[TTS]Fix losses of StarGAN v2 VC

examples/vctk/vc3/local/train.sh

@@ -8,4 +8,5 @@ python3 ${BIN_DIR}/train.py \
     --dev-metadata=dump/dev/norm/metadata.jsonl \
     --config=${config_path} \
     --output-dir=${train_output_path} \
-    --ngpu=1
+    --ngpu=1 \
+    --speaker-dict=dump/speaker_id_map.txt

paddlespeech/t2s/datasets/am_batch_fn.py

@@ -820,12 +820,13 @@ def __init__(self, latent_dim: int=16, max_mel_length: int=192):
         self.max_mel_length = max_mel_length
 
     def random_clip(self, mel: np.array):
-        # [80, T]
-        mel_length = mel.shape[1]
+        # [T, 80]
+        mel_length = mel.shape[0]
         if mel_length > self.max_mel_length:
             random_start = np.random.randint(0,
                                              mel_length - self.max_mel_length)
-            mel = mel[:, random_start:random_start + self.max_mel_length]
+
+            mel = mel[random_start:random_start + self.max_mel_length, :]
         return mel
 
     def __call__(self, exmaples):
@@ -843,7 +844,6 @@ def starganv2_vc_batch_fn(self, examples):
         mel = [self.random_clip(item["mel"]) for item in examples]
         ref_mel = [self.random_clip(item["ref_mel"]) for item in examples]
         ref_mel_2 = [self.random_clip(item["ref_mel_2"]) for item in examples]
-
         mel = batch_sequences(mel)
         ref_mel = batch_sequences(ref_mel)
         ref_mel_2 = batch_sequences(ref_mel_2)

paddlespeech/t2s/exps/starganv2_vc/train.py

@@ -113,6 +113,16 @@ def train_sp(args, config):
     model_version = '1.0'
     uncompress_path = download_and_decompress(StarGANv2VC_source[model_version],
                                               MODEL_HOME)
+    # 根据 speaker 的个数修改 num_domains
+    # 源码的预训练模型和 default.yaml 里面默认是 20
+    if args.speaker_dict is not None:
+        with open(args.speaker_dict, 'rt', encoding='utf-8') as f:
+            spk_id = [line.strip().split() for line in f.readlines()]
+        spk_num = len(spk_id)
+        print("spk_num:", spk_num)
+        config['mapping_network_params']['num_domains'] = spk_num
+        config['style_encoder_params']['num_domains'] = spk_num
+        config['discriminator_params']['num_domains'] = spk_num
 
     generator = Generator(**config['generator_params'])
     mapping_network = MappingNetwork(**config['mapping_network_params'])
@@ -123,7 +133,7 @@ def train_sp(args, config):
     jdc_model_dir = os.path.join(uncompress_path, 'jdcnet.pdz')
     asr_model_dir = os.path.join(uncompress_path, 'asr.pdz')
 
-    F0_model = JDCNet(num_class=1, seq_len=192)
+    F0_model = JDCNet(num_class=1, seq_len=config['max_mel_length'])
     F0_model.set_state_dict(paddle.load(jdc_model_dir)['main_params'])
     F0_model.eval()
 
@@ -234,6 +244,11 @@ def main():
     parser.add_argument("--output-dir", type=str, help="output dir.")
     parser.add_argument(
         "--ngpu", type=int, default=1, help="if ngpu == 0, use cpu.")
+    parser.add_argument(
+        "--speaker-dict",
+        type=str,
+        default=None,
+        help="speaker id map file for multiple speaker model.")
 
     args = parser.parse_args()
 

paddlespeech/t2s/models/starganv2_vc/losses.py

@@ -19,35 +19,38 @@
 
 from .transforms import build_transforms
 
-
 # 这些都写到 updater 里
-def compute_d_loss(nets: Dict[str, Any],
-                   x_real: paddle.Tensor,
-                   y_org: paddle.Tensor,
-                   y_trg: paddle.Tensor,
-                   z_trg: paddle.Tensor=None,
-                   x_ref: paddle.Tensor=None,
-                   use_r1_reg: bool=True,
-                   use_adv_cls: bool=False,
-                   use_con_reg: bool=False,
-                   lambda_reg: float=1.,
-                   lambda_adv_cls: float=0.1,
-                   lambda_con_reg: float=10.):
+
+
+def compute_d_loss(
+        nets: Dict[str, Any],
+        x_real: paddle.Tensor,
+        y_org: paddle.Tensor,
+        y_trg: paddle.Tensor,
+        z_trg: paddle.Tensor=None,
+        x_ref: paddle.Tensor=None,
+        # TODO: should be True here, but r1_reg has some bug now 
+        use_r1_reg: bool=False,
+        use_adv_cls: bool=False,
+        use_con_reg: bool=False,
+        lambda_reg: float=1.,
+        lambda_adv_cls: float=0.1,
+        lambda_con_reg: float=10.):
 
     assert (z_trg is None) != (x_ref is None)
     # with real audios
     x_real.stop_gradient = False
     out = nets['discriminator'](x_real, y_org)
     loss_real = adv_loss(out, 1)
-
     # R1 regularizaition (https://arxiv.org/abs/1801.04406v4)
     if use_r1_reg:
         loss_reg = r1_reg(out, x_real)
     else:
-        loss_reg = paddle.to_tensor([0.], dtype=paddle.float32)
+        # loss_reg = paddle.to_tensor([0.], dtype=paddle.float32)
+        loss_reg = paddle.zeros([1])
 
     # consistency regularization (bCR-GAN: https://arxiv.org/abs/2002.04724)
-    loss_con_reg = paddle.to_tensor([0.], dtype=paddle.float32)
+    loss_con_reg = paddle.zeros([1])
     if use_con_reg:
         t = build_transforms()
         out_aug = nets['discriminator'](t(x_real).detach(), y_org)
@@ -118,9 +121,10 @@ def compute_g_loss(nets: Dict[str, Any],
         s_trg = nets['style_encoder'](x_ref, y_trg)
 
     # compute ASR/F0 features (real)
-    with paddle.no_grad():
-        F0_real, GAN_F0_real, cyc_F0_real = nets['F0_model'](x_real)
-        ASR_real = nets['asr_model'].get_feature(x_real)
+    # 源码没有用 .eval(), 使用了 no_grad()
+    # 我们使用了 .eval(), 开启 with paddle.no_grad() 会报错
+    F0_real, GAN_F0_real, cyc_F0_real = nets['F0_model'](x_real)
+    ASR_real = nets['asr_model'].get_feature(x_real)
 
     # adversarial loss
     x_fake = nets['generator'](x_real, s_trg, masks=None, F0=GAN_F0_real)

paddlespeech/t2s/models/starganv2_vc/starganv2_vc_updater.py

@@ -259,7 +259,7 @@ def evaluate_core(self, batch):
             y_org=y_org,
             y_trg=y_trg,
             z_trg=z_trg,
-            use_r1_reg=False,
+            use_r1_reg=self.use_r1_reg,
             use_adv_cls=use_adv_cls,
             **self.d_loss_params)
 
@@ -269,7 +269,7 @@ def evaluate_core(self, batch):
             y_org=y_org,
             y_trg=y_trg,
             x_ref=x_ref,
-            use_r1_reg=False,
+            use_r1_reg=self.use_r1_reg,
             use_adv_cls=use_adv_cls,
             **self.d_loss_params)