mirror of
https://github.com/storytold/vits-finetuning.git
synced 2026-10-09 00:09:52 +00:00
ZDisket
74 lines
2.1 KiB
JSON
74 lines
2.1 KiB
JSON
{
|
|
"train": {
|
|
"log_interval": 200,
|
|
"eval_interval": 3000,
|
|
"seed": 1234,
|
|
"epochs": 20000,
|
|
"learning_rate": 2e-4,
|
|
"betas": [0.8, 0.99],
|
|
"eps": 1e-9,
|
|
"batch_size": 32,
|
|
"fp16_run": true,
|
|
"lr_decay": 0.999875,
|
|
"segment_size": 8192,
|
|
"init_lr_ratio": 1,
|
|
"warmup_epochs": 0,
|
|
"c_mel": 45,
|
|
"c_kl": 1.0
|
|
},
|
|
"data": {
|
|
"training_files":"filelists/train_filelist.txt.cleaned",
|
|
"validation_files":"filelists/val_filelist.txt.cleaned",
|
|
"text_cleaners":["english_cleaners2"],
|
|
"max_wav_value": 32768.0,
|
|
"sampling_rate": 44100,
|
|
"filter_length": 2048,
|
|
"hop_length": 512,
|
|
"win_length": 2048,
|
|
"n_mel_channels": 80,
|
|
"mel_fmin": 20,
|
|
"mel_fmax": 11025,
|
|
"add_blank": true,
|
|
"n_speakers": 0,
|
|
"cleaned_text": true
|
|
},
|
|
"model": {
|
|
"inter_channels": 192,
|
|
"hidden_channels": 192,
|
|
"filter_channels": 768,
|
|
"n_heads": 2,
|
|
"n_layers": 6,
|
|
"kernel_size": 3,
|
|
"p_dropout": 0.1,
|
|
"resblock": "1",
|
|
"resblock_kernel_sizes": [3,7,11],
|
|
"resblock_dilation_sizes": [[1,3,5], [1,3,5], [1,3,5]],
|
|
"upsample_rates": [8,8],
|
|
"upsample_initial_channel": 512,
|
|
"upsample_kernel_sizes": [16,16],
|
|
"n_layers_q": 3,
|
|
"use_spectral_norm": false,
|
|
"gen_istft_n_fft" : 32,
|
|
"gen_istft_hop_size": 8
|
|
},
|
|
"disc": {
|
|
"combd_channels": [16, 64, 256, 1024, 1024, 1024],
|
|
"combd_kernels" : [[7, 11, 11, 11, 11, 5], [11, 21, 21, 21, 21, 5], [15, 41, 41, 41, 41, 5]],
|
|
"combd_groups" : [1, 4, 16, 64, 256, 1],
|
|
"combd_strides": [1, 1, 4, 4, 4, 1],
|
|
|
|
"tkernels" : [7, 5, 3],
|
|
"fkernel" : 5,
|
|
"tchannels" : [64, 128, 256, 256, 256],
|
|
"fchannels" : [32, 64, 128, 128, 128],
|
|
"tstrides" : [[1, 1, 3, 3, 1], [1, 1, 3, 3, 1], [1, 1, 3, 3, 1]],
|
|
"fstride" : [1, 1, 3, 3, 1],
|
|
"tdilations" : [[[5, 7, 11], [5, 7, 11], [5, 7, 11], [5, 7, 11], [5, 7, 11], [5, 7, 11]], [[3, 5, 7], [3, 5, 7], [3, 5, 7], [3, 5, 7], [3, 5, 7]], [[1, 2, 3], [1, 2, 3], [1, 2, 3], [1, 2, 3], [1, 2, 3]]],
|
|
"fdilations" : [[1, 2, 3], [1, 2, 3], [1, 2, 3], [2, 3, 5], [2, 3, 5]],
|
|
"pqmf_n" : 16,
|
|
"pqmf_m" : 64,
|
|
"freq_init_ch" : 128,
|
|
"tsubband" : [6, 11, 16]
|
|
}
|
|
}
|