conv_shortcut = use_conv_shortcut
# SCH: MAGVIT uses GroupNorm by default
- self.norm1 = nn.GroupNorm(num_groups, in_out_channels, dtype=dtype)
+ self.norm1 = nn.GroupNorm(num_groups, in_out_channels, dtype=dtype, device=self.device)
self.conv1 = conv_fn(in_out_channels, self.filters, kernel_size=(3, 3, 3), bias=False)
- self.norm2 = nn.GroupNorm(num_groups, self.filters, dtype=dtype)
+ self.norm2 = nn.GroupNorm(num_groups, self.filters, dtype=dtype, device=self.device)
self.conv2 = conv_fn(self.filters, self.filters, kernel_size=(3, 3, 3), bias=False)
if self.use_conv_shortcut:
self.conv3 = conv_fn(self.filters, self.filters, kernel_size=(3, 3, 3), bias=False)
@@ -170,7 +170,7 @@ class Encoder(nn.Module):
prev_filters = filters # update in_channels
# MAGVIT uses Group Normalization
- self.norm1 = nn.GroupNorm(self.num_groups, prev_filters, dtype=dtype) # SCH: separate <prev_filters> channels into 32 groups
+ self.norm1 = nn.GroupNorm(self.num_groups, prev_filters, dtype=dtype, device = self.device) # SCH: separate <prev_filters> channels into 32 groups
self.conv2 = self.conv_fn(prev_filters, self.embedding_dim, kernel_size=(1, 1, 1))
@@ -308,7 +308,7 @@ class Decoder(nn.Module):
else:
raise NotImplementedError(f'Unknown upsampler: {self.upsample}')
- self.norm1 = nn.GroupNorm(self.num_groups, prev_filters, dtype=dtype)
+ self.norm1 = nn.GroupNorm(self.num_groups, prev_filters, dtype=dtype, device=self.device)
self.conv2 = self.conv_fn(prev_filters, self.output_dim, kernel_size=(3, 3, 3))