DiffWave-Vocoder/config_64.json at master · philsyn/DiffWave-Vocoder · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
{
    "diffusion_config":{
        "T": 50,
        "beta_0": 0.0001,
        "beta_T": 0.05
    },
    "wavenet_config": {
        "in_channels": 1,
        "res_channels": 64,
        "skip_channels": 64,
        "out_channels": 1,
        "num_res_layers": 30,
        "dilation_cycle": 10,
        "diffusion_step_embed_dim_in": 128,
        "diffusion_step_embed_dim_mid": 512,
        "diffusion_step_embed_dim_out": 512
    },
    "train_config": {
        "output_directory": "logs/checkpoint",
        "tensorboard_directory": "logs/tensorboard",
        "ckpt_iter": "max",
        "iters_per_ckpt": 10000,
        "iters_per_logging": 100,
        "n_iters": 1000001,
        "learning_rate": 2e-4,
        "batch_size_per_gpu": 2
    },
    "trainset_config": {
        "segment_length": 16000,
        "data_path": "/tmp2/LJSpeech-1.1/train",
        "valid": false,
        "sampling_rate": 22050,
        "filter_length": 1024,
        "hop_length": 256,
        "win_length": 1024,
        "mel_fmin": 0.0,
        "mel_fmax": 8000.0
    },
    "gen_config":{
        "tensorboard_directory": "logs/tensorboard",
        "mel_path": "./mel_spectrogram",
        "output_directory": "speeches",
        "ckpt_path": "logs/checkpoint"
    },
    "dist_config": {
        "dist_backend": "nccl",
        "dist_url": "tcp://localhost:54321"
    }
}