ref: d54b9fb49af339c8ee72a8f54ee7e5beadbd724f
parent: fb570ed8bb2648e07e84faf40f30d93b7a0311d7
author: Jean-Marc Valin <jmvalin@amazon.com>
date: Tue Sep 5 08:16:45 EDT 2023
Adds skip connections
--- a/dnn/torch/fargan/fargan.py
+++ b/dnn/torch/fargan/fargan.py
@@ -140,7 +140,7 @@
print("has_gain:", self.has_gain) print("passthrough_size:", self.passthrough_size)- self.sig_dense1 = nn.Linear(4*self.subframe_size+self.passthrough_size+self.cond_size+4, self.cond_size, bias=False)
+ self.sig_dense1 = nn.Linear(4*self.subframe_size+self.passthrough_size+self.cond_size, self.cond_size, bias=False)
self.sig_dense2 = nn.Linear(self.cond_size, self.cond_size, bias=False)
self.gru1 = nn.GRUCell(self.cond_size, self.cond_size, bias=False)
self.gru2 = nn.GRUCell(self.cond_size, self.cond_size, bias=False)
@@ -151,11 +151,11 @@
self.gru1_glu = GLU(self.cond_size)
self.gru2_glu = GLU(self.cond_size)
self.gru3_glu = GLU(self.cond_size)
- self.ptaps_dense = nn.Linear(self.cond_size, 5)
+ self.ptaps_dense = nn.Linear(4*self.cond_size, 5)
- self.sig_dense_out = nn.Linear(self.cond_size, self.subframe_size+self.passthrough_size, bias=False)
+ self.sig_dense_out = nn.Linear(4*self.cond_size, self.subframe_size+self.passthrough_size, bias=False)
if self.has_gain:
- self.gain_dense_out = nn.Linear(self.cond_size, 1)
+ self.gain_dense_out = nn.Linear(4*self.cond_size, 1)
self.apply(init_weights)
@@ -173,30 +173,35 @@
pred = pred/(1e-5+gain)
prev = prev/(1e-5+gain)
- #prev = prev*0
dump_signal(prev, 'pitch_exc.f32')
dump_signal(exc_mem, 'exc_mem.f32')
passthrough = states[3]
- tmp = torch.cat((cond, pred, prev, passthrough, phase), 1)
+ tmp = torch.cat((cond, pred[:,2:-2], prev, passthrough, phase), 1)
tmp = self.dense1_glu(torch.tanh(self.sig_dense1(tmp)))
- tmp = self.dense2_glu(torch.tanh(self.sig_dense2(tmp)))
- gru1_state = self.gru1(tmp, states[0])
- gru2_state = self.gru2(self.gru1_glu(gru1_state), states[1])
- gru3_state = self.gru3(self.gru2_glu(gru2_state), states[2])
+ dense2_out = self.dense2_glu(torch.tanh(self.sig_dense2(tmp)))
+ gru1_state = self.gru1(dense2_out, states[0])
+ gru1_out = self.gru1_glu(gru1_state)
+ #gru1_out = torch.cat([gru1_out, fpitch], 1)
+ gru2_state = self.gru2(gru1_out, states[1])
+ gru2_out = self.gru2_glu(gru2_state)
+ #gru2_out = torch.cat([gru2_out, fpitch], 1)
+ gru3_state = self.gru3(gru2_out, states[2])
gru3_out = self.gru3_glu(gru3_state)
+ gru3_out = torch.cat([gru1_out, gru2_out, gru3_out, dense2_out], 1)
sig_out = torch.tanh(self.sig_dense_out(gru3_out))
if self.passthrough_size != 0:
passthrough = sig_out[:,self.subframe_size:]
sig_out = sig_out[:,:self.subframe_size]
dump_signal(sig_out, 'exc_out.f32')
+ taps = self.ptaps_dense(gru3_out)
+ taps = .2*taps + torch.exp(taps)
+ taps = taps / (1e-2 + torch.sum(torch.abs(taps), dim=-1, keepdim=True))
+ dump_signal(taps, 'taps.f32')
+ fpitch = taps[:,0:1]*pred[:,:-4] + taps[:,1:2]*pred[:,1:-3] + taps[:,2:3]*pred[:,2:-2] + taps[:,3:4]*pred[:,3:-1] + taps[:,4:]*pred[:,4:]
+
if self.has_gain:
- taps = self.ptaps_dense(gru3_out)
- taps = .2*taps + torch.exp(taps)
- taps = taps / (1e-2 + torch.sum(torch.abs(taps), dim=-1, keepdim=True))
- dump_signal(taps, 'taps.f32')
- fpitch = taps[:,0:1]*pred[:,:-4] + taps[:,1:2]*pred[:,1:-3] + taps[:,2:3]*pred[:,2:-2] + taps[:,3:4]*pred[:,3:-1] + taps[:,4:]*pred[:,4:]
pitch_gain = torch.exp(self.gain_dense_out(gru3_out))
dump_signal(pitch_gain, 'pgain.f32')
sig_out = (sig_out + pitch_gain*fpitch) * gain
--
⑨