Upload config.json

configs/config.json

@@ -1,17 +1,18 @@
 {
     "train": {
-        "log_interval": 200,
-        "eval_interval": 1800,
+        "log_interval": 1000,
+        "eval_interval": 2000,
         "seed": 1234,
         "epochs": 10000,
-        "learning_rate": 0.0001,
+        "learning_rate": 0.00015,
         "betas": [
             0.8,
             0.99
         ],
         "eps": 1e-09,
-        "batch_size": 32,
+        "batch_size": 5,
         "fp16_run": false,
+        "half_type": "fp16",
         "lr_decay": 0.999875,
         "segment_size": 10240,
         "init_lr_ratio": 1,
@@ -21,7 +22,7 @@
         "use_sr": true,
         "max_speclen": 512,
         "port": "8001",
-        "keep_ckpts": 40,
+        "keep_ckpts": 50,
         "all_in_mem": false,
         "vol_aug": true
     },
@@ -35,7 +36,8 @@
         "win_length": 2048,
         "n_mel_channels": 80,
         "mel_fmin": 0.0,
-        "mel_fmax": 22050
+        "mel_fmax": 22050,
+        "unit_interpolate_mode": "nearest"
     },
     "model": {
         "inter_channels": 192,
@@ -84,13 +86,18 @@
             4
         ],
         "n_layers_q": 3,
+        "n_flow_layer": 4,
         "use_spectral_norm": false,
         "gin_channels": 768,
         "ssl_dim": 768,
-        "n_speakers": 62,
+        "n_speakers": 61,
+        "vocoder_name": "nsf-hifigan",
         "speech_encoder": "vec768l12",
         "speaker_embedding": false,
-        "vol_embedding": true
+        "vol_embedding": true,
+        "use_depthwise_conv": false,
+        "flow_share_parameter": false,
+        "use_automatic_f0_prediction": true
     },
     "spk": {
         "AKIMOTO_MANATSU": 0,
@@ -153,7 +160,6 @@
         "WATANABE_MIRIA": 57,
         "YAKUBO_MIO": 58,
         "YAMASHITA_MIZUKI": 59,
-        "YAMAZAKI_RENA": 60,
-        "YODA_YUUKI": 61
+        "YODA_YUUKI": 60
     }
 }