Revert unresolved changes in Bias initialization
it should be zeros_ or parameterized in future properly.
This commit is contained in:
parent
cc56df996e
commit
029d7c7543
|
@ -75,7 +75,7 @@ class HypernetworkModule(torch.nn.Module):
|
|||
w, b = layer.weight.data, layer.bias.data
|
||||
if weight_init == "Normal" or type(layer) == torch.nn.LayerNorm:
|
||||
normal_(w, mean=0.0, std=0.01)
|
||||
normal_(b, mean=0.0, std=0.005)
|
||||
normal_(b, mean=0.0, std=0)
|
||||
elif weight_init == 'XavierUniform':
|
||||
xavier_uniform_(w)
|
||||
zeros_(b)
|
||||
|
|
Loading…
Reference in New Issue
Block a user