vits inference success run

decoder/interface.py

@@ -2,7 +2,6 @@
 from .taco_ar.model import Model as TacoAR
 from .taco_mol.model import MelDecoderMOLv2 as TacoMOL
 from .vits.models import VITS
-from .vits.utils import load_checkpoint as load_vits_checkpoint
 from .grad_tts.grad_tts_model import GradTTS
 import torch
 import yaml
@@ -23,8 +22,12 @@ def load_VITS(ckpt = None, config = None, device = 'cpu'):
     with open(config) as f:
         model_config = yaml.safe_load(f)
         f.close()
-    model = VITS(model_config)    
-    model = load_vits_checkpoint(ckpt, model, None)
+    model = VITS(model_config['decoder_params'])    
+    params = torch.load(ckpt, map_location = torch.device(device))
+    params = params['model']
+    params = remove_module_from_state_dict(params)
+    model.load_state_dict(params)
+
     generator = model.generator
     generator.dec.remove_weight_norm()
     generator.to(device)

decoder/vits/models.py

@@ -427,9 +427,11 @@ def forward(self, spec, spec_lengths, ling, spk, pros):
         return o, ids_slice, spec_mask, (z, z_p, m_p, logs_p, m_q, logs_q)
 
     def infer(self, ling, lengths, pros, spk):
+        if self.prosodic_net is not None and pros is not None:
+            pros = self.prosodic_net(pros)
         z_p, m_p, logs_p, mask = self.enc_p(ling, lengths, pros)
         z = self.flow(z_p, mask, g=spk, reverse=True)
-        o = self.dec((z * mask), g=spk, pros = pros)
+        o = self.dec((z * mask), g=spk)
         return o 
 
 

submit_inference.sh

@@ -6,11 +6,11 @@ split=eval_all
 ling_enc=vqwav2vec
 spk_enc=uttdvec
 pros_enc=ppgvcf0
-dec=gradtts
-vocoder=ppgvchifigan
+dec=vits
+vocoder=none
 
 # exp setup
-exp_name=vctk_first_train
+exp_name=vctk_no16fp_split
 exp_dir=exp/${dataset}_${ling_enc}_${spk_enc}_${pros_enc}_${dec}_${vocoder}/${exp_name}