Add files via upload

configs/config.json

@@ -0,0 +1,106 @@
+{
+  "train": {
+    "log_interval": 50,
+    "eval_interval": 1000,
+    "seed": 1234,
+    "port": 8001,
+    "epochs": 10000,
+    "learning_rate": 0.0002,
+    "betas": [
+      0.8,
+      0.99
+    ],
+    "eps": 1e-09,
+    "batch_size": 6,
+    "accumulation_steps": 1,
+    "fp16_run": false,
+    "lr_decay": 0.998,
+    "segment_size": 10240,
+    "init_lr_ratio": 1,
+    "warmup_epochs": 0,
+    "c_mel": 45,
+    "keep_ckpts":4
+  },
+  "data": {
+    "data_dir": "dataset",
+    "dataset_type": "SingDataset",
+    "collate_type": "SingCollate",
+    "training_filelist": "filelists/train.txt",
+    "validation_filelist": "filelists/val.txt",
+    "max_wav_value": 32768.0,
+    "sampling_rate": 44100,
+    "n_fft": 2048,
+    "fmin": 0,
+    "fmax": 22050,
+    "hop_length": 512,
+    "win_size": 2048,
+    "acoustic_dim": 80,
+    "c_dim": 256,
+    "min_level_db": -115,
+    "ref_level_db": 20,
+    "min_db": -115,
+    "max_abs_value": 4.0,
+    "n_speakers": 200
+  },
+  "model": {
+    "hidden_channels": 192,
+    "spk_channels": 192,
+    "filter_channels": 768,
+    "n_heads": 2,
+    "n_layers": 4,
+    "kernel_size": 3,
+    "p_dropout": 0.1,
+    "prior_hidden_channels": 192,
+    "prior_filter_channels": 768,
+    "prior_n_heads": 2,
+    "prior_n_layers": 4,
+    "prior_kernel_size": 3,
+    "prior_p_dropout": 0.1,
+    "resblock": "1",
+    "use_spectral_norm": false,
+    "resblock_kernel_sizes": [
+      3,
+      7,
+      11
+    ],
+    "resblock_dilation_sizes": [
+      [
+        1,
+        3,
+        5
+      ],
+      [
+        1,
+        3,
+        5
+      ],
+      [
+        1,
+        3,
+        5
+      ]
+    ],
+    "upsample_rates": [
+      8,
+      8,
+      4,
+      2
+    ],
+    "upsample_initial_channel": 256,
+    "upsample_kernel_sizes": [
+      16,
+      16,
+      8,
+      4
+    ],
+    "n_harmonic": 64,
+    "n_bands": 65
+  },
+  "spk": {
+    "jishuang": 0,
+    "huiyu": 1,
+    "nen": 2,
+    "paimon": 3,
+    "yunhao": 4
+  }
+}

dataset_raw/wav_structure.txt

@@ -0,0 +1,20 @@
+数据集准备
+
+raw
+├───speaker0
+│   ├───xxx1-xxx1.wav
+│   ├───...
+│   └───Lxx-0xx8.wav
+└───speaker1
+    ├───xx2-0xxx2.wav
+    ├───...
+    └───xxx7-xxx007.wav
+
+此外还需要编辑config.json
+
+"n_speakers": 10
+
+"spk":{
+    "speaker0": 0,
+    "speaker1": 1,
+}

filelists/test.txt

@@ -0,0 +1,4 @@
+./dataset/44k/taffy/000562.wav
+./dataset/44k/nyaru/000011.wav
+./dataset/44k/nyaru/000008.wav
+./dataset/44k/taffy/000563.wav

filelists/train.txt

@@ -0,0 +1,15 @@
+./dataset/44k/taffy/000549.wav
+./dataset/44k/nyaru/000004.wav
+./dataset/44k/nyaru/000006.wav
+./dataset/44k/taffy/000551.wav
+./dataset/44k/nyaru/000009.wav
+./dataset/44k/taffy/000561.wav
+./dataset/44k/nyaru/000001.wav
+./dataset/44k/taffy/000553.wav
+./dataset/44k/nyaru/000002.wav
+./dataset/44k/taffy/000560.wav
+./dataset/44k/taffy/000557.wav
+./dataset/44k/nyaru/000005.wav
+./dataset/44k/taffy/000554.wav
+./dataset/44k/taffy/000550.wav
+./dataset/44k/taffy/000559.wav

filelists/val.txt

@@ -0,0 +1,4 @@
+./dataset/44k/nyaru/000003.wav
+./dataset/44k/nyaru/000007.wav
+./dataset/44k/taffy/000558.wav
+./dataset/44k/taffy/000556.wav